New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium 9780253007797, 0253007798

Citation preview

paleontology

Michael J. Ryan is Vice-Chair Curator and Head of Vertebrate Paleontology at the Cleveland Museum of Natural History.

Easily distinguished by the horns and frills on their skulls, ceratopsids were one of the most successful of all dinosaurs. This volume presents a broad range of cutting-edge research on the functional biology, behavior, systematics, paleoecology, and paleogeography of the horned dinosaurs, and includes descriptions of newly identified species.

“New Perspectives on Horned Dinosaurs records a landmark event and makes clear that our understanding of this group is undergoing truly explosive growth. To give just one measure, the number of ceratopsids discussed at this meeting represented a doubling of species compared to a comprehensive review of this clade published just three years earlier. The remarkable abundance of newly discovered forms was augmented by presentation of rigorous studies of stratigraphy, phylogeny, ontogeny, biomechanics, taphonomy, paleogeography, and paleoenvironment. These results, including descriptions of ten new taxa, are captured in this volume, which will be a must-own for dinosaur paleontologists and enthusiasts alike.” —Scott Sampson, University of Utah

r ya n chinner yallgeier eberth

New Perspectives on

Horned Dinosaurs The Royal Tyrrell Museum Ceratopsian Symposium

New Perspectives on

Horned dinosaurs

—Larry Witmer, Ohio University

“From Archaeoceratops to Zuniiceratops, from Alaska to Mexico, and from sediments to functional morphology, this book covers much of present-day research on ceratopsians. These horned dinosaurs are rendered as living, behaving, and evolving organisms throughout the thirty-six chapters of this book. I encourage everyone interested in how a myriad of incredible fossils can inform about life of the past to read it.”

Ceratopsids, or horned dinosaurs, are a group of large-bodied, quadruped herbivores that lived roughly 65–70 million years ago. Part of a larger group of dinosaurs that include stegosaurs, ankylosaurs, ornithopods, and pachychephalosaurs, the better known members of the ceratopsids include centrosaurs, chasmosaurs, and triceratopsians. All are easily distinguished by the horns and frills on their skulls; in fact, ceratopsids have among the largest, most elaborate skulls found in vertebrates. They were one of the most successful of all dinosaurs and their remains are well known from many locations in the United States, Canada, and Mexico. They died out at the end of the Cretaceous along with the rest of the dinosaurs.

—David Weishampel, co-editor of The Dinosauria and co-author of Dinosaurs: A Concise Natural History “This book captures an explosion of new and exciting research on one of the most fascinating groups of dinosaurs. It will be a landmark in the study of ceratopsians.”

Brenda J. Chinnery-Allgeier is Lecturer in the School of Biological Sciences at The University of Texas at Austin.

—David C. Evans, University of Toronto

Edited by m i c h a e l j

.

r ya n ,

b r e n da j . c h i n n e r y - a l l g e i e r ,

indiana

INDIANA David A. Eberth is a senior research scientist at the Royal Tyrrell Museum in Drumheller, Alberta, Canada.

“Triceratops and its kin may hail from the dim and distant past, but this new volume brings them fully into the light of today. An all-star and comprehensive list of authors not only effectively puts horned dinosaurs in the context of their own time and place, but also brings them alive as living, breathing biological organisms. New Perspectives on Horned Dinosaurs is able proof of the vitality of modern dinosaur science, bringing to bear twenty-first-century ideas and approaches to ask—and answer—questions that once would have been thought to be out of reach.”

The outgrowth of the Royal Tyrrell Museum’s Ceratopsian Symposium, this volume presents a broad range of cutting-edge research on the functional biology and behavior, systematics, paleoecology, and paleogeography of the horned dinosaurs, including descriptions of newly identified species. There is also a history of collecting these dinosaurs, plus a supplementary CD-ROM containing a history of ceratopsian discoveries in Canada and a list of the specimens recovered to date.

a n d dav i d a . e b e r t h

University Press

Bloomington & Indianapolis www.iupress.indiana.edu 1-800-842-6796

Life of the Past

James O. Farlow, editor

s u p p l e m e n ta ry c d - r o m i n c l u d e d

Cover illustration: ©2009 Donna Sloan

NEW PERSPECTIVES ON

HORNED DINOSAURS

L I F E O F T H E PA S T

James O. Farlow, editor

NEW PERSPECTIVES ON

HORNED DINOSAURS The Royal Tyrrell Museum Ceratopsian Symposium

EDITED BY

M I C H A E L J . R YA N B R E N D A J . C H I N N E R Y- A L L G E I E R D AV I D A . E B E R T H Indiana University Press Bloomington and Indianapolis

PAT R I C I A E . R A L R I C K E D I T O R I A L A S S I S TA N T

This book is a publication of Indiana University Press 601 North Morton Street Bloomington, IN 47404-3797 USA www.iupress.indiana.edu Telephone orders Fax orders Orders by e-mail

800-842-6796 812-855-7931 [email protected]

∫ 2010 by Indiana University Press All rights reserved No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage and retrieval system, without permission in writing from the publisher. The Association of American University Presses’ Resolution on Permissions constitutes the only exception to this prohibition. ! The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1992. Manufactured in the United States of America Library of Congress Cataloging-in-Publication Data Royal Tyrrell Museum Ceratopsian Symposium (2007 : Drumheller, Alta.) New perspectives on horned dinosaurs : the Royal Tyrrell Museum Ceratopsian Symposium / edited by Michael J. Ryan, Brenda J. Chinnery-Allgeier, and David A. Eberth ; Patricia E. Ralrick, editorial assistant. p. cm. Includes bibliographical references and index. ISBN 978-0-253-35358-0 (cloth : alk. paper) 1. Ceratopsidae—Congresses. I. Ryan, Michael J., [date]II. Chinnery-Allgeier, Brenda J. III. Eberth, David A. IV. Title. QE862.O65R695 2010 567.915—dc22 2009019913 1 2 3 4 5 15 14 13 12 11 10

this volume is dedicated to the late

Halska Osmólska, whose work on basal neoceratopsians set a standard for excellence, and to

Wann Langston, Jr., for his work on horned dinosaurs that now spans more than five decades and continues to inspire new research.

What seest thou else in the dark backward and abysm of time? Shakespeare, The Tempest, act I, scene 2

CONTENTS Preface xiii Acknowledgments xv List of Contributors xvii List of Reviewers xxi

PART ONE § OVERVIEW

1. Forty Years of Ceratophilia / P E T E R

DODSON

3

PART TWO § SYSTEMATICS AND NEW CERATOPSIANS

2. Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs (Ceratopsia: Psittacosaurus) / PA U L C . S E R E N O 21 3. A New Species of Archaeoceratops (Dinosauria: Neoceratopsia) from the Early Cretaceous of the Mazongshan Area, Northwestern China / H A I - L U Y O U , K Y O TA N O U E , A N D P E T E R D O D S O N 59 4. A Redescription of the Montanoceratops cerorhynchus Holotype with a Review of Referred Material / P E T E R J . M A K O V I C K Y 68 5. First Basal Neoceratopsian from the Oldman Formation (Belly River Group), Southern Alberta / T E T S U T O M I YA S H I TA , P H I L I P J . C U R R I E , A N D B R E N D A J . C H I N N E R Y- A L L G E I E R

83

6. Zuniceratops christopheri: The North American Ceratopsid Sister Taxon Reconstructed on the Basis of New Data / D O U G L A S G . W O L F E , J A M E S I . K I R K L A N D , D AV I D S M I T H , K A R E N P O O L E , B R E N D A J . C H I N N E R Y- A L L G E I E R , A N D A N D R E W M C D O N A L D 91 7. Horned Dinosaurs (Ornithischia: Ceratopsidae) from the Upper Cretaceous (Campanian) Cerro del Pueblo Formation, Coahuila, Mexico / M A R K A . L O E W E N , S C O T T D . S A M P S O N , ERIC K. LUND, ANDREW A. FARKE, MARTHA C. AGUILLÓN-MARTÍNEZ, CLAUDIO A. DE LEON, R U B É N A . R O D R Í G U E Z - D E L A R O S A , M I C H A E L A . G E T T Y, A N D D AV I D A . E B E R T H

99

8. New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation (Middle Campanian), Grand Staircase–Escalante National Monument, Southern Utah / J A M E S I . K I R K L A N D A N D D O N A L D D . D E B L I E U X 117 9. A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation (Late Campanian) of Southern Alberta, Canada / M I C H A E L J . R YA N , D AV I D A . E B E R T H , D O N A L D B . B R I N K M A N , P H I L I P J . C U R R I E , A N D D A R R E N H . TA N K E 141 10. New Material of ‘‘Styracosaurus’’ ovatus from the Two Medicine Formation of Montana / A N D R E W T. M C D O N A L D A N D J O H N R . H O R N E R 156 11. A New Chasmosaurine (Ceratopsidae, Dinosauria) from the Upper Cretaceous Ojo Alamo Formation (Naashoibito Member), San Juan Basin, New Mexico / R O B E R T M . A N D S P E N C E R G . L U C A S 169

S U L L I VA N

12. A New Chasmosaurine Ceratopsid from the Judith River Formation, Montana / M I C H A E L A N T H O N Y P. R U S S E L L , A N D S C O T T H A R T M A N 181

J . R YA N ,

13. Description of a Complete and Fully Articulated Chasmosaurine Postcranium Previously Assigned to Anchiceratops (Dinosauria: Ceratopsia) / J O R D A N C . M A L L O N A N D R O B E R T H O L M E S 189 14. A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation, Northwest South Dakota, United States: A Preliminary Description / C H R I S T O P H E R J . O T T A N D P E T E R L . L A R S O N 203

PART THREE § ANATOMY, FUNCTIONAL BIOLOGY, AND BEHAVIOR

15. Comments on the Basicranium and Palate of Basal Ceratopsians / P E T E R K Y O TA N O U E 221 16. Mandibular Anatomy in Basal Ceratopsia / K Y O

DODSON, HAI-LU YOU, AND

TA N O U E , H A I - L U Y O U , A N D P E T E R D O D S O N

234

17. Histological Evaluation of Ontogenetic Bone Surface Texture Changes in the Frill of Centrosaurus apertus / A L L I S O N R . T U M A R K I N - D E R A T Z I A N 251 18. Modeling Structural Properties of the Frill of Triceratops / A N D R E W A R T A N D E R S E N 264

A. FARKE, RALPH E. CHAPMAN, AND

19. New Evidence Regarding the Structure and Function of the Horns in Triceratops (Dinosauria: Ceratopsidae) / J O H N W. H A P P 271 20. Evolutionary Interactions between Horn and Frill Morphology in Chasmosaurine Ceratopsians / D AV I D A . K R A U S S , A N T O I N E P E Z O N , P E T E R N G U Y E N , I S S A S A L A M E , A N D S H A N T I B . R Y W K I N 282 21. Skull Shapes as Indicators of Niche Partitioning by Sympatric Chasmosaurine and Centrosaurine Dinosaurs / D O N A L D M . H E N D E R S O N 293 22. The Function of Large Eyes in Protoceratops: A Nocturnal Ceratopsian? / N I C K 23. A Semi-Aquatic Life Habit for Psittacosaurus / T R A C Y

LONGRICH

308

328

L. FORD AND LARRY D. MARTIN

24. Habitual Locomotor Behavior Inferred from Manual Pathology in Two Late Cretaceous Chasmosaurine Ceratopsid Dinosaurs, Chasmosaurus irvinensis (CMN 41357) and Chasmosaurus belli (ROM 843) / E L I Z A B E T H R E G A , R O B E R T H O L M E S , A N D A L E X T I R A B A S S O 340 25. Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance / D A R R E N B R U C E M . R O T H S C H I L D 355

H . TA N K E A N D

PART FOUR § HORNED DINOSAURS IN TIME AND SPACE

Paleobiogeography, Taphonomy, and Paleoecology 26. An Update on the Paleobiogeography of Ceratopsian Dinosaurs / B R E N D A J A M E S I . K I R K L A N D 387

J . C H I N N E R Y- A L L G E I E R A N D

27. Unraveling a Radiation: A Review of the Diversity, Stratigraphic Distribution, Biogeography, and Evolution of Horned Dinosaurs (Ornithischia: Ceratopsidae) / S C O T T D . S A M P S O N A N D M A R K A . L O E W E N 405 28. A Review of Ceratopsian Paleoenvironmental Associations and Taphonomy / D AV I D 29. Behavioral Interpretations from Ceratopsid Bonebeds / R E B E C C A

A. EBERTH

428

K. HUNT AND ANDREW A. FARKE

447

30. Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry (Prince Creek Formation: Late Cretaceous), Northern Alaska: A Multi-Disciplinary Study of a High-Latitude Ceratopsian Dinosaur Bonebed / A N T H O N Y R . F I O R I L L O , PA U L J . M C C A R T H Y, P E T E R P. F L A I G , E R I K B R A N D L E N , D AV I D W. N O R T O N , P I E R R E Z I P P I , L O U I S J A C O B S , A N D R O L A N D A . G A N G L O F F 456

x contents

31. Taphonomy of Horned Dinosaurs (Ornithischia: Ceratopsidae) from the Late Campanian Kaiparowits Formation, Grand Staircase–Escalante National Monument, Utah / M I K E A . G E T T Y, M A R K A . L O E W E N , E R I C R O B E R T S , A L A N L . T I T U S , A N D S C O T T D . S A M P S O N 478 32. A Centrosaurine Mega-Bonebed from the Upper Cretaceous of Southern Alberta: Implications for Behavior and Death Events / D AV I D A . E B E R T H , D O N A L D B . B R I N K M A N , A N D VA I A B A R K A S 495 33. Insect Trace Fossils Associated with Protoceratops Carcasses in the Djadokhta Formation (Upper Cretaceous), Mongolia / J A M E S I . K I R K L A N D A N D K E N N E T H B A D E R 509 34. Faunal Composition and Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis and Other Dinosaurs, Aguja Formation (Upper Cretaceous), Big Bend, Texas / J U L I A T. S A N K E Y 520

PART FIVE § HISTORY OF HORNED DINOSAUR COLLECTION

35. Lost in Plain Sight: Rediscovery of William E. Cutler’s Missing Eoceratops / D A R R E N 36. Historical Collecting Bias and the Fossil Record of Triceratops in Montana / M A R K J O H N R . H O R N E R 551 Afterword / P H I L I P J . C U R R I E

Index

H . TA N K E

541

B. GOODWIN AND

565

569

Supplemental CD-ROM 1. A Ceratopsian Compendium / T R A C Y L . F O R D 2. Ceratopsian Discoveries and Work in Alberta, Canada: Historical Review and Census / D A R R E N

H . TA N K E

contents

xi

P R E FA C E

Horned dinosaurs (ceratopsians) are among the best-loved

saurs, and here he reflects on all he has learned over the years,

and better-known groups within the Dinosauria. We are still

while providing historical perspective about those he has ‘‘in-

learning more about them every day, which is why the Cera-

fected’’ with his joy of all things ceratopsian. His chapter elo-

topsian Symposium was convened at the Royal Tyrrell Mu-

quently describes why we study horned dinosaurs, and just

seum of Palaeontology (Drumheller, Alberta) on September

how fascinating they really are.

22–24, 2007. The symposium was a joint venture among three

Part 2, Systematics and New Ceratopsians, includes 13

groups: (1) Don Brinkman, Dave Eberth, and Phil Currie from

chapters that describe new taxa and present phylogenetic

the Royal Tyrrell Museum in Alberta (most of you are aware

analyses. An astonishing and unprecedented 10 new cera-

that Phil is now at the University of Alberta); (2) Michael Ryan

topsian taxa are described, bringing the number of named

from the Cleveland Museum of Natural History; and (3) Bren-

ceratopsian species to around 70! In addition, new informa-

da Chinnery-Allgeier from the University of Texas at Austin.

tion on previously known taxa adds greatly to this section,

With so many new specimens and putatively new taxa start-

including a benchmark paper by Paul Sereno on the system-

ing to appear, as well as the unveiling of a new exhibit dedi-

atic relationships of psittacosaurs.

cated to ceratopsians at the Tyrrell Museum, we realized that it

Part 3, Anatomy, Functional Biology, and Behavior, in-

was an opportune time to bring together the community of

cludes 11 chapters that burrow into the details of new and old

horned-dinosaur workers (and lovers) to share in these discov-

specimens in an attempt to address classic and new questions.

eries, and to explore the significance of the new data. More-

Horn and frill function, patterns of growth, niche partition-

over, with so many graduate students exploring new and old

ing, life habits, and the application of paleopathology to be-

ideas about ceratopsians and applying emerging and new

havior inference are all included here. Many of these studies

technologies in their studies, there was no question that the

employ evolving techniques in histology, computer model-

symposium would be lively, exciting, and challenging.

ing, and engineering that we hope will spark ongoing discus-

The symposium was a resounding success. Over the course

sion and interest. Of particular interest here (and sure to spark

of two days of talks in a filled-to-capacity auditorium (≈190

debate) are the challenges posed by Nick Longrich and Tracy

participants), and a day in the field with the Tyrrell research

Ford and Larry Martin to our notions about the life habits of

team, we learned about recent discoveries, evaluated rein-

Protoceratops and Psittacosaurus.

terpretations about known specimens and localities, and were

Part 4, Horned Dinosaurs in Time and Space: Paleobiogeo-

exposed to new ideas about ceratopsian anatomy and be-

graphy, Taphonomy, and Paleoecology, includes 9 separate

havior. The menu also included biostratigraphy, taphonomy,

discussions about large-scale diversity and migration patterns

and new approaches for studying horned dinosaurs.

among ceratopsians, and where and how ceratopsians lived,

This volume comprises those presentations and more, and

died, and were preserved. Here the chapters by Chinnery-

its size alone should hint at just how much novel information

Allgeier and James Kirkland; Sampson and Loewen; and

was presented over three short days. Given the abundance

Eberth stand out because of their ambitious integrative ap-

and breadth of new information, we chose to call this volume

proaches, whereas the others are particularly satisfying be-

New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum

cause of their more detailed approaches.

Ceratopsian Symposium and organized the contributions into six parts.

Part 5, History of Horned Dinosaur Collection, is the last section of the print volume and includes two chapters that

Part 1, Overview, includes only one chapter, written by the

focus on practical and historical issues surrounding the collec-

irrefutable ‘‘king’’ of ceratopsian dinosaurs, Peter Dodson. Pe-

tion of ceratopsians. Although this section is brief, it presents

ter has spent a large part of his career studying horned dino-

engaging treatments on prospecting and collecting bias in the

xiii

Hell Creek Formation by Goodwin and Horner, and Tanke’s

compile this volume, and its wide range of content. We have

documentation of how a lost specimen from Alberta was

tried our best to present a group of well-balanced and consis-

never really lost at all.

tently edited manuscripts, while allowing the authors to ex-

Part 6 is a CD-ROM that includes Ford’s exhaustive listing of ceratopsian specimens collected and published up to January

press their individual styles. We hope that you all enjoy the volume and find it useful for years to come.

2008, and Tanke’s thorough historical account of ceratopsian discoveries in Alberta, Canada.

Michael J. Ryan

For us, this project was particularly satisfying because of the

Brenda J. Chinnery-Allgeier

great response, the relative ease with which we were able to

David A. Eberth

xiv preface

ACKNOWLEDGMENTS

We thank all of the participants who attended the Ceratopsian

seum, which provided the FTP site and physical space that

Symposium in 2007 and helped us realize that this volume

were critical for the compilation of this volume. Both the

would be a successful venture. We thank the authors for their

Cleveland Museum of Natural History and the Royal Tyrrell

attention to detail and timely responses to our numerous ha-

Museum Cooperating Society provided funds in support of

rangues about deadlines. We also thank the reviewers who did

this project and we thank them for that.

their jobs, ensuring that these contributions were of high

Don Brinkman, Donna Sloan, and George Olshevsky were

quality both scientifically and editorially, and turned the

instrumental in the successful completion and presentation

manuscripts around so efficiently. In particular, we thank

of this volume.

Andy Farke for his diligence and efforts in cheerfully and effectively reviewing so many contributions. Our sincere gratitude goes to Bob Sloan and his team at Indiana University Press for all their help with this project. Our home organizations gave freely of their time, physical resources, and manpower, in particular the Royal Tyrrell Mu-

We especially would like to recognize Patty E. Ralrick for her huge contribution in serving as Editorial Assistant for this project. Without her dedication, consistency, and good humor, this volume would have been significantly delayed, and the sanity of the editors would have been compromised (even more).

xv

CONTRIBUTORS

Martha C. Aguillón-Martínez

Andrew A. Farke

Coordinación de Paleontología, Secretaría de Educación y

Department of Anatomical Sciences, Stony Brook

Cultura, Museo del Desierto, Saltillo, Coahuila, Mexico.

University, T8 040 Health Sciences Center, Stony Brook, NY

Art Andersen Virtual Surfaces, Inc., Mt. Prospect, IL 60056 USA. Kenneth Bader Museum of Natural History, University of Kansas, 1345 Jayhawk Blvd., Lawrence, KS 66045 USA. Vaia Barkas P.O. Box 119, Whitehall, MT 59759 USA. Erik Brandlen Department of Geology and Geophysics, and Geophysical Institute, University of Alaska, P.O. Box 755780, Fairbanks, AK 99775 USA. Donald B. Brinkman Royal Tyrrell Museum of Palaeontology, P.O. Box 7500, Drumheller, AB T0J 0Y0 Canada. Ralph E. Chapman

11794 USA. Anthony R. Fiorillo Museum of Nature and Science, P.O. Box 151469, Dallas, TX 75315 USA. Peter P. Flaig Department of Geology and Geophysics, and Geophysical Institute, University of Alaska, P.O. Box 755780, Fairbanks, AK 99775 USA. Tracy L. Ford P.O. Box 1171, Poway, CA 92074 USA. Roland A. Gangloff Museum of Paleontology, University of California, Berkeley, CA 94720 USA. Michael A. Getty Utah Museum of Natural History and Department of

Eryops Consulting, 295 Bryce Ave., Los Alamos, NM 87544

Geology and Geophysics, University of Utah, 1390 East

USA.

Presidents Circle, Salt Lake City, UT 84112 USA.

Brenda J. Chinnery-Allgeier School of Biological Sciences, University of Texas at Austin, 1 University Station, A6500, Austin, TX 78712 USA. Philip J. Currie Department of Biological Sciences, University of Alberta, Edmonton, AB T6G 2E9 Canada. Donald D. DeBlieux Utah Geological Survey, P.O. Box 146100, Salt Lake City, UT 84114 USA. Claudio A. de Leon

Mark B. Goodwin Museum of Paleontology, University of California, Berkeley, CA 94720 USA. John W. Happ Shenandoah University, 1460 University Dr., Winchester, VA 22601 USA. Scott Hartman Wyoming Dinosaur Center, 110 Carter Ranch Rd., P.O. Box 868, Thermopolis, WY 82441 USA. Donald M. Henderson

Coordinación de Paleontología, Secretaría de Educación y

Royal Tyrrell Museum of Palaeontology, P.O. Box 7500,

Cultura, Museo del Desierto, Saltillo, Coahuila, Mexico.

Drumheller, AB T0J 0Y0 Canada.

Peter Dodson School of Veterinary Medicine, University of Pennsylvania, 3800 Spruce St., Philadelphia, PA 19104 USA. David A. Eberth

Robert Holmes Earth Sciences, Canadian Museum of Nature, P.O. Box 3443, STN ‘‘D,’’ Ottawa, ON K1P 6P4 Canada. John R. Horner

Royal Tyrrell Museum of Palaeontology, P.O. Box 7500,

Museum of the Rockies, Montana State University, 600

Drumheller, AB T0J 0Y0 Canada.

West Kagy Blvd., Bozeman, MT 59717 USA.

xvii

ReBecca K. Hunt Augustana College, 639 38th St., Department of Geology, Rock Island, IL 61201 USA. Louis Jacobs Department of Geosciences, Southern Methodist University, Dallas, TX 75275 USA. James I. Kirkland

Peter Nguyen Science Department, B.M.C.C., City University of New York, 199 Chambers St., New York, NY 10007 USA. David W. Norton Arctic Rim Research, 1749 Red Fox Dr., Fairbanks, AK 99709 USA. Christopher J. Ott

Utah Geological Survey, P.O. Box 146100, Salt Lake City,

Black Hills Institute of Geological Research, P.O. Box 643,

UT 84114 USA.

117 Main St., Hill City, SD 57745 USA.

David A. Krauss Science Department, B.M.C.C., City University of New York, 199 Chambers St., New York, NY 10007 USA. Peter L. Larson Black Hills Institute of Geological Research, P.O. Box 643, 117 Main St., Hill City, SD 57745 USA. Mark A. Loewen Utah Museum of Natural History and Department of Geology and Geophysics, University of Utah, 1390 East Presidents Circle, Salt Lake City, UT 84112 USA. Nick Longrich Department of Biological Sciences, University of Calgary, 2500 University Dr. NW, Calgary, AB T2N 1N4 Canada. Spencer G. Lucas New Mexico Museum of Natural History, 1801 Mountain Rd. NW, Albuquerque, NM 87104 USA. Eric K. Lund Utah Museum of Natural History and Department of Geology and Geophysics, University of Utah, 1390 East Presidents Circle, Salt Lake City, UT 84112 USA. Peter J. Makovicky Department of Geology, Field Museum of Natural History, 1400 S. Lake Shore Dr., Chicago, IL 60605 USA. Jordan C. Mallon Faculty of Veterinary Medicine, University of Calgary, 3330 Hospital Dr., Calgary, AB T2N 4N1 Canada. Larry D. Martin Natural History Museum and Department of Ecology and Evolutionary Biology, University of Kansas, Lawrence, KS 66045 USA. Paul J. McCarthy Department of Geology and Geophysics, and Geophysical

Antoine Pezon Science Department, B.M.C.C., City University of New York, 199 Chambers St., New York, NY 10007 USA. Karen Poole Department of Earth and Planetary Sciences, Washington University, St. Louis, MO 63130 USA. Elizabeth Rega Western University of Health Sciences, 309 E. 2nd St., Pomona, CA 92506 USA. Eric Roberts School of Geosciences, University of Witwatersrand, Private Bag 3, Johannesburg, South Africa. Rubén A. Rodríguez-de la Rosa Secretaría de Educación y Cultura, Saltillo, Coahuila, Mexico. Bruce M. Rothschild Arthritis Center, Biodiversity Institute, University of Kansas, Lawrence, KS 66045 USA. Anthony P. Russell Department of Biological Sciences, University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4 Canada. Michael J. Ryan Department of Vertebrate Paleontology, Cleveland Museum of Natural History, 1 Wade Oval Dr., University Circle, Cleveland, OH 44106 USA. Shanti B. Rywkin Science Department, B.M.C.C., City University of New York, 199 Chambers St., New York, NY 10007 USA. Issa Salame Chemistry Department, the City College of New York, 138th St. at Convent Ave., New York, NY 10031 USA. Scott D. Sampson

Institute, University of Alaska, P.O. Box 755780, Fairbanks,

Utah Museum of Natural History and Department of

AK 99775 USA.

Geology and Geophysics, University of Utah, 1390 East

Andrew McDonald Department of Earth and Environmental Science,

Presidents Circle, Salt Lake City, UT 84112 USA. Julia T. Sankey

University of Pennsylvania, 240 S. 33rd St., Philadelphia,

Department of Physics and Geology, California State

PA 19104 USA.

University, Stanislaus, One University Circle, Turlock, CA

Tetsuto Miyashita Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, AB T6G 2E9 Canada.

95382 USA. Paul C. Sereno Department of Organismal Biology and Anatomy, University of Chicago, 1027 East 57th St., Chicago, IL 60637 USA.

xviii contributors

David Smith Northland Pioneer College, Show Low, AZ 85901 USA. Robert M. Sullivan Section of Paleontology and Geology, the State Museum of Pennsylvania, 300 North St., Harrisburg, PA 17120 USA. Darren H. Tanke Royal Tyrrell Museum of Palaeontology, P.O. Box 7500, Drumheller, AB T0J 0Y0 Canada. Kyo Tanoue Department of Earth and Environmental Science, University of Pennsylvania, 240 S. 33rd St., Philadelphia, PA 19104 USA. Alex Tirabasso

Allison R. Tumarkin-Deratzian Temple University, Department of Geology, Beury Hall (016-00), 1901 North 13th St., Philadelphia, PA 19122 USA. Douglas G. Wolfe New Mexico Museum of Natural History and Science, 1801 Mountain Rd. NW, Albuquerque, NM 87104 USA. Hai-Lu You Institute of Geology, Chinese Academy of Geological Sciences, 26 Baiwanzhuang Rd., Beijing 100037, P. R. China. Pierre Zippi Biostratigraphy.com LLC, 7518 Twin Oaks Ct., Garland, TX 75044 USA.

Canadian Museum of Nature, P.O. Box 3443, Station D, Ottawa ON K1P 6P4 Canada. Alan L. Titus Grand Staircase–Escalante National Monument, 190 E. Center St., Kanab, UT 84741 USA.

contributors

xix

REVIEWERS

Donald B. Brinkman Royal Tyrrell Museum of Palaeontology, P.O. Box 7500, Drumheller, AB T0J 0Y0 Canada. Brooks Britt Brigham Young University, Department of Geology S-387 ESC, P.O. Box 24606, Provo, UT 85602 USA. Kenneth Carpenter Denver Museum of Natural History, Department of Earth Sciences, 2001 Colorado Blvd., Denver, CO 80205 USA. Brenda J. Chinnery-Allgeier

Jerry Harris Dixie State College, Science Bldg., 225 S. 700 East, St. George, UT 84770 USA. Donald M. Henderson Royal Tyrrell Museum of Palaeontology, P.O. Box 7500, Drumheller, AB T0J 0Y0 Canada. Robert Holmes Earth Sciences, Canadian Museum of Nature, P.O. Box 3443, STN ‘‘D,’’ Ottawa, ON K1P 6P4 Canada. ReBecca K. Hunt

School of Biological Sciences, University of Texas at Austin,

Augustana College, 639 38th St., Department of Geology,

1 University Station, A6500, Austin, TX 78712 USA.

Rock Island, IL 61201 USA.

Philip J. Currie Department of Biological Sciences, University of Alberta, Edmonton, AB T6G 2E9 Canada. Peter Dodson School of Veterinary Medicine, University of Pennsylvania, 3800 Spruce St., Philadelphia, PA 19104 USA. David A. Eberth Royal Tyrrell Museum of Palaeontology, P.O. Box 7500, Drumheller, AB T0J 0Y0 Canada. David C. Evans Royal Ontario Museum, Department of Natural History, 100 Queen’s Park, Toronto, ON M5S 2C6 Canada. Andrew A. Farke Department of Anatomical Sciences, Stony Brook

James I. Kirkland Utah Geological Survey, P.O. Box 146100, Salt Lake City, UT 84114 USA. Andrew Lee Ohio University, 228 Irvine Hall, Athens, OH 45701 USA. Peter J. Makovicky Department of Geology, Field Museum of Natural History, 1400 S. Lake Shore Dr., Chicago, IL 60605 USA. Ralph Molnar 402 W. Apache Rd., Flagstaff, AZ 86001 USA. Anthony P. Russell Department of Biological Sciences, University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4 Canada. Michael J. Ryan

University, T8 040 Health Sciences Center, Stony Brook, NY

Department of Vertebrate Paleontology, Cleveland

11794 USA.

Museum of Natural History, 1 Wade Oval Dr., University

Catherine A. Forster George Washington University, Department of Biological Sciences, Washington DC, 20052 USA. Denver Fowler Museum of the Rockies, Paleo Department, 600 W. Kagy Blvd., Bozeman, MT 59717 USA. Michael A. Getty Utah Museum of Natural History and Department of Geology and Geophysics, University of Utah,

Circle, Cleveland, OH 44106 USA. Scott D. Sampson Utah Museum of Natural History and Department of Geology and Geophysics, University of Utah, 1390 East Presidents Circle, Salt Lake City, UT 84112 USA. Craig Scott Royal Tyrrell Museum of Palaeontology, P.O. Box 7500, Drumheller, AB T0J 0Y0 Canada. Eric Snively

1390 East Presidents Circle, Salt Lake City,

Department of Biosciences, CW405 Biosci Centre,

UT 84112 USA.

University of Alberta, Edmonton, AB T6G 2E9 Canada.

xxi

Kent Stevens

Matt Vickaryous

University of Oregon, Department of Computer and

Biomedical Sciences, Ontario Veterinary College,

Information Science, 301 Deschutes Hall, Eugene, OR

University of Guelph, 50 Stone Road, Guelph, ON, N1G

97403 USA. François Therrien

2W1 Canada. Larry M. Witmer

Royal Tyrrell Museum of Palaeontology, P.O. Box 7500,

Department of Biomedical Sciences, College of Osteopathic

Drumheller, AB T0J 0Y0 Canada.

Medicine, Life Sciences Bldg., Rm. 123, Ohio University,

Allison R. Tumarkin-Deratzian Temple University, Department of Geology,

Athens, OH 45701 USA. Xiao-chun Wu

Beury Hall (016-00), 1901 North 13th St.,

Canadian Museum of Nature, P.O. Box 3443, Station D,

Philadelphia, PA 19122 USA.

Ottawa, ON K1P 6P4 Canada.

xxii reviewers

NEW PERSPECTIVES ON

HORNED DINOSAURS

PART ONE OVER VIEW

1 Forty Years of Ceratophilia PETER DODSON

With the death of my beloved and highly esteemed mentor

at the Carnegie Museum. Matt Lamanna is certainly living

John Ostrom (1928–2005), I seem to have become the dean,

proof for the need to take children seriously!

or at least the senior citizen, of ceratopsian studies. Of course

My family moved to Aylmer, Québec, a suburb of Ottawa,

my interest in dinosaurs came from my childhood in the

Canada’s capital, when I was 11 years old. I lived at home and

1950s, at a time when there was not nearly so much dinosaur

attended the University of Ottawa, where my father taught,

‘‘stuff.’’ A vivid early encounter with dinosaurs came when my

enjoying the benefits of home cooking and free tuition. I

mother, a lover of classical music, took me to see Walt Disney’s

date my own entry into the field of paleontology from 1968,

Fantasia. I was enthralled by the paleontology segment that

the year of my graduation from university and my entry into

started with the beginning of unicellular life in the primordial

graduate school at the University of Alberta. Thus my profes-

seas and ended with the death march of the dinosaurs to the

sional activities span parts of five decades and, God willing

stirring chords of Stravinsky’s Sacre du Printemps. Growing up

(shudder!), soon enough to be six decades. Moreover, my per-

in South Bend, Indiana, we also visited the Field Museum in

sonal acquaintances in paleontology range not only to the

Chicago, where I remember the mummies but not the dino-

beginning of the twentieth century in the persons of such

saurs. I really liked the coal mine at the Museum of Science

luminaries as George Gaylord Simpson (b. 1902); E. H. Colbert

and Industry! My father, a professor of evolutionary biology,

(b. 1905); E. C. Olson (b. 1910), but back into the nineteenth

nourished me with Colbert’s (1951) The Dinosaur Book and

century (A. S. Romer, b. 1894). These historical greats still at-

Roy Chapman Andrews’s (1953) All About Dinosaurs. My early

tended Society of Vertebrate Paleontology annual meetings

experiences were very important to my formation. In conse-

when I began to attend, beginning in 1967. As a ceratophile

quence I always consider it a privilege when children (of any

(I seem to have coined this fairly obvious term, meaning

age) visit my lab or when I am invited into an elementary

‘‘lover of horns,’’ and applied it in my book to Darren Tanke

school to share some of the excitement of my field with them.

and myself [Dodson 1996: 180]), I am proudest of all to have

Who knows which of them will become one of the next gener-

known Charles M. Sternberg (1885–1981). He still came into

ation of paleontologists? I corresponded briefly with a bright

the paleo labs of the National Museum of Canada once a week

fifth-grader from rural New York State years ago, met him as a

during the summer that I worked in the prep lab under the

freshman at his college a decade later, supervised his Ph.D. at

supervision of Dale Russell, my ever-ebullient mentor and

the University of Pennsylvania, and now enjoy his collegiality

friend. One of my great regrets is that I never took C. M.’s

3

photograph. My sense of history was deficient in those days—

My first research encounter with ceratopsians arose when I

only as I matured did I gain a proper sense of awe at the hu-

went to Yale for my Ph.D. in 1970. It may be said that in those

man dimension of discovery and scholarship. It is difficult to

days dinosaur studies were in a ‘‘quiet’’ phase in the United

think of oneself as part of history. However, one can at least

States. Colbert had retired from the American Museum of Nat-

accept that each of us is part of a continuum. In an academic

ural History and had moved to Flagstaff. It was only at Yale

sense, I did not spring fresh from the brow of Job, but I am

that dinosaur studies were being pursued with vigor by young

the product of my academic mentors, Dale Russell (National

John Ostrom, whose studies on hadrosaurs (Ostrom 1961,

Museum of Canada—now the Canadian Museum of Nature),

1964a) and ceratopsians (1964b, 1966) certainly attracted my

Richard C. Fox (University of Alberta), and John Ostrom (Yale

attention. John had just completed five years of fieldwork

University)—thanks, Dads! Through John Ostrom I can trace

in the Early Cretaceous Cloverly Formation of the Bighorn

my lineage back through E. H. Colbert, to Henry Fairfield Os-

Basin, and in announcing the exciting discovery of the hot-

born and W. K. Gregory, from the former to E. D. Cope, and

blooded predator Deinonychus (Ostrom 1969) introduced an

then back to Joseph Leidy (1823–1891), who is revered as the

entire new dinosaur fauna including the basal iguanodontian,

father of American paleontology. Note that through my Yale

Tenontosaurus, the nodosaurid Sauropelta, and the problematic

and Philadelphia sojourns, there is an interesting blending of

small theropod Microvenator (Ostrom 1970a). My arrival at

both the Marsh and Cope legacies!

Yale coincided with Ostrom’s rediscovery of a specimen of

I did not set out to become a student of horned dinosaurs. It

Archaeopteryx, the first new one to be recognized since 1877

just sort of happened. It was a quiet field when I entered it, and

(Ostrom 1970b). Unknown to me until my arrival in New

remained so for many years, but for the last decade and a half

Haven was a certain long-haired, bearded, and chromatically

it has been very active. I majored in geology at the University

attired gentleman named Robert Bakker, who filled the air

of Ottawa, and although paleontology was part of the curricu-

with colorful descriptions of warm-blooded dinosaurs. Bob

lum, I enjoyed sedimentology and geomorphology as well,

was a stimulating friend and an important intellectual influ-

and considered these fields for graduate work. I participated in

ence on me. These were exciting times at Yale. Jim Farlow

fieldwork on Somerset Island in the Canadian Arctic in 1965

arrived at Yale two years later, and he looms large in my early

and 1966. Here I collected Silurian shelly invertebrates and

ceratopsian studies.

Devonian ostracoderms and placoderms with David Dineley,

My work at Yale involved biometric studies of growth series

as my introduction to field paleontology. In 1967, I spent a

of dinosaurs (Dodson 1974). Ostrom wisely counseled me to

magical summer before my senior year in the ancient paleo

begin my work with alligators, to establish my knowledge of

laboratory at the National Museum of Canada. It was in this

basic reptilian osteology, to hone my practical techniques of

limestone block building at Sussex and Georgestreets where

measurement with the tools of the trade, calipers and tape

Lawrence Lambe began to put Canadian dinosaurs on the

measure, and to get a handle on analytical techniques. Alliga-

map in the early 1900s. Here I met C. M. Sternberg, a hoary

tors are to paleontologists what Drosophila is to geneticists

link to a glorious past. He was old and stooped, but his eyes

or Caenorhabditis elegans is to developmental biologists. I hesi-

still sparkled. I also met Wann Langston, Jr., who spent a

tate to admit just how primitive my analytical techniques

month there in his study of Pachyrhinosaurus canadensis. There

were. I resisted acquiring competence with computers long

Dale Russell became my mentor and friend. That summer

past the point of reasonableness. This was, of course, long

Gilles Danis and Gerry FitzGerald began their careers in pale-

before the advent of desktop computing. The personal com-

ontology as well. This auspicious alignment of the planets, so

puter had not been developed (I acquired my first personal

to speak, confirmed my desire to become a paleontologist.

computer only in 1984, and it was unspeakably crude by to-

Initially I planned to go to Austin and do a master’s degree

day’s standards—word processing programs did not yet have

with Wann Langston at the University of Texas, but in 1968,

spell checkers!).

the Vietnam War was raging, and it seemed to make a great

The alligator study (Dodson 1975a) proved to be a very use-

deal of sense to remain in Canada for a while. Dale Russell

ful baseline, and it continues to be cited in a variety of con-

encouraged me to apply to the University of Alberta to study

texts by paleontologists whom I admire, even in recent years

with Richard C. Fox, which I did. Dale provided me with ideas

(e.g., Houck et al. 1990; Clark et al. 2000; Brochu 2001; Erick-

both for my master’s research and my Ph.D. I spent the sum-

son et al. 2003; Chinnery 2004; Farlow et al. 2005; Evans et al.

mer of 1968, 10 weeks of it, living in a tent at Dinosaur Pro-

2005). I also published a study of two closely related species of

vincial Park, Alberta, with my new bride and field assistant,

Sceloporus lizards, based on a data set kindly provided by Er-

Dawn. Here I walked over the decaying remains of ceratop-

nest Lundelius (Dodson 1975b). This study provided a careful

sians while collecting taphonomic data for my master’s thesis

analysis of a data set that was complicated by both sex and

(Dodson 1971), but did not otherwise latch onto ceratopsians

taxonomy. The sexual difference was real but proved difficult

as the objects of my desire.

to uncover. At last I was ready to take on dinosaurs in two

4 dodson

nov (1972), who had only seven skulls to work with; but my study (Dodson 1976) was the most rigorous mathematically and had the largest sample size and growth range. The specimens in my study also all came from the same Mongolian locality, the Flaming Cliffs of Bayn Zag. Claims of sexual dimorphism in dinosaurs are properly greeted with skepticism (e.g., Sampson and Ryan 1997; Padian et al. 2005). In order to sustain a strong claim for sexual dimorphism, a significant sample of specimens is required from the same locality (thus ruling out geographic or chronological complications), ideally spanning a significant size range so that ontogenetic variation can be accounted for. Rigorous morphometric analysis is also required. These exacting requirements have rarely been met in dinosaur studies. The case for sexual dimorphism in Protoceratops is arguably the best in all of dinosaur paleontology. My attempt to infer sexes in lambeosaurine hadrosaurs (Dodson 1975c) held up for 30 years, but fell apart two years ago, not because of a faulty morphoPrincipal coordinates analysis of a growth series of Protoceratops andrewsi showing juveniles (open circles), adult males (solid circles), and adult females (horizontal hatching). The identity of specimen 21 (vertical hatching) is ambiguous. Redrawn from Dodson (1976).

FIGURE 1.1.

metric analysis but because new stratigraphic evidence from the Dinosaur Park Formation carefully analyzed by David Evans and colleagues (2006) showed that my putative sexes of Corythosaurus are likely time-successive species. I was wrong (Dodson 2007)! No study in science, and least of all in paleontology, is definitive, in the sense of constituting the final word on a subject.

conceptually related studies. The easier case was that of Proto-

New methods and fresh minds bring new insights. For exam-

ceratops, the more difficult case that of Canadian lambeosau-

ple, my use of multivariate statistics brought new insights into

rine hadrosaurs.

Protoceratops three decades ago. Inevitably, other researchers

Protoceratops andrewsi is one of the great treasures of the

would eventually revisit the topic. Greg Erickson developed

American Museum of Natural History’s famed Central Asiatic

along with Tatiyana Tumanova (Erickson and Tumanova

Expedition of the 1920s, headed by the flamboyant Roy Chap-

2000) a technique named Developmental Mass Extrapolation

man Andrews. Using the magnificent display skulls, I was

(DME) that combines skeletochronology and mass estimates

ready to tackle growth in size from small specimens (basal

to produce a growth curve. DME was first applied to Psittaco-

length 76 mm) to large (basal length 357 mm). Using both

saurus mongoliensis, providing important insight into its life

bivariate plots and multivariate statistics (Fig. 1.1), I showed

history. Psittacosaurus shows growth rates ranging from 2.6 g

that there is a distinct dimorphism in the large specimens.

to 12.5 g/day to a maximum calculated body size of about

One group of specimens showed an arch over the nose (site

20 kg at age 9 years. Similar statistics will soon be available for

of the presumptive nose horn), broadly flaring jugals, and a

Protoceratops. Erickson has joined a consortium led by Peter

broad and elevated frill. The second group consisted of speci-

Makovicky and including Rud Sadleir and Mark Norell, as well

mens with less arching over the nose and with frills that were

as myself (Makovicky et al. 2007). Preliminary reanalysis of my

slightly lower and narrower (Fig. 1.2). The smaller specimens

data failed to reveal the dimorphism that I claimed. However,

plotted with the latter group of large specimens. The dimor-

by way of documentation, I still possess the computer print-

phism demands explanation, and no one has ever argued for

outs that I produced at the Yale Computer Center in 1973

two species for the specimens from Shabarak Usu (Lambert

(now I know why I never throw anything out!). It was in fact

et al. 2001 recently named P. hellenikorhinus from Inner Mon-

quite difficult for me to extract the dimorphism I reported in

golia). Accordingly, my interpretation is that the dimorphism

1976. Even an analysis of series of alligators, crocodiles, and

is sexual in nature, and that males had the taller, wider, show-

gharials failed to separate three genera of extant reptiles in a

ier skulls. The crest of Protoceratops is eggshell thin, and fits

principal coordinates analysis until I used ratios rather than

well, in my judgment, the paradigm of a display structure

raw measurements, which have the effect of swamping out

rather than a mechanical framework for jaw muscles. Sexual

everything else in size studies (Dodson 1978). I feel confident

dimorphism had been suggested before for Protoceratops, ten-

that the dimorphism I reported in 1976 will be sustained when

tatively by Brown and Schlaikjer (1940), analytically by Kurza-

all is considered, but I am hardly an objective commentator.

Forty Years of Ceratophilia 5

frilled chasmosaurine, and was often placed in the wrong lineage in the past (e.g., Lull 1933; Colbert 1948; Ostrom 1966). In Farlow and Dodson (1975: fig. 1), mirabile dictu, we followed our famous mentors rather than the less distinguished but certainly correct C. M. Sternberg (1949), who clearly saw Triceratops as a chasmosaurine, not a centrosaurine. In 1975, I was certainly no expert on horned dinosaurs, only a rank newcomer. Our interest then was primarily paleobiological not phylogenetic, although we certainly understand more clearly today that comparative biology must be predicated on accurate phylogeny. My dissertation work completed, Jim Farlow and I went our separate ways, and we did collaborate further for several decades. It was not for another decade that ceratopsians troubled me greatly. To my considerable surprise (and slight measure of dismay at the time), I assumed my post as a gross anatomist at the University of Pennsylvania School of Veterinary Medicine. I won this position by virtue of auditing the human gross anatomy course at Yale Medical School. I always counsel my students to expect the unexpected. Over the years, I have come to enjoy greatly the teaching of veterinary gross anatomy, so rich in paleobiological implications. The University of Pennsylvania has attained elite status among American universities, facilitating the process of attracting high-quality students from around the country and around the world. My Protoceratops andrewsi male and female specimens redrawn from Dodson (1976).

FIGURE 1.2.

professorship has been fecund, and I have the pleasure now of seeing the students of my students rising within the academy (Fig. 1.3). David B. Weishampel (Ph.D. 1981) was my first student (and who could ask for a better or more productive one?), and Larry Witmer (Ph.D. 1991) was his academic first-born.

In my dissertation, I stuck close to the data and resisted wild

Although neither David nor Larry is primarily a student of

speculation, which John Ostrom most definitely did not en-

ceratopsians, Larry at least illustrated the skull of Triceratops,

courage. However, Jim Farlow was impressed by the apparent

and analyzed its narial position (Witmer 2001). David’s stu-

sexual dimorphism, coupled with the eggshell thinness of

dent and my doctoral grandchild, Brenda Chinnery-Allgeier

the Protoceratops frill, and encouraged me to examine display

(Ph.D. 2002), is decidedly a ceratopsian specialist (Chinnery

structures in modern ungulates, especially as analyzed by Ca-

and Weishampel 1998; Chinnery 2004; Chinnery and Horner

nadian wildlife biologist Valerius Geist (Geist 1966). In the

2007). Frank Varriale in the Weishampel laboratory is com-

rashness of our youth, Jim and I (Farlow and Dodson 1975)

pleting an ambitious survey of mastication in marginocepha-

extrapolated from Protoceratops to the use of the frill for sexual

lians, and his Ph.D. is expected soon. Another fecund line of

display in ceratopsians generally. We posited a social model

descent is delineated by Catherine Forster (Ph.D. 1990), an

for chasmosaurines, and a more solitary model for centro-

expert on Triceratops, the subject of her dissertation (Forster

saurines, based on the differing potential for display and safe

1996a, b), and generally knowledgeable about ceratopsians

coupling of the horns. According to this model, one might

(Forster et al. 1993; Dodson et al. 2004), although she does

predict bonebeds for chasmosaurines and none for centro-

many other things as well. Her splendid student is Andy Farke

saurines. Although the chasmosaurine Agujaceratops occurs in

(Ph.D. 2008), a dedicated ceratophile since high school (Farke

bonebeds (e.g., Lehman 1989), so also do Centrosaurus (Cur-

2004, 2006; Farke and Williamson 2006). Tony Fiorillo (Ph.D.

rie and Dodson 1984) and Pachyrhinosaurus (Langston 1975;

1989) earned his Ph.D. on the taphonomy of the Careless

Tanke 1988), thus effectively falsifying that prediction. Tri-

Creek bonebed (Fiorillo 1991), but has since become expert on

ceratops seems to present an interesting case, in that bonebeds

Alaskan dinosaurs (Fiorillo and Gangloff 2000; Fiorillo 2004).

are extremely rare (Dodson 1996), which is consonant with its

Tony is currently excavating a Pachyrhinosaurus bonebed on

short frill suggesting a solitary nature, but not consonant with

the North Slope. Paul G. Penkalski, Jr., earned his M.Sc. in

its membership in the Chasmosaurinae. It is truly a short-

1994, with a fine study of Avaceratops lammersi (Penkalski and

6 dodson

FIGURE 1.3.

A genealogy of Dodson ceratopsian students.

Dodson 1999). Hai-Lu You earned his Ph.D. in 2002. He has

1975a, b, c, 1976, 1978), but this does not necessarily result in

studied Chinese Cretaceous faunas generally, including many

any deep knowledge of any particular animal or group of ani-

basal ceratopsians: Hongshanosaurus (You et al. 2003; You and

mals. It was the next episode that provided the entrée to a new

Xu 2005); Magnirostris (You and Dong 2003): and Auroracera-

phase of my career. A serendipitous phone call to my home led

tops (You et al. 2005). Allison Tumarkin finished her Ph.D.

to a visit in October 1981 with Eddie and Ava Cole in their

in 2003. She is expert on the bone texture of archosaurs

fossil shop in Wall, South Dakota, where I recognized the par-

(Tumarkin-Deratzian et al. 2006, 2007), but with a distinct in-

tial remains of a small ceratopsid that they had just collected

terest in ceratopsians (Tumarkin and Dodson 1998; Tumarkin-

from the Judith River Formation on a ranch in south central

Deratzian and Dodson 2005). My current ceratopsian student

Montana. I was excited on several accounts—small dinosaurs

is Kyo Tanoue (Ph.D. expected 2008), who is also being men-

in general and small ceratopsids in particular are very rare,

tored by Hai-Lu You, in whose lab in Beijing he has collected

especially in North America, and south central Montana is not

much data. This mentorship is reflected in a series of papers,

the classic region for the Judith River Formation. The Judith

three of which appear in this volume, and two of which have

River Formation is widespread in north central Montana but

already appeared (You et al. 2007; You et al. in press). Thus my

historically has been rather stingy in high-quality specimens

students and ‘‘grandstudents’’ have contributed very signifi-

compared to formations such as the Dinosaur Park Formation

cantly to the literature of Ceratopsia over the past two de-

of Alberta and the Two Medicine Formation of northwestern

cades. In fairness I must acknowledge that my ceratopsian

Montana. To give a very telegraphic account, I visited the field

students are but a subset of all of my students. It has been the

site on the Careless Creek Ranch near Harlowton with the

quality and consistency of my students over the years that

Coles in 1982, finished collecting the specimen, and even-

have maintained my interest in the academic life.

tually took the entire collection to the Academy of Natural

My first post-Yale project was in the Upper Jurassic Mor-

Sciences in Philadelphia, the home of the first American dino-

rison Formation (notably short on ceratopsians, but rich in

saurs, described by Leidy and by Cope in the nineteenth cen-

many other dinosaurs) with Bob Bakker, Kay Behrensmeyer,

tury. I described Avaceratops lammersi (Fig. 1.4) in 1986 as

and Jack McIntosh (Dodson et al. 1980), and then I began

a basal centrosaurine (honoring Ava Cole and the Lammers

several years of faunal studies at Dinosaur Park, Alberta, at the

family, on whose land we worked). This was my first taxo-

invitation of Phil Currie. I at least was able to determine the

nomic offering to the field of paleontology. The splendid little

contribution of ceratopsians to the very rich Judithian dino-

animal measuring 2.3 m long and standing only 800 mm at

saur fauna, 24.7% of adjusted skeletal census, and 26.4% of

the hip is about 70% complete. A cast of Avaceratops stands on

the microfaunas (Dodson 1983, 1987). Ceratopsians are about

permanent exhibit at the Academy of Natural Sciences. Per-

half as abundant as hadrosaurs in Alberta.

versely, my first horned dinosaur lacked horns—or rather,

At this time in my life I suppose I was more a dinosaur gen-

they were not preserved, so we can only speculate on their

eralist than a specialist. I really did not have in-depth knowl-

appearance. Subsequently Tony Fiorillo earned his Ph.D. in

edge of the anatomy and systematics of any group of dino-

1989 on the taphonomy of the Careless Creek quarry and

saurs. I had employed the techniques of taphonomy (Dodson

analysis of the fauna of the region (Fiorillo 1989, 1991), and

1971; Dodson et al. 1980) and of morphometrics (Dodson

Paul Penkalski earned his master’s on Avaceratops in 1994

Forty Years of Ceratophilia 7

FIGURE 1.4.

Avaceratops lammersi (Dodson 1986), type specimen ANSP 15800, Academy of Natural Sciences, Philadelphia.

(Penkalski and Dodson 1999). Granting that Avaceratops is a

topsians. Students are the best things that have ever happened

juvenile, it certainly has autapomorphies that ensure its valid-

in my professional life. David Weishampel is the sort of stu-

ity despite some initial skepticism. Nonetheless, to my cha-

dent one can get used to. David graduated with his Ph.D.

grin there has been a considerable reluctance to include Ava-

in 1981, and began his professional career at Florida Inter-

ceratops in current phylogenetic analyses (e.g., Sampson et al.

national University in 1983, following a NATO postdoctoral

1997; Dodson et al. 2004—phylogenetic analysis by ‘‘et al.’’

fellowship in Tübingen. As he cast about in 1984 to begin

[i.e., Forster and Sampson, not Dodson]; Ryan 2007). It consis-

making his indelible mark on science, he hit on a very am-

tently occupies a basal position either within or just outside

bitious idea, and enlisted my support. That idea we know to-

of Centrosaurinae. Granted that juvenile specimens may be

day as The Dinosauria (Weishampel et al. 1990, 2004), which

more difficult to work with than adults, there is no a priori

is generally acclaimed as an enormously valuable reference

reason to justify the exclusion of juveniles or subadults merely

source—certainly one I use every day myself. My major con-

because they are difficult. Australopithecus africanus, Raymond

tribution to that book was the chapter ‘‘Neoceratopsia,’’ co-

Dart’s revolutionary ‘‘ape of the south’’ is based on the 3-year-

authored with Phil Currie (Dodson and Currie 1990). I wrote

old Taung child, fortunately not discarded because of its ju-

my material during a sabbatical year (1985–1986) with Dale

venile status. The good news is that juvenile alligators grow

Russell in Ottawa at the National Museum of Natural Science

up to be adult alligators not adult crocodiles, and vice versa

(now Canadian Museum of Nature). I did not have to rely on

(Tumarkin and Dodson 1998). It is also well to remind our-

published accounts of ceratopsian taxa. What a pleasure it

selves that a partial skull and skeleton, in this case one 70%

was to rise from my computer screen and stroll back into the

complete, is incomparably better than mere fragments or no

cavernous collection room, where the specimen cabinets and

specimen at all. It is possibly not a strength of cladistics that

open shelves housed the fabled specimens that I was writing

juveniles invariably end up as basal forms. The ‘‘problem’’

about: Centrosaurus apertus, Chasmosaurus belli, Styracosaurus

with Avaceratops is that it lacks the nasal, postorbital, and pari-

albertensis, Pachyrhinosaurus canadensis, Monoclonius lowei, all

etal of adult form (Michael Ryan pers. com. 2008). Of course,

the while ignoring equally splendid skulls of the hadrosaurs

difficulty in assessing phylogenetic position is not the same as

Corythosaurus casuarius, Lambeosaurus lambei, and Edmonto-

questioning taxonomic validity.

saurus regalis. Editorially, we decided to include the Proto-

In any case, after Avaceratops came into my life, I now felt

ceratopsidae in the same chapter (thus ‘‘Neoceratopsia’’), and

fully committed, both scientifically and emotionally, to cera-

assigned the newly minted Ph.D. Paul Sereno to describe Psit-

8 dodson

FIGURE 1.5.

Clustered results from an RFTRA analysis of ceratopsian skulls. This purely phenetic analysis recovers clusters that are taxonomically congruent with the monophyletic taxa Centrosaurinae and Chasmosaurinae plus a basal group. Redrawn from Dodson (1993).

tacosauria in a separate chapter. As knowledge evolves and

held at the Tyrrell Museum of Palaeontology (now the Royal

new workers enter the field, we realigned material for the sec-

Tyrrell Museum—where, oh where, has the ‘‘palaeontology’’

ond edition. We thus wrote a chapter on basal ceratopsians

gone?) in Drumheller June 3–5, 1986 (Dodson 1990). Al-

including Psittacosauria (You and Dodson 2004) and a sepa-

though I was not hooted off the stage, there has been no de-

rate chapter on Ceratopsidae (Dodson et al. 2004). Part of

tectable enthusiasm for this explanation of morphological

the editorial charge to the authors of The Dinosauria was

variability within the Centrosaurinae, and Sampson et al.

to be critical and authoritative at the species level. Accord-

(1997) specifically argue against the attempt to link Styraco-

ingly I attempted to rationalize the species of Centrosaurus, of

saurus albertensis and C. nasicornus. I would not try very hard

which there were a plethora, including C. flexus, C. longirostris,

to defend my suggestions today, but certainly no one today

C. nasicornus, and C. dawsoni. I naturally attempted to use

tries to argue for the validity of such species as C. flexus, C. lon-

the methodology that was at least somewhat successful in ra-

girostris, C. nasicornus, and C. dawsoni.

tionalizing variability in Protoceratops and Corythosaurus. I

My generation of paleontologists has not been eager to em-

measured every skull I could, including specimens of Mono-

brace cladistics, although my students from Dave Weishampel

clonius and Styracosaurus, and chasing skulls in New York, New

onward have not hesitated to do so, with my blessing. In a de-

Haven, Toronto, and Alberta, as well as the Ottawa skulls. The

liberately polemical piece (Dodson 2000), I articulated some

study was much less satisfactory because there was a dearth of

of my discontents with the method, but I recognize its value.

small ceratopsid skulls. The size range was 73 cm to 83.5 cm

However, I have in general chosen to exercise my limited tal-

basal length, or 138 cm to 158 cm total skull length, meaning

ents in other directions. In 1993, my Yale classmate (and fel-

that all specimens were either adult or close to adulthood, and

low cladistics-resister!) Phil Gingerich and I edited a special

thus bivariate plots had little explanatory power. I attempted

volume of the American Journal of Science in honor of John

to discern clusters, and concluded that there were two groups

Ostrom’s 25th anniversary of editorship (Dodson and Ginge-

within Centrosaurus apertus, the genotypic species: one includ-

rich 1993). In this volume, I presented my RFTRA (Resistant-

ing the robust C. ‘‘flexus’’ and the other including the remain-

Fit Theta-Rho Analysis) study of ceratopsian skulls (Dodson

ing species as more gracile, which I tentatively posited as, re-

1993), which I believe to be my most important rarely cited

spectively, male and female, in keeping with my inclinations.

paper. This strictly landmark-based morphometric analysis

I found that C. nasicornus shared at least one character with

provides a cluster analysis with a strong phylogenetic signal

Styracosaurus albertensis, a very tall straight nasal horncore,

without the baggage of cladistic assumptions about the polar-

and I posited C. ‘‘nasicornus’’ as a female Styracosaurus alber-

ity of character states. Basal Ceratopsia, Centrosaurinae, and

tensis. Both are very rare types. I also attempted to recognize

Chasmosaurinae are recovered as separate clusters (Fig. 1.5).

sexes in Monoclonius, although I had but a single complete

Moreover, with RFTRA one can easily strip away selected char-

skull, that of the enigmatic M. ‘‘lowei’’ to work with. I pre-

acters and determine how robust the phylogenetic signal is;

sented these findings at the Dinosaur Systematics Conference

thus we can answer the question, do we put too much reliance

Forty Years of Ceratophilia 9

on horns and frills in the diagnosis of ceratopsian taxa? The

white art by my friend Philadelphia artist Robert Walters, who

answer is a resounding ‘‘No!’’ For example, when the nasal and

was the first one to present drawings of Avaceratops in 1986.

orbital horns and the parietal are removed, the three clus-

In preparing The Horned Dinosaurs I dabbled in the long-

ters remain, and in fact the two specimens of Chasmosaurus

vexed question of ceratopsian posture. The old approach to

are brought together, whereas in the first analysis long-horned

skeletal mounts of ceratopsids, namely Triceratops at the Amer-

Chasmosaurus ‘‘kaiseni’’ plots with Pentaceratops. Especially

ican Museum of Natural History (Osborn 1933) and Chasmo-

interestingly, Triceratops now plots with the Centrosaurinae,

saurus at the National Museum of Canada (Sternberg 1927),

where it has long been placed incorrectly, because it is cer-

involved a wide sprawl of the forelimbs with horizontal hu-

tainly a short-frilled chasmosaurine. In a third analysis I

meri. As the modern dinosaur was refashioned in brilliant

focused on the masticatory system, removing the face in addi-

drawings and paintings by Robert Bakker, Gregory Paul, and a

tion to the horns, but reinstating the parietal due to its poten-

host of dinosaur art specialists, dinosaurs in general and cera-

tial role in the masticatory system, as postulated by Ostrom

topsids in particular were recast with absolutely vertical, para-

(1966). Once again the three clusters are recovered, and Tri-

sagittal limbs. Bakker (1986, 1987) insisted that ceratopsian

ceratops is correctly classified as a chasmosaurine. Thus cera-

limbs compared favorably with those of elephants and rhinoc-

topsian classification is not unduly based only on horns and

eros, and Paul (1987) also provided explicit and detailed in-

frills, and is quite robust even with incomplete specimens, a

structions to artists for restoring skeletons with vertical limbs.

most encouraging finding. The morphometric analysis also

I had reservations about the comparison of ceratopsids with

emphasized trends in the evolution from basal ceratopsians to

large mammals, particularly in the shoulder region. The prom-

centrosaurines to chasmosaurines involving not merely the

inence of the ceratopsid deltopectoral crest and the eccen-

obvious developments of facial horns and parietal elongation

tric head of the humerus both seemed impediments to upright

and ornamentation but also the reorientation of the cheek

posture in the style of ungulate mammals, as these features

region, postquadrate expansion of squamosal, and finally ele-

also seemed to those who actually mounted ceratopsid skele-

vation of the squamosal by bending along its long axis.

tons, something the aforementioned artists did not do. A good

A much-appreciated sabbatical in 1994 allowed me the

footprint record might potentially provide important insights

opportunity to pursue a writer’s dream, writing a book on

into the problem. Oddly, despite a wealth of tracks of other

one’s favorite subject. It was actually proposed to me by artist

kinds of dinosaurs, ceratopsid trackways are exceedingly rare.

Wayne Barlowe, with whom I had worked on a children’s al-

Only a single trackway has been described (Lockley and Hunt

phabet book for the late Byron Preiss (Dodson and Barlowe

1995), and it consists of only three prints. I teamed up with

1995). Wayne’s wife, Shawna McCarthy, is a literary agent,

my old friend Jim Farlow to analyze ceratopsid posture (Dod-

and soon we had a contract, and a wonderful sponsoring edi-

son and Farlow 1997), and we concluded that the footprints,

tor, Jack Repcheck. Unfortunately, Jack switched publishers

claimed as definitive proof of strict parasagittal posture, are

before the book was finished, and the publisher rejected the

nothing of the sort. Although the latter posture may be aes-

book they had contracted. Fortunately, Jack’s new employer,

thetically superior, it is not necessarily supported by the

Princeton University Press, was pleased to publish the book.

anatomy. We opted for an intermediate semi-erect posture.

Accordingly, The Horned Dinosaurs—A Natural History, was

Happily, this study has been corroborated by the detailed bio-

published in 1996, a labor of love if ever there were one (Dod-

mechanical analysis of Thompson and Holmes (2007); see also

son 1996). I was able to delve deeply into the literature of the

Rega at al. this volume.

Ceratopsia, to trace the faltering early steps toward under-

I more or less saw The Horned Dinosaurs as a swan song, or a

standing the nature of the beast. My Philadelphia hero E. D.

summary statement of my work in horned dinosaurs. I was

Cope described the first three ceratopsians, the best of which

not then doing fieldwork, and with two major book projects

was Monoclonius crassus Cope 1876, which wasn’t really very

in the way, I had not done any since 1988. Likewise, I did not

good. Simply put, Cope didn’t have a clue what ceratopsians

have any ceratopsian projects. Within a couple of years in the

were. He thought the parietal shield of Monoclonius was a

mid-1990s, I became mentor to Allison Tumarkin, Josh Smith,

breast bone! It was only when Marsh described Triceratops in

Hai-Lu You, Matt Lamanna, and Jerry Harris, and soon my

1889 that the Ceratopsia began to take form. In any case my

plate had become very full. I invited Allison and Matt to the

book was about as comprehensive as it was possible to be as of

field in Montana, Josh and Matt invited Allison and me to

1994—of course every printed word always becomes dated the

Egypt, Matt invited me to Argentina, and Hai-Lu invited Josh,

moment the publisher receives it. The book was graced with

Matt, Jerry, and me to China. Montana and Egypt involve

six exquisite original color plates by Wayne featuring Psittaco-

sauropods (Smith et al. 2001; Harris and Dodson 2004). My

saurus, Leptoceratops, Styracosaurus, Chasmosaurus, Pachyrhino-

career took a Sinocentric shift in 1995 when I attended a con-

saurus, and Triceratops. It was also enhanced by black-and-

ference in China, where I met Hai-Lu You (Fig. 1.6), a young

10 dodson

FIGURE 1.6.

Hai-Lu You calculates phylogenetic relationships of basal ceratopsians in his office at the Chinese Academy of Geological Sciences, Beijing. Photo by Kyo Tanoue.

master’s student at the Institute of Vertebrate Paleontology

ince, Matt Lamanna, Hai-Lu, and I visited provincial geologist

and Paleoanthropology (IVPP) in Beijing. So impressed was I

Li Daqing (Fig. 1.7) in Lanzhou, and he presented us with a

that I invited him to be my student at the University of Penn-

beautiful skull of a basal ceratopsian from Mazongshan. In

sylvania. Hai-Lu completed his Ph.D. in 2002 and returned to

2005, our team named the animal Auroraceratops (You et al.

China, taking a position at the Institute of Geology of the

2005; Fig. 1.8) in honor of my wife of 37 years, Dawn (aurora in

Chinese Academy of Geological Sciences, under the direction

Latin). In studying Auroraceratops, once again I was tugged

of Ji Qiang, one of the great movers of contemporary Chinese

down into the rich world of basal ceratopsians, a world in

paleontology. One of the first things Hai-Lu did on returning

which I have reveled for 30 years. I was accompanied in China

to China was to name the basal ceratopsian Magnirostris dod-

by my new student Kyo Tanoue, who is looking at jaw func-

soni (You and Dong 2003), meaning ‘‘Dodson’s big nose!’’ He

tion in basal ceratopsians, making use of the splendid new

and I redescribed Archaeoceratops from Gansu (You and Dod-

specimens (Fig. 1.9). The first paper resulting from this col-

son 2003). It can hardly have escaped notice except by sin-

laboration is a description of a tiny skull of Liaoceratops (You

cere troglodytes that dinosaur paleontology has exploded in

et al. 2007).

China during the past decade. Although theropod and bird

In 2006, I took another turn in course. When I returned to

finds have grabbed the headlines, especially but not exclu-

China that summer, waiting for me in Hai-Lu’s office was a

sively from northeastern China, discoveries and descriptions

splendid large skull of Psittacosaurus from Liaoning. The qual-

of new basal ceratopsians have just about kept pace: Archaeo-

ity of preservation and of the preparation was exquisite—no

ceratops (Dong and Azuma 1997); Chaoyangsaurus (Zhao et al.

sediment remained to obscure details either inside or out—

1999); Liaoceratops (Xu et al. 2002); Hongshanosaurus (You et

this in contrast to the chunks of rock that so often pass for

al. 2003); Magnirostris (You and Dong 2003); Auroraceratops

psittacosaur skulls. It was hard for me to imagine that there

(You et al. 2005); Yinlong (Xu et al. 2006); Xuanhuaceratops

was anything left to learn about Psittacosaurus, one of the very

(Zhao et al. 2006), as well as Yamaceratops (Makovicky and

well-known ceratopsian dinosaurs of Mongolia and one of the

Norell 2006) from Mongolia. In our 2004 foray to Gansu Prov-

most common dinosaurs in the world. However, it became

Forty Years of Ceratophilia 11

As I look back over the field of ceratopsian studies, I am impressed indeed. It once seemed a tired and dated field of study with few dedicated students. All of the ceratopsids seemed to have been discovered, with no new genera described between 1950 (Pachyrhinosaurus) and 1986 (Avaceratops). No basal ceratopsian genera were described from China until Archaeoceratops was described in 1997. Who could foresee the explosion of discoveries and renewed interest of the past decade and a half? There is so much material from China, Mongolia, and North America. Hopefully, taxonomically significant material will soon be forthcoming from Mexico. I see nothing but opportunity for years to come, spiced by new finds at every taxonomic level. The harvest is great but the workers are few! References Cited

Peter Dodson with Li Daqing, discoverer of Archaeoceratops, Auroraceratops, and many other specimens in the Gobi Desert, Mazongshan Region, Gansu Province, northwestern China. Photo by Hai-Lu You.

FIGURE 1.7.

pretty clear that anatomical details were available now that were not available to Paul Sereno in his superb doctoral studies now more than two decades old (Sereno 1987; summarized in Sereno 1990a, b). I had tended on a priori philosophical grounds to be dismissive of the idea of multiple species of Psittacosaurus. However, Sereno’s housecleaning of old species included the naming of two new species, one from northwestern China: P. xinjiangensis (Sereno and Chao 1988), and one from northeastern China: P. meileyingensis (Sereno et al. 1988). Several new species were named during the 1990s on less than ideal material: P. neimongoliensis from Inner Mongolia (Russell and Zhao 1996) and P. mazongshanensis from Gansu (Xu 1997). Two more species were added in the last several years: P. lujiatunensis (Zhou et al. 2006) and P. major (Sereno et al. 2007). The geographic range of Psittacosaurus has been extended to Siberia by P. sibiricus (Averianov et al. 2006). This level of speciation in a dinosaur is unprecedented. What is now required is a major restudy of Psittacosaurus within a holistic account that takes into consideration the temporal, geographic, and environmental context.

12 dodson

Andrews, R. C. 1953. All About Dinosaurs. New York: Random House. Averianov, A. O., A. V. Voronkevich, S. V. Leshchinskiy, and A. V. Fayngertz. 2006. A ceratopsian dinosaur Psittacosaurus sibiricus from the Early Cretaceous of West Siberia, Russia and its phylogenetic relationships. Journal of Systematic Palaeontology 4: 359–395. Bakker, R. T. 1986. Dinosaur Heresies. New York: William Morrow. ———. 1987. The return of the dancing dinosaurs. In S. J.Czerkas and E. C. Olsen, eds., Dinosaurs Past and Present, Vol. 1, pp. 38– 69. Los Angeles: Natural History Museum. Brochu, C. A. 2001. Crocodylian snouts in space and time: Phylogenetic approaches toward adaptive radiation. American Zoologist 41: 564–585. ———. 2004. Morphometric analysis of evolutionary trends in the ceratopsian postcranial skeleton. Journal of Vertebrate Paleontology 24: 591–609. Brochu, C. A., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Brown, B., and E. M. Schlaikjer. 1940. The structure and relationships of Protoceratops. Annals of the New York Academy of Science 40: 133–266. Chinnery, B. J. 2004. Description of Prenoceratops pieganensis gen. et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Chinnery, B. J., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Chinnery, B. J., and D. B. Weishampel. 1998. Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) and relationships among basal neoceratopsians. Journal of Vertebrate Paleontology 18: 569–585. Clark, J. D., H.-D. Sues, and D. S. Berman. 2000. A new specimen of Hesperosuchus agilis from the Upper Triassic of New Mexico and the interrelationships of basal crocodylomorph archosaurs. Journal of Vertebrate Paleontology 20: 683–704. Colbert, E. 1948. Evolution of the horned dinosaurs. Evolution 2: 145–163.

Auroraceratops rugosus from the Early Cretaceous of Mazongshan, Gansu Province, China. From You et al. (2005). This basal ceratopsian is named after Dawn Dodson.

FIGURE 1.8.

Forty Years of Ceratophilia 13

Kyo Tanoue studies the skull of Archaeoceratops at the Chinese Academy of Geological Sciences, Beijing. Photo by Peter Dodson.

FIGURE 1.9.

———. 1951. The Dinosaur Book. New York: McGraw-Hill. Cope, E. 1876. Descriptions of some vertebrate remains from the Fort Union Beds of Montana. Proceedings of the Academy of National Sciences of Philadelphia 28: 248–261. Currie, P. J., and P. Dodson. 1984. Mass death of a herd of ceratopsian dinosaurs. In W. E. Reif and F. Westphal, eds., Third Symposium of Mesozoic Terrestrial Ecosystems, pp. 52–60. Tubingen: Attempto Verlag. Dodson, P. 1971. Sedimentology and taphonomy of the Oldman Formation (Campanian), Dinosaur Provincial Park, Alberta (Canada). Palaeogeography, Palaeoclimatology, Palaeoecology 10: 21–74. ———. 1974. A study of relative growth in some living and fossil reptiles. Ph.D. diss., Yale University, New Haven. ———. 1975a. Functional and ecological aspects of relative growth in Alligator. Journal of Zoology London 175: 315–355. ———. 1975b. Relative growth in two sympatric species of Sceloporus. American Midland Naturalist 94: 121–150. ———. 1975c. Taxonomic implications of relative growth in lambeosaurine hadrosaurs. Systematic Zoology 24: 37–54. ———. 1976. Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. Journal of Paleontology 50: 929– 940. ———. 1978. On the use of ratios in growth studies. Systematic Zoology 27: 62–67. ———. 1983. A faunal review of the Judith River (Oldman) Formation, Dinosaur Provincial Park, Alberta. The Mosasaur 1: 89–118.

14 dodson

———. 1986. Avaceratops lammersi: A new ceratopsid from the Judith River Formation of Montana. Proceedings of the Academy of Natural Sciences of Philadelphia 138: 305–317. ———. 1987. Microfaunal studies of dinosaur paleoecology, Judith River Formation of southern Alberta. In P. J. Currie and E. H. Koster, eds., Fourth Symposium on Mesozoic Terrestrial Ecosystems, pp. 70–75. Tyrrell Museum of Palaeontology, Occasional Paper 3. ———. 1990. On the status of the ceratopsids Monoclonius and Centrosaurus. In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 231–243. Cambridge: Cambridge University Press. ———. 1993. Comparative craniology of the Ceratopsia. American Journal of Science 293A: 200–234. ———. 1996. The Horned Dinosaurs. Princeton, N.J.: Princeton University Press. ———. 2000. Origin of birds: The final solution? American Zoologist 40: 504–512. ———. 2007. On being wrong—or the diverting history of Corythosaurus. American Paleontologist 15: 27–35. Dodson, P., and W. Barlowe. 1995. An Alphabet of Dinosaurs. New York: Scholastic. Dodson, P., A. K. Behrensmeyer, R. T. Bakker, and J. S. McIntosh. 1980. Taphonomy and paleoecology of the Upper Jurassic Morrison Formation. Paleobiology 6: 208–232. Dodson, P., and P. J. Currie. 1990. Neoceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 593–618. Berkeley: University of California Press.

Dodson, P., and J. O. Farlow. 1997. The forelimb carriage of ceratopsid dinosaurs. In D. L. Wolberg, E. Stump, and G. D. Rosenberg, eds., Dinofest International: Proceedings of a Symposium sponsored by Arizona State University, pp. 393–398. Philadelphia: Academy of Natural Sciences. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Dodson, P., and P. D. Gingerich. 1993. Functional Morphology and Evolution. Special Volume of the American Journal of Science 293A: 1–478. Dong, Z., and Y. Azuma. 1997. On a primitive neoceratopsian from the Early Cretaceous of China. In Z. Dong, ed., SinoJapanese Silk Road Dinosaur Expedition, pp. 68–89. Beijing: China Ocean Press. Erickson, G. M., A. K. Lappin, and K. A. Vliet. 2003. The ontogeny of bite-force performance in American alligator (Alligator mississippiensis). Journal of Zoology London 260: 317–327. Erickson, G. M., and T. A. Tumanova. 2000. Growth curve of Psittacosaurus mongoliensis Osborn (Ceratopsia: Psittacosauridae) inferred from long bone histology. Zoological Journal of the Linnean Society 130: 551–566. Evans, D. C., P. J. Currie, D. A. Eberth, and M. J. Ryan. 2006. High resolution lambeosaurine dinosaur biostratigraphy, Dinosaur Park Formation, Alberta: Sexual dimorphism reconsidered. Journal of Vertebrate Paleontology 26(3, Suppl.): 59A. Evans, D. C., C. A. Forster, and R. R. Reisz. 2005. The type specimen of Tetragonosaurus erectofrons (Ornithischia: Hadrosauridae) and the identification of juvenile lambeosaurines. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 349–365. Bloomington: Indiana University Press. Farke, A. A. 2004. Horn use in Triceratops (Dinosauria: Ceratopsidae): Testing behavioral hypotheses using scale models. Palaeontologia Electronica 7: 1–10. ———. 2006. Morphology and ontogeny of the cornual sinuses in chasmosaurine dinosaurs (Ornithischia: Ceratopsidae). Journal of Paleontology 80: 780–785. Farke, A. A., and T. E. Williamson. 2006. A ceratopsid dinosaur parietal from New Mexico and its implications for ceratopsid dinosaur biogeography and systematics. Journal of Vertebrate Paleontology 26: 1018–1020. Farlow, J. O., and P. Dodson. 1975. The behavioral significance of frill and horn morphology in ceratopsian dinosaurs. Evolution 29: 353–361. Farlow, J. O., G. R. Hurlburt, R. M. Elsey, A. R. C. Britton, and W. Langston, Jr. 2005. Femoral dimensions and body size of Alligator mississippiensis: Estimating the size of extinct mesoeucrocodylians. Journal of Vertebrate Paleontology 25: 354– 369. Fiorillo, A. R. 1989. The vertebrate fauna from the Judith River Formation (Late Cretaceous) of Wheatland and Golden Valley Counties, Montana. The Mosasaur 4: 127–142. ———. 1991. Taphonomy and depositional setting of Careless Creek Quarry ( Judith River Formation), Wheatland County,

Montana, U.S.A. Palaeogeography, Palaeoclimatology, Palaeoecology 81: 281–311. ———. 2004. The dinosaurs of Arctic Alaska. Scientific American (December 2004): 84–91. Fiorillo, A. R., and R. A. Gangloff. 2000. Theropod teeth from the Prince Creek Formation (Cretaceous) of northern Alaska, with speculations on arctic dinosaur paleoecology. Journal of Vertebrate Paleontology 20: 675–682. Forster, C. A. 1996a. New information on the skull of Triceratops. Journal of Vertebrate Paleontology 16: 246–258. ———. 1996b. Species resolution in Triceratops: Cladistic and morphometric approaches. Journal of Vertebrate Paleontology 16: 259–270. Forster, C. A., P. C. Sereno, T. W. Evans, and T. Rowe. 1993. A complete skull of Chasmosaurus mariscalensis (Dinosauria: Ceratopsidae) from the Aguja Formation (late Campanian) of West Texas. Journal of Vertebrate Paleontology 13: 161–170. Geist, V. 1966. The evolution of horn-like organs. Behaviour 27: 175–214. Harris, J. D., and P. Dodson. 2004. A new diplodocoid sauropod dinosaur from the Upper Jurassic Morrison Formation of Montana, U.S.A. Acta Palaeontologica Polonica 49: 197–210. Houck, M. A., J. A. Gauthier, and R. E. Strauss. 1990. Allometric scaling in the earliest Archaeopteryx lithographica. Science 247: 195–198. Kurzanov, S. M. 1972. Sexual dimorphism in protoceratopsians. Paleontological Journal 1972: 91–97. Lambert, O., P. Godefroit, H. Li, C. Shang, and Z. Dong. 2001. A new species of Protoceratops (Dinosauria, Neoceratopsia) from the Late Cretaceous of Inner Mongolia (P. R. China). Bulletin de L’Institut Royal des Sciences Naturelles de Belgique, Sciences de la Terre 71: 5–28. Langston, W., Jr. 1975. The ceratopsian dinosaurs and associated lower vertebrates from the St. Mary River Formation (Maestrichtian) at Scabby Butte, southern Alberta. Canadian Journal of Earth Sciences 12: 1576–1608. Lehman, T. M. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. Lockley, M. G., and A. P. Hunt. 1995. Ceratopsid tracks and associated ichnofauna from the Laramie Formation (Upper Cretaceous: Maastrichtian) of Colorado. Journal of Vertebrate Paleontology 15: 592–614. Lull, R. S. 1933. A revision of the Ceratopsia or horned dinosaurs. Memoirs Peabody Museum of Natural History 3: 1–175. Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Makovicky, P. J., R. Sadleir, P. Dodson, G. M. Erickson, and M. A. Norell. 2007. Life history of Protoceratops andrewsi from Bayn Zag, Mongolia. Journal of Vertebrate Paleontology 27(3, Suppl.): 109A. Marsh, O. C. 1889. Notice of gigantic horned Dinosauria from the Cretaceous. American Journal of Science, Series 3, 38: 173– 175.

Forty Years of Ceratophilia 15

Osborn, H. F. 1933. Mounted skeleton of Triceratops elatus. American Museum Novitates 654: 1–14. Ostrom, J. H. 1961. Cranial morphology of the hadrosaurian dinosaurs of North America. Bulletin of the American Museum of Natural History 122: 33–186. ———. 1964a. A reconsideration of the paleoecology of hadrosaurian dinosaurs. American Journal of Science 262: 975–997. ———. 1964b. A functional analysis of jaw mechanics in the dinosaur Triceratops. Postilla 88: 1–35. ———. 1966. Functional morphology and evolution of the ceratopsian dinosaurs. Evolution 20: 290–308. ———. 1969. Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Bulletin of the Peabody Museum of Natural History 30: 1–165. ———. 1970a. Stratigraphy and paleontology of the Cloverly Formation (Lower Cretaceous) of the Bighorn Basin area, Wyoming and Montana. Bulletin of the Peabody Museum of Natural History 35: 1–234. ———. 1970b. Archaeopteryx: Notice of a ‘‘new’’ specimen. Science 170: 537–538. Padian, K., J. R. Horner, and A. Lee. 2005. Sexual dimorphism in dinosaurs? A review of the evidence. Journal of Vertebrate Paleontology 25(3, Suppl.): 98A. Paul, G. S. 1987. The science and art of restoring the life appearance of dinosaurs and their relatives. In S. Czerkas and E. C. Olson, eds., Dinosaurs Past and Present, Vol. II, pp. 4–49. Seattle: University of Washington Press. Penkalski, P. G., Jr., and P. Dodson. 1999. The morphology and systematics of Avaceratops, a primitive horned dinosaur from the Judith River Formation (late Campanian) of Montana, with the description of a second skull. Journal of Vertebrate Paleontology 19: 692–711. Rega, E., R. Holmes, and A. Tirabasso. 2010. Habitual locomotor behavior inferred from manual pathology in two Late Cretaceous chasmosaurine ceratopsid dinosaurs, Chasmosaurus irvinensis (CMN 41357) and Chasmosaurus belli (ROM 843). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 340–354. Bloomington: Indiana University Press. Russell, D. A., and X. Zhao. 1996. New psittacosaur occurrences in Inner Mongolia. Canadian Journal of Earth Sciences 33: 637– 648. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, Southeastern Alberta. Journal of Paleontology 81: 376–396. Sampson, S. D., and M. J. Ryan. 1997. Variation. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, pp. 773–780. San Diego: Academic Press. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Sereno, P. C. 1987. The ornithischian dinosaur Psittacosaurus from the Lower Cretaceous of Asia and the relationships of the Ceratopsia. Ph.D. diss., Columbia University, New York.

16 dodson

———. 1990a. New data on parrot-beaked dinosaurs (Psittacosaurus). In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 203–210. Cambridge: Cambridge University Press. ———. 1990b. Psittacosauridae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 579–592. Berkeley: University of California Press. Sereno, P. C., and S. Chao. 1988. Psittacosaurus xinjiangensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northwestern China. Journal of Vertebrate Paleontology 8: 353–365. Sereno, P. C., S. Chao, Z. Cheng, and C. Rao. 1988. Psittacosaurus meileyingensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northwestern China. Journal of Vertebrate Paleontology 8: 366–377. Sereno, P. C., X. Zhao, L. Brown, and T. Lin. 2007. New psittacosaurid highlights skull enlargement in horned dinosaurs. Acta Paleontologica Polonica 52: 275–284. Smith, J. B., M. C. Lamanna, K. J. Lacovara, P. Dodson, J. R. Smith, J. C. Poole, R. Giegengack, and Y. Attia. 2001. A giant sauropod dinosaur from an Upper Cretaceous mangrove deposit in Egypt. Science 292: 1704–1706. Sternberg, C. M. 1927. Horned dinosaur group in the National Museum of Canada. Canadian Field-Naturalist 41: 67–73. ———. 1949. The Edmonton fauna and description of a new Triceratops from the Upper Edmonton member; phylogeny of the Ceratopsidae. Bulletin National Museum of Canada 113: 33–46. Tanke, D. 1988. Ontogeny and dimorphism in Pachyrhinosaurus (Reptilia, Ceratopsidae), Pipestone Creek, N.W. Alberta, Canada. Journal of Vertebrate Paleontology 8(3, Suppl.): 27A. Thompson, S., and R. Holmes. 2007. Forelimb stance and step cycle in Chasmosaurus irvinenesis (Dinosauria: Neoceratopsia). Palaeontologia Electronica 10: 1–17. Tumarkin, A. R., and P. Dodson. 1998. A heterochronic analysis of enigmatic ceratopsids. Journal of Vertebrate Paleontology 18(3, Suppl.): 83A. Tumarkin-Deratzian, A. R., and P. Dodson, P. 2005. A new look at old faces: Revisiting Monoclonius and Brachyceratops. Journal of Vertebrate Paleontology 25(3, Suppl.): 125A. Tumarkin-Deratzian, A. R., D. R.Vann, and P. Dodson. 2006. Bone surface texture as an ontogenetic indicator in long bones of the Canada goose Branta canadensis (Anseriformes: Anatidae). Zoological Journal of the Linnean Society 148: 133–168. ———. 2007. Growth and textural aging in long bones of the American alligator Alligator mississippiensis (Crocodylia: Alligatoridae). Zoological Journal of the Linnean Society 150: 1–39. Weishampel, D. B., P. Dodson, and H. Osmólska, eds. 1990. The Dinosauria. Berkeley: University of California Press. ———, eds. 2004. The Dinosauria, 2nd ed. Berkeley: University of California Press. Witmer, L. M. 2001. Nostril position in dinosaurs and other vertebrates and its significance for nasal function. Science 293: 850–853. Xu, X. 1997. A new psittacosaur (Psittacosaurus mazongshanensis sp. nov.) from Mazongshan area, Gansu Province, China. In Z.

Dong, ed., Sino-Japanese Silk Road Dinosaur Expedition, pp. 48– 67. Beijing: China Ocean Press. Xu, X., C. A. Forster, J. M. Clark, and J. Mo. 2006. A basal ceratopsian with transitional features from the Late Jurassic of northwestern China. Proceedings of the Royal Society of London B 273: 2135–2140. Xu, X., P. J. Makovicky, X. Wang, M. A. Norell, and H. You. 2002. A ceratopsian dinosaur from China and the early evolution of Ceratopsia. Nature 416: 314–317. You, H., and P. Dodson. 2003. Redescription of neoceratopsian dinosaur Archaeoceratops and early evolution of Neoceratopsia. Acta Paleontologica Polonica 48: 261–272. ———. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press. You, H., and Z. Dong. 2003. A new protoceratopsid (Dinosauria: Neoceratopsia) from the Late Cretaceous of Inner Mongolia. Acta Geologica Sinica 77: 299–303. You, H., D. Li, Q. Ji, M. C. Lamanna, and P. Dodson. 2005. On a new genus of basal neoceratopsian dinosaur from the Early Cretaceous of Gansu Province, China. Acta Geologica Sinica 79: 593–597. You, H., K. Tanoue, and P. Dodson. 2007. A new specimen of Liaoceratops yanzigouensis (Dinosauria: Neoceratopsia) from

the Early Cretaceous of Liaoning Province, P. R. China. Acta Geologica Sinica 81: 898–904. ———. In press. A new cranial specimen of the ceratopsian dinosaur Psittacosaurus major from the Early Cretaceous Yixian Formation of Liaoning Province, China. Acta Paleontologica Polonica 53. You, H., and X. Xu. 2005. An adult specimen of Hongshanosaurus houi (Dinosauria: Psittacosauridae) from the Lower Cretaceous of western Liaoning Province, China. Acta Geologica Sinica 79: 168–173. You, H., X. Xu, and X. Wang. 2003. A new genus of Psittacosauridae (Dinosauria: Ornithopoda) and the origin and early evolution of marginocephalian dinosaurs. Acta Geologica Sinica 77: 15–20. Zhao, X., Z. Cheng, and X. Xu. 1999. The earliest ceratopsian from the Tuchengzi Formation of Liaoning, China. Journal of Vertebrate Paleontology 19: 681–691. Zhao, X., Z. Cheng, X. Xu, and P. J. Makovicky. 2006. A new ceratopsian from the Upper Jurassic Houcheng Formation of Hebei, China. Acta Geologica Sinica 80: 467–473. Zhou C., K. Gao, R. C. Fox, and S. Chen. 2006. A new species of Psittacosaurus (Dinosauria: Ceratopsia) from the Early Cretaceous Yixian Formation, Liaoning, China. Palaeoworld 15: 100–114.

Forty Years of Ceratophilia 17

PART TWO SYSTEMATICS AND NEW CERATOPSIANS

2 Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs (Ceratopsia: Psittacosaurus) PA U L C . S E R E N O

in 1922, well-preserved fossils of the first parrot-

(Coombs 1980, 1982). For many years, Osborn’s two brief

beaked dinosaur were discovered in Early Cretaceous

notes on P. mongoliensis (Osborn 1923, 1924) and a descrip-

horizons in the Gobi Desert of Mongolia. Now referred to

tion of P. sinensis (Young 1958) provided most of the informa-

a single species, Psittacosaurus mongoliensis, these remains

tion available on psittacosaur morphology.

include a growth series from hatchlings to adults. In sub-

Recent Work. Sereno (1987) provided an overview of psit-

sequent years, 15 species have been added to the genus

tacosaur morphology. Portions of this dissertation were pub-

Psittacosaurus and a second genus, Hongshanosaurus, was

lished, including the description of two new species (P. meiley-

recently described, all from Early Cretaceous rocks in

ingensis, P. xinjiangensis; Sereno and Zhao 1988; Sereno et al.

Asia. Although the second genus and about one-half of

1988), the synonomy of several poorly known species (Sereno

the species attributed to Psittacosaurus are potentially in-

1990a), and an overview of the morphology of the clade Psit-

valid, Psittacosaurus remains the most species-rich dino-

tacosauridae (Sereno 1990b). Although most of this overview

saurian genus, with interspecific variation concentrated

can be found in You and Dodson (2004), reference is made

in the skull and dentition. This paper reviews evidence

only to the original source (Sereno 1990b).

differentiating the named genera and species of psit-

Russian psittacosaurs, including a partial skull first reported

tacosaurs, outlines major cranial changes in a growth se-

by Rozhdestvensky (1955, 1960) at Shestakovo in Siberia, be-

ries from hatchling to adult in Psittacosaurus

came the subject of a dissertation by Xijin Zhao under his

mongoliensis, and provides evidence of two species

direction. Renewed work at Shestakovo in 1994 has yielded

groups within the genus.

more complete skeletal remains described as P. sibiricus (Averianov et al. 2006).

Introduction

In China, P. meileyingensis (Sereno et al. 1988) was the first dinosaur described from the mid-Cretaceous Jehol fauna in the

Exceptional psittacosaur skeletons were discovered in 1922 as

Yixian and Jiufotang Formations of Liaoning Province (Xu and

the first major paleontological find of the Asiatic Expedi-

Norell 2006). In subsequent years, scores of non-avian dino-

tions led by Roy Chapman Andrews of the American Museum

saurs and basal avians have been described, including two new

of Natural History (Andrews 1932). Now attributed to a single

species of Psittacosaurus, P. lujiatunensis (Zhou et al. 2006b) and

species, Psittacosaurus mongoliensis, this material includes

P. major (Sereno et al. 2007), and a new genus and species,

complete skulls and skeletons of hatchlings as well as adults

Hongshanosaurus houi (You et al. 2003; You and Xu 2005).

21

Paleogeographic distribution of psittacosaurids and location of particularly fossiliferous or singular localities. Locality abbreviations, 1: Shestakovo (Siberia); 2: Delunshan (Xinjiang); 3: Bulasutuin; 4: Ondai Sair; 5: Khobur; 6: Oshih; 7: Ulan Osh; 8: Khuren Dukh; 9: Sharalin Ula; 10: Tsakurt; 11: Kharmin Us; 12: Suhongtu (Inner Mongolia); 13: Haratologay (Inner Mongolia); 14: Guyang (Inner Mongolia); 15: Hangginqi (Inner Mongolia); 16: Laiyang (Shandong); 17: Meileyingzi, Shangyuan (Liaoning). Four localities have yielded contemporaneous psittacosaur species: 12: P. mongoliensis, P. sp.; 15: P. neimongoliensis, P. sinensis (= P. ordosensis); 17 (Meileyingzi): P. meileyingensis, P. mongoliensis; 17 (Shangyuan): P. lujiatunensis, P. major.

FIGURE 2.1.

Finally, a new species, P. sp., was discovered in 2001 in the

Psittacosaur ‘‘Biochron.’’ The genus Psittacosaurus is so com-

Bayan Gobi Formation of the Inner Mongolia Autonomous

mon in mid-Cretaceous vertebrate faunas of northern and

Region (Fig. 2.1, locality 12; Fig. 2.2; Sereno et al. in review).

central Asia that a ‘‘psittacosaur’’ fauna (Dong 1973) or ‘‘bio-

In recent years hundreds of specimens of Psittacosaurus have

chron’’ (Lucas 2006) has been proposed, beginning in the

been collected in the vicinity of the type locality Oshih in

mid-Barremian (ca. 125 Ma) and extending to the mid-Albian

Mongolia, and perhaps thousands more have been collected

(ca. 105 Ma; Lucas 2006), for a duration of about 20 million

from the Lujiatun Beds of the Yixian Formation in Liaoning

years. The interval with abundant psittacosaurs, however,

Province in China (Xu and Norell 2006). Among the many

may only have been half as long as previously reported. New

specimens from the Lujiatun Beds is a nest of hatchlings with

radiometric dates from Liaoning Province (He at al. 2006;

an adult (Meng et al. 2004), a juvenile social group (Qi et al.

Zhou et al. 2007) confirm that the lowermost beds of the Yi-

2007), and an adult preserving integumentary bristles (Mayr

xian Formation (Lujiatun, Jianshangou), which include the

et al. 2002).

oldest psittacosaurs, are no older than 123–125 Ma, straddling

22 sereno

Table 2.1. Psittacosaurid Genera and Species Taxon

Author

Skull

Skeleton

Senior synonym

Taxa regarded as valid Psittacosaurus sp.

Sereno et al. in review

兹

兹

Psittacosaurus lujiatunensis

Zhou et al. 2006b

兹

兹

Psittacosaurus major

Sereno et al. 2007

兹

兹

Psittacosaurus meileyingensis

Sereno et al. 1988

兹

—

Psittacosaurus mongoliensis

Osborn 1923

兹

兹

Psittacosaurus neimongoliensis

Russell and Zhao 1996

兹

兹

Psittacosaurus sibiricus

Averianov et al. 2006

兹

兹

Psittacosaurus sinensis

Young 1958

兹

兹

Psittacosaurus xinjiangensis

Sereno and Zhao 1988

兹

兹

Junior synonyms Protiguanodon

Osborn 1923

Psittacosaurus

Hongshanosaurus

You et al. 2003

Psittacosaurus

Protiguanodon mongoliense

Osborn 1923

P. mongoliensis

Psittacosaurus

Young 1958

P. mongoliensis

Protiguanodonensis Psittacosaurus youngi

Zhao 1962

Psittacosaurus osborni

Young 1931

—

—

P. mongoliensis

P. sinensis

Psittacosaurus guyangensis

Cheng 1983

—

—

P. mongoliensis

Hongshanosaurus houi

You et al. 2003

兹

—

? P. lujiatunensis

Psittacosaurus mazongshanensis

Xu 1997

兹

—

Psittacosaurus ordosensis

Russell and Zhao 1996

兹

兹

Nomina dubia

Psittacosaurus tingi

Young 1931

—

—

Psittacosaurus sattayaraki

Buffetaut and

—

—

Suteethorn 1992

? ? P. sinensis ? Ceratopsia, incertae sedis

Note: The quality of cranial and postcranial remains for a given species is indicated by a checkmark for a relatively complete skull or articulated postcranial skeleton or a dash for more fragmentary material.

the Barremian-Aptian boundary (Gradstein et al. 2004). The

Nothing can be done about problem 1, except to realize

youngest psittacosaurs in Liaoning (P. mongoliensis, P. meiley-

in future taxonomic work that the proper identification of

ingensis) lie below the Aptian-Albian boundary (ca. 112 Ma),

most psittacosaur species requires relatively complete, well-

about 13 million years later.

prepared holotypic specimens for taxonomic resolution. For

Taxonomic Issues. Two genera and 17 species have been

example, P. guyangensis (Cheng 1983) and P. mazongshanensis

named over the years, all from Lower Cretaceous rocks in Asia

(Xu 1997) suffer from fragmentary holotypic specimens that

(Table 2.1). As I will argue below, more than one-half of these

allow only a limited number of comparisons. Unless there is a

are junior synonyms, the principal taxonomic shortcomings

compelling reason, future holotypic material within Psittaco-

being four in kind:

saurus should be limited to relatively complete skulls or skulls with associated skeletons.

(1) Fragmentary holotypic material that is difficult to assess (2) Immature material that may not yet have acquired

Problem 2 is well exemplified by a new genus and species, Hongshanosaurus houi (You et al. 2003; You and Xu 2005).

species-specific cranial proportions or full

Based on a compressed, immature, isolated skull that is ap-

development of processes and horns

proximately one-fourth adult size, H. houi is difficult to vali-

(3) Character states presented as diagnostic that have a

date as a taxon or to serve as the basis for referral of additional specimens. The solution to this problem is to identify post-

multi-species distribution (4) Character states presented as diagnostic that are

hatching features that appear with maturity, down-weighting

incompletely formulated or clearly correlated with

their absence in material that is clearly immature. In this re-

other character states

view, I attempt to establish the outlines of post-hatching cra-

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 23

nial transformation in Psittacosaurus mongoliensis, in order to

others that might characterize ceratopsians. In this review,

identify relative shape or other changes that may have been

I enumerate the most diagnostic features that are either

misinterpreted as taxonomic signal in immature specimens.

unique or thought to be derived for the species at hand. These

Future holotypic referral should avoid specimens that are

are often followed by other features that help to distinguish

clearly immature, because many of the nuanced features that

the species that have a limited distribution within Psittacosau-

distinguish species, such as raised edges, eminences, flanges,

rus. The characters in this paper that were compiled above the

and horns, appear late in post-hatching growth.

species level (Tables 2.2, 2.3) were formulated and edited with

Problem 3 requires review, as in Xu and Zhao (1999) and in the present work, in order to better understand character dis-

standards for completeness, testability, and independence in mind (Sereno 2007).

tributions. In several instances in this review, we do not have

Institutional Abbreviations. AMNH: American Museum of

enough comparative information to decide with confidence

Natural History, New York; BNHM: Beijing Natural History

whether several features currently functioning as species dif-

Museum, Beijing; CAGS-IG: Chinese Academy of Geological

ferentia, such as the width of the distal end of the ischium in P.

Sciences, Institute of Geology, Beijing; GI SPS: Geologic Insti-

neimongoliensis, are truly diagnostic or just further examples

tute, Section of Paleontology and Stratigraphy, Ulan Bator;

of individual or size-related variation.

IVPP: Institute of Vertebrate Paleontology and Paleoanthro-

Problem 4 requires a more diligent, guarded approach to the

pology, Beijing; JZMP: Jingzhou Museum of Paleontology,

formulation of characters. Sereno (1987: table 18), for ex-

Jingzhou; LH: Long Hao Institute for Stratigraphic Paleon-

ample, listed a ‘‘dorsally positioned’’ external naris as a syn-

tology, Hohhot; PIN: Paleontological Institute, Moscow;

apomorphy for Psittacosaurus. This has been classified as a

PKUP: Peking University Paleontological Collections, Beijing;

‘‘relative-geometric’’ character by Sereno (2007), because it

PM TGU: Paleontological Museum, Tomsk State University,

draws attention to the dorsal position of the external naris

Tomsk; UCRC: University of Chicago Research Collection,

relative to some other feature. That other feature, however, is

Chicago; UGM: Urümqi Geological Museum, Urümqi; ZMNH:

not specified in the character, although that specification is

Zhejiang Museum of Natural History, Hangzhou.

key to its evaluation in any terminal taxon. A complete char-

Anatomical Abbreviations. +: positive sclerotic plate; –: nega-

acter can be constructed from associated text (Sereno 1987:

tive sclerotic plate; 1–12: tooth or vertebral number; I–IV:

254), which pinpoints the ventral margins of the external

quadrants I–IV in a sclerotic ring; a: angular; ac: acromion; ai:

naris and orbit as the relative relationship of interest (Table

atlantal intercentrum; almf: anterolateral maxillary foramen;

2.2, character 2). A more considered (or complete) formu-

alp: alveolar pedestal; ana: atlantal neural arch; apd: articular

lation of the character in this case makes it possible for an-

surface for the predentary; apmf: anterior premaxillary fora-

other taxonomist to evaluate and score the condition among

men; aqj: articular surface for the quadratojugal; ar: articular;

specimens.

ari: attachment ridge; bo: basioccipital; bp: basal plate; bpt:

The second aspect of problem 4, character correlation, vio-

basipterygoid process; bs: basisphenoid; bt: basal tubera; C:

lates character independence, an underlying assumption for

cervical vertebra; c: coronoid; ce: centrum; ci: crista inter-

character data under a parsimony criterion (Sereno 2007).

fenestralis; cnIII–XII: cranial nerves III–XII; co: coracoid; cof:

Averianov et al. (2006: 363), for example, listed ‘‘skull width

coracoid foramen; d: dentary; D: dorsal vertebra; de: denticle;

exceeds skull length’’ as well as ‘‘premaxilla length to height

den: denticule; df: dentary flange; ec: ectopterygoid; emf: ex-

ratio less than 60%’’ as derived character states for two charac-

ternal mandibular fenestra; en: external naris; eo: exoccipital;

ters in P. sibiricus. The premaxilla in this species is particularly

f: frontal; flc: fenestra of the lacrimal canal; fm: foramen mag-

short anteroposteriorly. This condition, however, also results

num; fob: fossa for the olfactory bulb; gl: glenoid; ic: in-

in a proportionately shorter skull, the postorbital portion of

tercoronoid; if: incisive foramen; imf: internal mandibular

which then appears proportionately wider. Premaxilla length

fenestra; j: jugal; jfo: jugal fossa; jh: jugal horn; k: keel; l: lacri-

and skull length in this case appear to be dependent.

mal (or left); lc: lacrimal canal; lf: lacrimal foramen; lhv: lateral

Present Approach. The approach to generic and species tax-

head vein; lpmf: lateral premaxillary foramen; ls: laterosphe-

onomy in the present work acknowledges the utility of a dif-

noid; m: maxilla; mfo: maxillary fossa; mp: maxillary pro-

ferential diagnosis (Mayr et al. 1953), which lists both auta-

tuberance; mpmf: medial premaxillary foramen; n: nasal; na:

pomorphies as well as particular feature combinations that

neural arch; ncr: nuchal crest; nsu: nasal sulci; oc: occipital

differentiate species. Listing numerous symplesiomorphies in

condyle; od: odontoid; ofo: occipital fossa; op: opisthotic; p:

a species diagnosis as if it were a taxonomic key, however, is

parietal; pap: palpebral; pd: predentary; pl: palatine; pm: pre-

not illuminating or effective (Sereno 1990c: 16). In the di-

maxilla; pmri: premaxilla-maxilla ridge; po: postorbital; pocr:

agnoses in their review of psittacosaur species, for example,

postorbital crest; poh: postorbital horn; pojcr: postorbital-

Xu and Zhao (1999) mix together unique features with others

jugal crest; pojfo: postorbital-jugal fossa; pojh: postorbital-

that have limited distribution among psittacosaurs and still

jugal horn; popr: paroccipital process; ppf: postpalatine fora-

24 sereno

men; pr: prootic; pra: prearticular; prf: prefrontal; prfcr: pre-

skull over 150 mm in length was subsequently described and

frontal crest; pri: primary ridge on crown; ps: parasphenoid;

referred to the same genus and species (You and Xu 2005; IVPP

psqs: parietosquamosal shelf; pt: pterygoid; ptfo: pterygoid

V12617). The latter specimen clearly exhibits many of the de-

fossa; ptmr: pterygoid mandibular ramus; q: quadrate; qf:

rived features common to all psittacosaurs in the genus Psit-

quadrate foramen; qj: quadratojugal; qjp: quadratojugal pro-

tacosaurus (Table 2.2). Only one feature was cited to support

tuberance; r: rostral (or right); rf: replacement foramen; rp:

their phylogenetic interpretation of Hongshanosaurus loui

retroarticular process; sa: surangular; saf: surangular foramen;

as the sister taxon to all known species of Psittacosaurus—

sc: scapula; scb: scapular blade; scr: sclerotic ring; sh: shaft; so:

preorbital snout length approximately 50% of skull length

supraoccipital; sp: splenial; sq: squamosal; st: stapes; stf: sta-

(You and Xu 2005: 172). In other psittacosaurs the snout is

pedial footplate; stp: sternal plate; sts: stapedial shaft; sym:

proportionately shorter, measuring 40% or less of skull length. Anteroventral crushing of the cranium in both specimens

symphysis; ts: triturating surface; v: vomer.

of Hongshanosaurus houi (You et al. 2003; You and Xu 2005),

Systematic Paleontology

however, brings into question the supposedly longer snout proportions, as well as the additional features (oval shape of

Dinosauria Owen 1842

the external naris, orbit, and laterotemporal fenestra) cited as

Ornithischia Seeley 1888

autapomorphies. In the better-preserved adult skull, the ante-

Ceratopsia Marsh 1890

rior end of the lower jaw protrudes slightly beyond the upper

Psittacosauridae Osborn 1923

jaw (You and Xu 2005: fig. 1a); the upper portion of the cranium thus has been displaced posteroventrally, lowering its

Phylogenetic Definition. The most inclusive clade containing

profile and lengthening the distance anterior to the orbit. Ma-

Psittacosaurus mongoliensis Osborn 1923 but not Triceratops

jor cracks in this specimen course posterodorsally at a right

horridus Marsh 1889 (Sereno 2005; Sereno et al. 2005).

angle to the direction of compression. Overlooked in the description of H. loui is the presence of

Diagnosis. Same as Psittacosaurus. Included Genera. Psittacosaurus.

autapomorphies subsequently used to differentiate another

Remarks. Sereno (2005) provided the first phylogenetic defi-

psittacosaur species from the same horizon, Psittacosaurus lu-

nition for this taxon, when it appeared there were two genera

jiatunensis (Zhou et al. 2006b). These include the narrow pre-

of psittacosaurs (Psittacosaurus, Hongshanosaurus). With Hong-

frontal width (relative to nasal width) and the apparent con-

shanosaurus reduced to a junior synonym of Psittacosaurus (see

tact between the jugal and quadrate (You et al. 2003; You and

below), the familial taxon is redundant with the genus Psit-

Xu 2005). Shared derived features between H. houi and P. lu-

tacosaurus. ‘‘Psittacosauridae’’ and its vernacular ‘‘psittacosau-

jiatunensis include the absence of an external mandibular

rid’’ thus are not used in the remainder of this paper but could

fenestra and, as preserved in the adult skulls of both taxa, the

become ‘‘active’’ (Sereno et al. 2005) if additional genera allied

presence of a prominent dentary flange with a squared ante-

with Psittacosaurus are described in the future.

rior corner. It seems quite likely that these specimens pertain to the same species of psittacosaur, the most common dino-

Psittacosaurus Osborn 1923

saur found in the Lujiatun Beds of the Yixian Formation. Dis-

Synonomy. Protiguanodon, Hongshanosaurus.

tortion of cranial remains from this horizon is well known, as

Protiguanodon (Osborn 1923). The second genus and species

can be seen by comparison of the holotypic and paratypic

described by Osborn (1923), Protiguanodon mongoliense, has

skulls attributed to P. lujiatunensis (Zhou et al. 2006b). Why

been both regarded as a junior synonym of P. mongoliensis

these similarities were not mentioned by Zhou et al. (2006b)

(Rozhdestvensky 1955, 1977; Coombs 1982) and referred to a

in their description of P. lujiatunensis is not clear, as elsewhere

new species, Psittacosaurus protiguanodonensis (Young 1958).

many of the same authors criticized the adequacy of the holo-

Osborn’s reasons for distinguishing these two Mongolian taxa

typic skull of H. houi (Zhou et al. 2006a).

were reevaluated by Sereno (1987, 1990a, b), and the two

If these specimens pertain to the same species, as seems pos-

holotypic skeletons were found to lie within the variation

sible, a case could be made that P. lujiatunensis (Zhou et al.

present in the large collection of Psittacosaurus mongoliensis

2006b) is the junior synonym of Psittacosaurus (= Hongshano-

discovered by the American and Soviet-Mongolian expedi-

saurus) houi (You et al. 2003). Reevaluation of the material,

tions. Protiguanodon and Psittacosaurus protiguanodonensis thus

however, favors the course taken here, which is to regard

are regarded as junior synonyms of Psittacosaurus and P. mon-

Hongshanosaurus as a junior synonym of Psittacosaurus and

goliensis, respectively (Table 2.1).

P. houi as a nomen dubium. The immaturity of the holotypic

Hongshanosaurus (You et al. 2003). The genus Hongshanosau-

skull of H. houi undermines any significance given to the ab-

rus was erected on the basis of a dorsoventrally crushed, im-

sence of features that appear with age, such as the prominence

mature skull (IVPP V12704) about 35 mm in length. An adult

of the jugal horn or presence of a dentary flange.

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 25

Several of the features linking H. houi and P. lujiatunensis, fur-

tion or contact present, (8) maxillary fossa, (9) maxillary pro-

thermore, are not uniformly present in their respective holo-

tuberance, (10) fenestra of the lacrimal canal, (11) antorbital

typic specimens. The prefrontal, for example, is proportion-

fenestra and fossa absent, (12) postorbital posterior process

ately narrow in the holotypic specimen of H. loui (You et al.

extends along the entire supratemporal bar, (13) end of squa-

2003; IVPP V12704) but is subequal in width to the nasal in the

mosal anterior process situated on the dorsal aspect of the

referred adult skull (You and Xu 2005; IVPP V12617). In P. lujia-

postorbital, (14) pterygoid with neomorphic palatal lamina

tunensis, the prefrontal is shown as narrow relative to the nasal

forming the basal plate, (15) pterygoid with hypertrophied

only in the paratypic skull (PKUP V1054); the nasal-prefrontal

mandibular ramus, (16) medial quadrate condyle planar,

suture does not appear to be preserved in the holotypic skull

(17) laterally divergent palpebral with transverse posterior

(Zhou et al. 2006b; ZMNH M8137). Thus it is not clear if this

margin, (18) predentary with very short, tongue-shaped ven-

feature is variable, or if the apparent variability is simply an

tral processes, (19) predentary with semicircular anterior mar-

artifact of preservation. The thin medial edge of the prefrontal

gin, (20) dentary with ventral ridge or flange, (21) articular

overlaps the nasal and is often partially broken away in speci-

with planar surface for quadrate condyles, and (22) dentary

mens of P. mongoliensis, giving the prefrontal a proportionately

teeth with bulbous cone-shaped primary ridge with secondary

narrower appearance. Finally, jugal-quadrate contact is an au-

ridging (see Table 2.2).

tapomorphy of P. lujiatunensis preserved in both holotypic and paratypic skulls (Zhou et al. 2006b). Such contact may occur in

Psittacosaurus sp. Sereno et al. (in review)

the holotypic specimen of H. loui (You et al. 2003), although

Figure 2.2

the posterior end of the bone is not well preserved. Jugal-

Holotype. LH PV2, skull and articulated skeleton.

quadrate contact is shown only on one side of the referred skull

Type Horizon and Locality. Bayan Gobi Formation (Aptian);

of H. loui (You and Xu 2005). The variability of this feature,

N 40\ 59% 42.4&, E104\ 3% 53.8&, southwest of Suhongtu, Nei

thus, may also be a consequence of preservational factors.

Mongol Autonomous Region, People’s Republic of China

Given the tremendous abundance of psittacosaur remains

(Fig. 2.1, locality 12).

that were available by 2003 from the Lujiatun Beds of the

Diagnosis. Psittacosaur characterized by autapomorphies in-

Yixian Formation, it is unfortunate that Hongshanosaurus loui

cluding (1) pyramidal horn on the postorbital bar composed

and Psittacosaurus lujiatunensis were not established on the

almost entirely of the postorbital, (2) postorbital-jugal fossa,

basis of mature, well-prepared skeletons with complete skulls

(3) minimum width of the postorbital bar approximately 50%

as in P. major (Sereno et al. 2007). If Hongshanosaurus loui and

the width of the base of the process, (4) retroarticular process

Psittacosaurus lujiatunensis represent the same species within

deflected posteromedially at an angle of 40\ from the axis of

Psittacosaurus, the latter would be a junior synonym of the

the mandible.

former. It seems most prudent, however, to tentatively recog-

Other features with limited distribution among psittacosaur

nize the species P. lujiatunensis on the basis of its more com-

species include preorbital snout length 35% the length of the

plete, more mature, and less distorted holotypic and paratypic

skull, a ventrolaterally projecting pyramidal jugal horn, a low

skulls (ZMNH M8137, PKUP V1054). Following this course,

quadratojugal eminence, no development of dentary flanges

the species H. loui is here regarded as a nomen dubium, be-

or an external mandibular fenestra, and maxillary and den-

cause of the crushed, immature state of the holotypic speci-

tary tooth rows limited to eight teeth.

men. It seems likely that the referred adult skull can be shown to pertain to P. lujiatunensis.

Remarks. The holotypic specimen of P. sp. is a fully mature individual showing complete fusion at the mandibular

Known Distribution. Psittacosaur distribution is currently

symphysis and within the axial column (Fig. 2.2A). The skull

limited to northeast Asia, with specimens found as far north

is slightly fractured and twisted in dorsal view but other-

as Siberia, as far east as Japan, as far west as Xinjiang, and as far

wise superbly preserved. The skeleton is preserved in three-

south as Shandong (Fig. 2.1; Lucas 2006). Poorly preserved jaw

dimensions, including a fully articulated ribcage with a gastro-

fragments from Thailand (Buffetaut and Suteethorn 1992) are

lith mass. The specimen is thus well preserved for identifying

not regarded here as referable to the genus Psittacosaurus (see

species-level features. The pyramidal postorbital horn is pre-

‘‘Systematic Paleontology’’).

served on both sides (Fig. 2.2B) and differs in shape and com-

Diagnosis. Ceratopsian dinosaurs with (1) preorbital skull

position from that in P. sinensis, where the horn is more elon-

less than 40% of skull length, (2) external naris with ventral

gate and split between the postorbital and jugal. Just dorsal to

margin dorsal to that of the orbit, (3) nasal internarial process

the horn, the postorbital bar is very narrow. The posterior

extending ventral to external naris, (4) rostral-nasal contact

process of the postorbital is deeply emarginated and rather

present, (5) premaxilla dorsolateral process maximum width

narrow in lateral view as a result. The edge of the emargination

subequal to dorsoventral orbital diameter, (6) premaxilla-

terminates as a straight, horizontal crest positioned below the

prefrontal contact present, (7) premaxilla-jugal approxima-

center of the body of the postorbital.

26 sereno

FIGURE 2.2.

Psittacosaurus sp. (LHPV2). Skull in (A) lateral view; (B) enlarged view of postorbital-jugal horn in lateral view; (C) enlarged view of maxillary crowns in lateral view. Scale bars are (A) 3 cm; (B) 2 cm; (C) 1 cm. Reproduced in color on the insert.

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 27

P. sp. constitutes yet another pattern of character states not

with features that have a much broader distribution, such as

previously reported within Psittacosaurus (Sereno et al. in re-

closure of the mandibular fenestra (Zhou et al. 2006b). One

view). The three autapomorphies listed above seem to be

feature is shared only with P. major, a fossa on the broad ante-

uniquely expressed in P. sp. The pyramidal horn on the post-

rior ramus of the jugal, here termed the central jugal fossa.

orbital bar has a rather smooth surface and is constructed

Another feature, a ridge ascending from the maxillary pro-

solely of the postorbital, rather than a combination of the

tuberance, is present in two other species, P. major and P. mei-

postorbital and jugal as in P. sinensis. The straight crest on the

leyingensis. P. lujiatunensis lacks the very narrow nasal and

body of the postorbital (Fig. 2.2B) contrasts with the arched

frontal proportions and the elongate basipterygoid processes

crest in P. mongoliensis and P. sinensis. The strong medial de-

in P. major (Sereno et al. 2007). P. lujiatunensis is probably clos-

flection of the retroarticular process is also unique among psit-

est to P. major, both of which come from the Lujiatun Beds of

tacosaur species. All of these features are preserved on both

the Yixian Formation. Future comparative study of additional

sides of the skull.

skulls and skeletons will either verify their distinguishing fea-

A relatively small-bodied species, P. sp. has skull and skeletal

tures or suggest that they represent variants of a single species.

lengths comparable to those of P. sinensis (Young 1958; Sereno 1987). The skull, in addition, has a preorbital length of about

Psittacosaurus major Sereno et al. 2007

35% of skull length, a postorbital horn and associated jugal

Figures 2.3–2.6

fossa, and low quadratojugal eminence. There is no devel-

Holotype. LH PV1, skull and skeleton (Sereno et al. 2007).

opment of a dentary flange, and only eight teeth in maxil-

Notable Referred Specimens. JZMP-V-11, skull and skeleton

lary and dentary tooth rows (Fig. 2.2C). These features more

(Lü et al. 2007); CAGS-IG-VD-004, partial skull (You et al.

closely resemble the condition in P. sinensis than in P. mon-

2008).

goliensis or P. lujiatunensis. However, P. sp. more closely re-

Type Horizon and Locality. Lujiatun Beds, lowermost Yixian

sembles P. mongoliensis than P. sinensis in other regards, such as

Formation (late Barremian or earliest Albian); Lujiatun (near

the lack of shortening of the lower jaw relative to the upper

Beipiao), Liaoning Province, People’s Republic of China (Fig.

jaw, maxillary and jugal fossae, a ventrolaterally projecting

2.1, locality 17).

jugal horn, and an arched crest on the body of the postorbital.

Revised Diagnosis. Psittacosaur characterized by autapomorphies including (1) maximum width across nasals and

Psittacosaurus lujiatunensis Zhou et al. 2006b

interorbital frontal width subequal to maximum width of the

Holotype. ZMNH M8137, skull with lower jaws (Zhou et al.

rostral, (2) tall subtriangular laterotemporal fenestra with an-

2006b: fig. 2).

teroposterior width of the ventral margin approximately 25%

Paratypes. ZMNH M8138, skull with jaws and cervicals 1–3;

of dorsoventral height, (3) anterior ramus of jugal convex

PKUP, V1053, juvenile skull with fragmentary lower jaws

(best seen in dorsal view), (4) elongate basipterygoid processes

(Zhou et al. 2006b: fig. 4); PKUP V1054, adult skull with lower

subequal in length to the body of the basisphenoid as mea-

jaws, proatlas and atlas (Zhou et al. 2006b: fig. 3).

sured from the notch between the processes to the basal tu-

Type Horizon and Locality. Lujiatun Beds, lowermost Yi-

bera, (5) hypertrophied dentary flange with anterior corner

xian Formation (late Barremian or earliest Albian); Lujiatun

approximately 30% of the depth of the dentary ramus and

(near Beipiao), Liaoning Province, People’s Republic of China

with only a short gap to the predentary, and (6) seven sacral

(Fig. 2.1, locality 17).

vertebrae (addition of one dorsosacral).

Revised Diagnosis. Psittacosaur characterized by autapomor-

Other features with limited distribution among psittacosaur

phies including (1) prefrontal width less than 50% that of

species include preorbital snout length 33% of skull length, a

the nasal, (2) quadratojugal-squamosal contact along anterior

ventrolaterally projecting pyramidal jugal horn, closure of the

margin of quadrate shaft, and (3) jugal-quadrate contact pos-

external mandibular fenestra, and maximum depth of the an-

teroventral to the laterotemporal fenestra.

gular greater than that for the surangular. Sereno et al. (2007)

Other features with limited distribution among psittacosaur

listed among diagnostic features the relative size of the skull,

species include preorbital snout length equal to 35% skull

which is approximately 40% of trunk length (Lü et al. 2007).

length, a ventrolaterally projecting pyramidal jugal horn, ab-

This proportion, however, characterizes several psittacosaur

sence of a horn on the postorbital bar, a central jugal fossa, a

species, including cf. P. lujiatunensis (Mayr et al. 2002), P. xin-

low quadratojugal eminence (paratypic skulls; PKUP V1053,

jiangensis (Sereno and Zhao 1988), and P. sinensis (Young

V1054), strong dentary flange with anterior corner, external

1958). P. mongoliensis, with a skull length only 30% of trunk

mandibular fenestra closed, and maxillary and dentary tooth

length, is either primitive among psittacosaurs or has reverted

row limited to eight teeth.

back to a proportion common to many other small-bodied

Remarks. The revised diagnosis emphasizes three autapo-

ornithischians. You et al. (2008: 195) also noted the narrow

morphies that were cited in the original diagnosis but mixed

width of the ventral portion of the laterotemporal fenestra as

28 sereno

200 mm, matched in size only by P. sibiricus (Averianov et al. 2006). Only a few of the many hundred specimens of P. mongoliensis approach the size of these taxa (Sereno 1987). P. major is distinguished by the proportions of the nasal and frontal, which are narrower in P. major than in any other psittacosaur; by the narrow ventral margin of the laterotemporal fenestra that gives it a subtriangular shape in lateral view; by the convexity of the broad anterior ramus of the jugal that appears to have accommodated enlarged adductor musculature en route to the dentary; by basipterygoid processes that are approximately twice as long as in other psittacosaurs; by the extreme depth and anterior extension of the dentary flange, which approaches the predentary; by the narrow width of the prefrontal; and by the absence of jugal-quadrate and quadratojugal-squamosal contact, which is present in P. lujiatunensis. The presence of seven sacral vertebrae in P. major, one more than recorded in other psittacosaurs, cannot be assessed in P. lujiatunensis from published information. A complete skull and skeleton that appears to be referable to P. major was published independently by Lü et al. (2007: figs. 1, 2). It shows similar development of head-trunk proportions, hypertrophy and anterior extension of the dentary flange, and absence of quadratojugal-squamosal contact. The jugal approaches and may contact the quadrate in this specimen. At least one trunk vertebra is missing, and the anteriormost sacral rib to the additional dorsosacral appears to have been broken (Lü et al. 2007: fig. 1). P. major and P. lujiatunensis are differentiated based on skulls of mature individuals. The difference in the length of the basipterygoid processes is significant. The feature most likely attributable to preservational factors is the narrow width of the prefrontal in P. lujiatunensis, which may be due to breakage along its thin medial edge and is well documented only in a single paratypic skull (Zhou et al. 2006b: fig. 3). Similarly, the novel sutural contacts of the jugal and quadratojugal dorsal to the jaw articulation in P. lujiatunensis may be subject to variaPsittacosaurus major (LH PV1). Skull in (A) lateral view; (B) dorsal view. See text for abbreviations. Grey tone indicates matrix; cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bar is 10 cm.

FIGURE 2.3.

tion; in P. major these bones are separated by only a narrow margin. These sutural contacts, however, have not been demonstrated to occur in other species. The condition in other well-preserved adult skulls and description of postcranial remains from the Lujiatun Beds are needed to verify the distinction between these two large psittacosaurs.

characteristic of P. major. They proposed, in addition, that the maximum length of the skull exceeds its width across the jugal

Psittacosaurus meileyingensis Sereno et al. 1988

horns in this species. This proportion, however, is often sub-

Holotype. IVPP V7705, skull with jaws lacking the jugal horns

ject to postmortem distortion and also changes with growth

and central body of the premaxilla (Sereno et. al. 1988: figs. 2, 3).

(the skull increasing in relative width). These two dimensions,

Notable Referred Specimens. CAGS-IG V330, adult skull lacking

furthermore, are subequal in the skull they referred to P. major

the anterior end of the snout and partial postcranial skeleton

(You et al. 2008: fig. 1C2).

(Sereno et. al. 1988: figs. 8D, 9).

Remarks. P. major and P. lujiatunensis, named independently

Type Horizon and Locality. Meileyingzi Beds, Jiufotang Forma-

from the same formation (Zhou et al. 2006b, Sereno 2007), are

tion (Aptian); Beipiao, Liaoning Province, People’s Republic of

large-bodied psittacosaurs with adult skull length in excess of

China (Fig. 2.1, locality 17).

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 29

Psittacosaurus major (LH PV1). (A) Stereopairs of anterior palate in ventral view; (B) stereopairs of posterior palate and braincase in posteroventral view. See text for abbreviations. Scale bar is 3 cm. Reproduced in color on the insert.

FIGURE 2.4.

30 sereno

Psittacosaurus major (LH PV1) stereopairs of occiput in posteroventral view. See text for abbreviations. Scale bar is 3 cm. Reproduced in color on the insert.

FIGURE 2.5.

Psittacosaurus major (LH PV1). Maxillary crown in (A) lateral and (B) medial views; (C) dentary crown in medial view. Scale bar is 5 mm. Reproduced in color on the insert.

FIGURE 2.6.

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 31

Diagnosis. Psittacosaur characterized by autapomorphies in-

(holotype of Protiguanodon mongoliense) (Sereno 1990b: fig.

cluding (1) preorbital length only approximately 30% of skull

15.3); AMNH 6257, fragmentary vertebrae, distal left hu-

length, (2) subtriangular orbit with acute ventral corner, and

merus, and the left carpus and manus lacking only the ulnare

(3) rugose quadratojugal eminence.

and terminal phalanx of digit IV; AMNH 6260, partial skull

Other features with limited distribution among psittacosaur

and skeleton with complete carpus and manus and skin im-

species include a ridge ascending from the maxillary protuber-

pressions on metatarsus; AMNH 6534, skull with lower jaws

ance, a strong dentary flange, and a small external mandibular

and hyoids (Colbert 1945: fig. 5B), atlas and three anterior

fenestra.

caudal vertebrae, fragmentary ribs, right scapulocoracoid,

Remarks. P. meileyingensis was described on the basis of two

fragmentary right ilium, right femur, and the phalanges of

adult skulls, one with a partial postcranial skeleton. As with

right pes digit I; AMNH 6535, hatchling skull with jaws and

P. major and P. lujiatunensis from the Yixian Formation, P. mon-

partial skeleton including the atlas, dorsal vertebrae, ribs,

goliensis has been recovered along with P. meileyingensis in the

both sternals, both scapulae, and both coracoids (Coombs

overlying Jiufotang Formation (Sereno et al. 1988). In the case

1982: fig. 4, pl. 14); AMNH 6536, crushed skull with lower

of the latter species, however, there is no question regarding

jaws, sclerotic ring, atlas and axis, and articulated and dis-

its distinction.

articulated postcrania pertaining to many juvenile individuals including vertebrae, numerous partial and complete gir-

Psittacosaurus mongoliensis

dle elements, limb bones, an articulated hind limb and tail,

Figure 2.7

and a partial left pes (Coombs 1982: figs. 1, 2, 5, 6).

Synonomy. P. tingi, P. osborni, P. guyangensis (Table 2.1). Young

Type Horizon and Locality. Khukhtek Formation (Aptian-

(1958) regarded P. tingi (Young 1931) as a junior synonym of

Albian); Oshih, Ovorkhangai, Mongolian People’s Republic

P. osborni (Young 1931), both species based on immature, frag-

(Fig. 2.1, locality 6; Lucas 2006).

mentary specimens from the Inner Mongolia Autonomous

Diagnosis. Psittacosaur characterized by autapomorphies in-

Region (Sereno 1987, 1990b). You and Dodson (2004: table

cluding (1) a raised lip on the orbital margin of the prefrontal

22.1, 490) listed P. osborni as a valid species, although it was

and (2) transverse expansion of the distal end of the ischial

not among the valid species cited in the text. In the partial

blade to approximately twice its width at mid-shaft.

skull referred to P. osborni (Sereno 1990a: fig. 15.5), the shallow

Other features with limited distribution among psittaco-

angular and open external mandibular fenestra favors synon-

saurs include the presence of a maxillary fossa, maxillary pro-

omy with P. mongoliensis (Sereno 1987, 1990b; Xu and Zhao

tuberance, and low dentary flange. The raised lip on the pre-

1999), although these features may occur in juvenile speci-

frontal appears among juveniles less than one-half adult size,

mens of other species. Because several psittacosaur species

and may show some variation in adults. It cannot be consid-

have been recorded from Inner Mongolia (P. mongoliensis,

ered a mark of immaturity, however, as it is present in several

P. neimongoliensis, P. ordosensis, P. mazongshanensis), reference

adults in which many of the elements of the skull roof are

to P. mongoliensis remains tentative.

coossified (e.g., AMNH 6254, 6534).

Psittacosaurus guyangensis is based on the anterior portion of

Remarks. P. mongoliensis has slightly longer preorbital pro-

a skull (CAGS-IG V351), which is about two-thirds the size of

portions than any other psittacosaur species, usually measur-

the holotypic skull of P. mongoliensis. Disarticulated post-

ing 37% rather than 30–35% of skull length. This propor-

cranial remains of several individuals of varying maturity

tional difference is subtle and is highly dependent on skull

were referred to this taxon from the same locality (Cheng

orientation and choice of relative metric, as discussed below

1983). In the skull piece, a maxillary fossa and pendant maxil-

(see description, general adult skull shape). For the measure-

lary protuberance are present, the predentary extends to the

ments in this paper, the skull is oriented with the maxillary

anterior end of the snout, and there are at least nine teeth in

tooth row held horizontal.

the maxillary series (Cheng 1983: fig. 30). These features are consistent with P. mongoliensis, to which it was tentatively referred (Sereno 1987, 1990b; Xu and Zhao 1999) (Table 2.1). As with P. osborni, You and Dodson (2004: table 22.1, 490) listed P. guyangensis as a valid species, although it was not among the valid species cited in the text. Holotype. AMNH 6254, skull that includes the right sclerotic

Psittacosaurus neimongoliensis Russell and Zhao 1996 Holotype. IVPP 12-0888-2, articulated skeleton with skull (Russell and Zhao 1996: figs. 1, 3, 4) Notable Referred Specimens. IVPP 12-0888-3, right side of skull and anterior portion of skeleton; IVPP 07-0888-11, partially disarticulated skull.

ring, left stapes, and ceratohyals, and an articulated skeleton

Type Horizon and Locality. Ejinhoro Formation (Aptian-

lacking only several anterior caudal vertebrae and most of the

Albian); 80 km west of Dongshen, Ordos Basin, Nei Mongol

right hind limb (Sereno 1990b: fig. 15.2).

Autonomous Region, People’s Republic of China (Fig. 2.1, lo-

Notable Referred Specimens. AMNH 6253, fragmentary skull with articulated postcranial skeleton with gastrolith mass 32 sereno

cality 15; Dong 1993). Diagnosis. Psittacosaur characterized by autapomorphies in-

Psittacosaurus mongoliensis skull reconstruction based on the holotype (AMNH 6254) and referred skulls (AMNH 6534, PI/3779/10, 3779/12, 3779/20; IVPP V7668). (A) lateral view; (B) lower jaw in medial view. Cross-hatching indicates section through bone. See text for abbreviations.

FIGURE 2.7.

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 33

cluding (1) posterior end of the nasal contacting its opposite

cluding (1) laterotemporal fenestra subequal in maximum

in the midline (not separated by the frontal), (2) frontal in-

height and anteroposterior length, (2) postorbital ventral pro-

terorbital width approximately 30% of frontal length, and

cess with subvertical orientation set at an angle of approxi-

(3) postorbital extending along the margin of the orbit (rather

mately 95\ to the posterior process, (3) postorbital with small

than inset from the margin by the frontal in dorsal view).

dorsal horn, (4) enlarged palpebral subequal in transverse

Other features with limited distribution among psittacosaur

width to the adjacent skull roof, (5) palpebral posterior mar-

species include skull length approximately 40% of trunk

gin nearly straight and angled anterolaterally, (6) predentary

length, preorbital snout length approximately 35% the length

dorsoventrally compressed with a wedge-shaped profile with

of the skull, and absence of the external mandibular fenestra.

external margins set at approximately 30\, (7) angular with

The upper temporal bar angles posterolaterally, a condition

arcuate ventral extension of the dentary flange, (8) angular

similar to that in P. sinensis, although this may be due to dam-

process projecting laterally at posterior end of the ventral

age to the posterior skull table.

flange of the mandible, and (9) 14 dorsal vertebrae (one added).

Remarks. P. neimongoliensis appears to be most closely related

Other features with limited distribution among psittacosaur

to P. mongoliensis. There is no development of a postorbital-

species include preorbital snout length 35% the length of the

jugal horn, and the dentary flange is modestly developed. Un-

skull, a ventrolaterally projecting pyramidal jugal horn, and

like P. mongoliensis, however, the skull is large relative to the

no development of a maxillary protuberance, maxillary fossa,

trunk, as in all other species. The preorbital portion of the

or external mandibular fenestra.

skull, in addition, is less than 35% of skull length, the prefron-

Remarks. The holotypic specimen of P. sibiricus is a fully ma-

tal does not have a raised lateral edge, the anterior process of

ture individual with coossification of several of the cranial

the squamosal stops short of the body of the postorbital, the

sutures (Averianov et al. 2006). A distinctive species, P. sibiricus

external mandibular fenestra is closed, and the distal end of

is characterized by a suite of autapomorphies. Of the 9 listed

the ischial blade is not broadened. The lack of maturity of the

above, Averianov et al. (2006) either listed them as autapo-

holotype is potentially problematic regarding a few of these

morphies (1, 8, and 9) or described them in the text. I modi-

features, such as the short preorbital proportions of the skull

fied the first autapomorphy, which describes the shape of the

and the raised prefrontal lip. These features may only appear

laterotemporal fenestra, to avoid the use of skull length as a

late in growth in P. mongoliensis, a species that is also present

relative measure, the preorbital portion of which is subject to

in similar-age rocks elsewhere in Inner Mongolia.

change in other species. I also combined the reorientation and

In addition to the second autapomorphy cited above, Rus-

straightness of the posterior margin of the palpebral, consid-

sell and Zhao (1996) cited as diagnostic characters the length of

ering both of these to describe its altered shape. Finally, I ex-

the ischium (longer than the femur), its proportionately nar-

cluded three autapomorphies listed by Averianov et al. (2006:

row distal end, and the shorter length of the squamosal ante-

363) as too vaguely expressed to adequately test. These in-

rior process. The former may be an artifact of measurement;

clude the deep proportions of the premaxilla, the short medial

when ischial length is measured from the acetabulum to the

process of the postorbital, and the deep cleft in the posterior

distal end, its length in P. neimongoliensis is comparable to that

ramus of the jugal. The first two cannot be evaluated in the

in several other psittacosaurs including P. mongoliensis (just

articulated skull as the sutures are not shown. The edges of

longer than the femur). The other two features are plesiomor-

these bones may well have been subject to breakage when

phic within Psittacosaurus. The diagnostic features of the spe-

disarticulated, especially the thin edges of the premaxilla,

cies are currently limited to minor sutural variation and pro-

but breakage is not indicated in available views of these bones.

portional differences on the skull table. More information is

The posterior cleft in the jugal is difficult to differentiate

needed on the morphology of the holotype and referred speci-

from several other species on available evidence. Thus, fur-

mens of P. neimongoliensis to confirm the distinctiveness of this

ther information is needed to shore up these features as viable

species.

autapomorphies.

Psittacosaurus sibiricus Averianov et al. 2006

Psittacosaurus sinensis Young 1958

Holotype. PM TGU 16/4-20, articulated skeleton with skull

Figures 2.8–2.10

(Averianov et al. 2006: fig. 2).

Synonomy. P. youngi. From the same beds and near the locali-

Notable Referred Specimens. PM TGU 16/4-21, skull with ar-

ties of the holotypic specimen of P. sinensis, a second species

ticulated postcranial skeleton; found with holotypic skeleton.

was described, P. youngi (Zhao [Chao] 1962), based on a par-

Type Horizon and Locality. Ilek Formation (Aptian-Albian);

tial skeleton with a well-preserved skull (BNHM BPV149; Fig.

Shestakovo 3, Shestakovo, Kemerovo Province, Russia (Fig.

2.10). The diagnosis, however, included no autapomorphies

2.1, locality 1).

to distinguish the specimen, which has more recently been

Diagnosis. Psittacosaur characterized by autapomorphies in-

34 sereno

referred to P. sinensis (Table 2.1; Sereno 1987, 1990a, b).

Psittacosaurus sinensis (IVPP V738) skull in (A) left lateral view; (B) right lateral view. See text for abbreviations. Grey tone indicates matrix; cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bar is 3 cm.

FIGURE 2.8.

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 35

FIGURE 2.9.

Psittacosaurus sinensis (IVPP V738) skull in (A) dorsal view; (B) anterior view. See text for abbreviations. Grey tone indicates matrix; cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bar is 3 cm.

Holotype. IVPP V738, articulated skeleton of an adult individual; skull virtually complete, lacking only the palpebrals

nals, pubes, ischia, and most of the phalanges of the right pes (Young 1958: figs. 50–52, pls. 4, 5).

and the distal tip of the left jugal horn; postcranium lacking

Notable Referred Specimens. IVPP V739 (Young 1958: fig. 54),

only the distal radius and ulna, carpus, manus, distal tarsals

partial articulated adult skeleton including the posterior dor-

on both sides, and the distal three phalanges of digit III of

sal vertebrae, sacrum, anterior caudal vertebrae, ribs, both

the right pes; parts of the skeleton obscured by matrix include

ilia, proximal right ischium, femora, tibiae and fibulae; IVPP

the occiput, vertebral centra, cervical ribs, coracoids, ster-

V740-1 (Young 1958: fig. 53), partial articulated adult skeleton

36 sereno

FIGURE 2.10.

Psittacosaurus sinensis (BNHM BPV149) skull in (A) right lateral view; (B) posterior view. See text for abbreviations. Grey tone indicates matrix; cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bar is 2 cm.

including 15 articulated vertebrae (10 dorsal and 5 sacral

mation; Doushan village, 5 km northwest of Laiyang, Shan-

vertebrae), left and fragmentary right scapulocoracoids, dis-

dong Province (Fig. 2.1, localities 15, 16).

tal left humerus, both ilia lacking the postacetabular pro-

Diagnosis. Small-bodied psittacosaur with cranial autapo-

cesses, proximal ischia, both pubes lacking only the distal por-

morphies including (1) pendant rostrum that positions the

tion of the prepubic process, right femur and right proximal

ventral edge of the rostral bone below the level of the maxill-

fibula, and complete left hind limb lacking only the dis-

ary tooth row, (2) anteroventral processes of the nasal sepa-

tal phalanges; IVPP V742, fragmentary skull with maxillary

rated in the midline by a narrow gap, (3) short lower jaw that

teeth; IVPP V743, maxilla with four teeth; several isolated

positions the anterior margin of the predentary in opposition

teeth possibly representing more than one individual; IVPP

to the premaxilla rather than the rostral, (4) posteriorly flaring

V743a, fragmentary teeth and postcrania; IVPP V744 (Young

skull roof with postorbital-squamosal bars diverging at an an-

1958: fig. 55), anterior portion of a skull and fragmentary

gle of approximately 30\, (5) absence of the maxillary fossa,

postcrania; IVPP V745, maxilla with four teeth and one iso-

(6) absence of the maxillary protuberance, (7) vertically elon-

lated tooth; IVPP V749 (Young 1958: figs. 56, 57), skeleton

gate horn on the postorbital bar split between jugal and post-

with disarticulated skull elements including both maxillae

orbital, (8) frontal participation in the supratemporal fossa,

and nasals, left jugal and exoccipital, supraoccipital, preden-

(9) ectopterygoid far removed from postpalatine foramen

tary, and articulated postcrania including 12 presacral ver-

by broad maxilla-pterygoid contact, (10) internal mandibular

tebrae, right atlantal rib, posterior cervical and dorsal ribs,

fenestra reduced to a foramen, (11) absence of ossified ten-

left sternal, both scapulocoracoids, right humerus, left ulna,

dons, (12) prepubic and postpubic processes transversely

and partial right manus; associated disarticulated postcrania

broad throughout their length (transversely wider than dorso-

including the sacrum, fragmentary right ilium, both pubes,

ventrally tall), and (13) prepubic process projecting anteriorly

right humerus, right radius, right femur, right fibula, and par-

as far as the preacetabular process of the ilium.

tial pes; IVPP V750, series of dorsal ribs; IVPP V752, maxilla

Remarks. The small-bodied Psittacosaurus sinensis, based on

with five teeth, partial sacrum, and two distal femora; IVPP

an articulated skeleton (Young 1958: fig. 52), is known from

V753, internal skull mold; BNHM BPV149 (type of Psitta-

considerable skeletal material. Previous diagnoses for this spe-

cosaurus youngi Zhao 1962; Fig. 2.10), complete skull with frag-

cies originally were short and sometimes inaccurate. Gastralia

mentary postcranium including the vertebral column from

are not present in P. sinensis (contra Steel 1969: 39), and it

the posterior cervical to the anterior caudal vertebrae, frag-

is not plausible that P. sinensis be reduced to a junior synonym

mentary dorsal ribs, left ilium, proximal left ischium, and

of P. mongoliensis (Rozhdestvensky 1955; Coombs 1982). The

a fragment of the left pubis; CAGS-IG V808, two subadult

pendant form of the rostral in the holotypic skull is less appar-

skulls with articulated lower jaws; CAGS-IG unnumbered, ex-

ent because the ventral tip of the rostral is broken away (Fig.

cellent adult skull with articulated lower jaws and cervical

2.8). When reconstructed, it would extend below the level of

vertebrae; CAGS-IG unnumbered, fragmentary skull with

the tips of the maxillary crowns, as in a referred skull (Fig.

lower jaws.

2.10). P. sinensis shares several features with one or more addi-

Type Horizon and Locality. Qingshan Group, Doushan For-

tional species, including preorbital skull length of approxi-

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 37

mately 30% of skull length, the absence of an external man-

The holotypic specimen of P. xinjiangensis appears to be a

dibular fenestra or mandibular flange, and a short prepubic

subadult individual. A referred adult individual form Urho (=

process.

Wuerho) in the Junggar Basin not far from the holotypic locality (Brinkman et al. 2001) has a femur length of 153 mm, or

Psittacosaurus xinjiangensis Sereno and Zhao 1988

approximately one-third greater than that of the holotype

Holotype. IVPP V7698, posterior portion of the skull includ-

(105 mm). Thus, some caution is in order regarding features

ing the jugal horn and palpebral and articulated skeleton lack-

observed in the holotype that may owe their lack of expres-

ing distal limb bones and tail (Sereno and Zhao 1988: figs. 2, 3,

sion to immaturity, such as the absence of a quadratojugal

4F, 6).

eminence. The elongate postacetabular process of the ilium

Notable Referred Specimens. IVPP V7702, jaw fragments of

appears to be diagnostic, a process that has a height of at least

unknown association including right and left maxillae pre-

30% or more of its length in P. mongoliensis (Osborn 1924),

serving five teeth and an anterior dentary fragment (Sereno

P. major (Lü et al. 2007), P. sinensis (Young 1958), and P. sibiri-

and Zhao 1988: fig. 5A); IVPP V7704, complete right maxilla

cus (Averianov et al. 2006). The more slender curved, or hook-

with eight alveoli and left dentary fragment with erupting

like, shape of the prefrontal seems to be a good character,

crown (Sereno and Zhao 1988: figs. 4A, B, 5C, D); IVPP field

because the shape of this bone is broader in immature speci-

number 64047, fragmentary bones of several individuals in-

mens of P. mongoliensis. A similar hooklike shape, however,

cluding portions of the jugal, basioccipital and basisphenoid

may occur in juvenile specimens referred to P. lujiatunensis

and an articulated series of mid-caudal vertebrae with ossi-

(P. Makovicky pers. com.). The holotype of P. xinjiangenesis,

fied tendons (Sereno and Zhao 1988: figs. 4C–E, 8F); UGM

nevertheless, has reached adult size, as shown by a second

XG94Kh201, skull fragments and partial articulated skeleton

specimen of similar size with fused sacral vertebrae (Brinkman

lacking the tail (Brinkman et al. 2001: figs. 3–7).

et al. 2001).

Type Horizon and Locality. Tugulu Group (Aptian), Delun-

The revised diagnosis excludes two characters previously

shan, Junggar Basin, Xinjiang Uygur Autonomous Region,

listed by Sereno and Zhao (1988), the anteriorly flattened ju-

People’s Republic of China (Fig. 2.1, locality 2).

gal horn and the curved distal end of the denticulate margin

Revised Diagnosis. Psittacosaur characterized by autapo-

on the maxillary crowns, characters that appear elsewhere.

morphies including (1) hook-shaped palpebral with V-shaped

The jugal horn, in particular, has a sharp edged anterior face in

posterior margin, (2) dentary teeth with as many 21 den-

fully mature specimens of P. mongoliensis, P. lujiatunensis,

ticles in the posterior center of the tooth row, (3) ossified

P. major, and probably other species as well. The diagnosis, on

tendons extending into mid-caudal vertebrae, and (4) narrow

the other hand, includes the unusual shape of the palpebral in

iliac postacetabular process with height at midlength less

the holotypic specimen, which has an unusually concave, or

than 25% of the length of the process (as measured from the

V-shaped, posterior margin (Sereno and Zhao 1988: fig. 4F).

posterior edge of the ischial peduncle to the distal tip of the

The palpebral varies considerably in shape among species but

process).

does not appear to undergo marked shape transformation

Although relative preorbital skull length is not known,

during growth. Averianov et al. (2006) noted a size differential

other features with limited distribution among psittacosaur

in the palpebral of P. sibiricus but did not note any shape dif-

species are discernable from preserved portions of the skull,

ferential. The occipital condyle in P. xinjiangensis appears

including the absence of a quadratojugal eminence and den-

small in both the holotype and a referred specimen, its trans-

tary flange. Maxillary and dentary tooth rows appear to be

verse width less than one-half the width of the basal tubera

limited to eight teeth.

(Sereno and Zhao 1988; Brinkman et al. 2001). In other species

Remarks. The denticle count of 21 in a dentary tooth of

the condyle is usually more than one-half the width of the

P. xinjiangensis is high compared to the figured crowns of other

basal tubera, although this character may be influenced by

psittacosaurs. Caution is warranted in assessing this feature, as

exactly how the condyle is oriented and circumscribed.

denticle count increases with maturity and varies according to tooth position. The highest denticle counts occur in crowns of mature individuals from the posterior center of the tooth rows,

NOMINA DUBIA

where they tend to reach maximum size. The ossified tendons

Psittacosaurus mazongshanensis. P. mazongshanensis is based on

are preserved in an isolated series of mid-caudal vertebrae.

a fragmentary specimen preserving the lower jaws and por-

The absence of ossified tendons in the mid- and distal tail in

tions of the maxilla (Xu 1997: fig. 3; IVPP V12165). The skull

other psittacosaurs is based on well-preserved specimens of

is preserved upside down, the dorsal skull roof eroded away,

P. mongoliensis (Sereno 1987), cf. P. lujiatunensis (Mayr et al.

crushed, or covered by the lower jaws. The ventral margins of

2002), and P. sinensis (Young 1958).

the dentaries are broken, eliminating the chance to determine

38 sereno

the presence or strength of a dentary flange. Xu (1997) em-

dentary about 50 mm in length (TF 2449a), to which was re-

phasized the strong development of the maxillary protuber-

ferred another jaw fragment possibly pertaining to a maxilla

ance, the high number of denticles, the length of the sec-

(presumably TF 2449b; Buffetaut and Suteethorn 2002). Ini-

ondary ridges in the dentary crowns, and the Y-shape and

tially referred to Psittacosaurus without specific attribution

potential length of the snout. An exact denticle count, how-

(Buffetaut et al. 1989), the dentary was later made the holo-

ever, was never specified, and the figured crowns appear trun-

type of a new species, P. sattayaraki (Buffetaut and Suteethorn

cated by wear (Xu 1997: fig. 4). Denticle count varies along the

1992). Although these fragments add tantalizing information

tooth row in other psittacosaur species and is not sufficiently

to the still poorly known terrestrial Cretaceous faunas of

high in this case for distinction. Sereno (2000) questioned the

southeast Asia, assigning the holotypic dentary to Ceratopsia

validity of this species, as none of the features listed in the

incertae sedis based on its stout proportions may be the most

original diagnosis are unique. Averianov et al. (2006) reit-

specific assignment possible (Sereno 2000). Unsatisfied with

erated the ‘‘Y-shaped lower jaw’’ as a diagnostic feature, al-

that reassessment, Buffetaut and Suteethorn (2002) have con-

though this appears to be an artifact of preservation (Zhou

tinued to claim that P. sattayaraki is a valid species referable to

et al. 2006b). Zhou et al. (2006b: 112) regarded the pendant

the genus Psittacosaurus and that the upper jaw fragment col-

form of the maxillary protuberance to be a distinguishing fea-

lected later at the site is properly referred to the species.

ture, although this process is also pendent to some degree in

First, there is no basis for reference of the additional upper

P. mongoliensis, P. sp., and other species. Unfortunately, it is no

jaw fragment to the taxon, which was established on a den-

longer possible to investigate these potential distinguishing

tary. The additional jaw fragment is very poorly preserved and

features, as the maxillary protuberance and all of the teeth

figured (Buffetaut and Suteethorn 1992: fig. 2A, B). The au-

in the holotype have been lost (Zhou et al. 2006b). As none

thors are unsure whether it represents a right or left jaw frag-

of these features are sufficiently documented for comparison,

ment and leave open the possibility that some of its teeth were

Psittacosaurus mazongshanensis is here considered a nomen

blunted by wear. They claimed that a ‘‘loose replacement

dubium (Table 2.1).

tooth’’ from this specimen has a ‘‘strong primary ridge’’ (Buf-

Psittacosaurus ordosensis. The holotypic specimen of P. or-

fetaut and Suteethorn 1992: 805). Unfortunately, there is no

dosensis (IVPP 07-08888-1) includes the ventral one-half of a

documentation of this crown in their papers, its possible wear

skull and partial hind limb, although additional unprepared

facets, the described angle between crown and root, or any

specimens were noted to exist (Russell and Zhao 1996). The

other feature beyond its comparable size and site of origin

limited information available is insufficient to distinguish this

to justify their interpretation that this specimen might per-

material from that of P. sinensis (Sereno 2000). Outstanding

tain to the same taxon, much less the same individual, as the

features shared by both include small body size (skull length

holotype.

approximately 100 mm), pendant anterior end of the upper

Second, their description of the ‘‘bulbous’’ primary ridge

jaw, short lower jaw with predentary tip opposing the premax-

and ‘‘incipient’’ dentary flange in the holotype is poorly pre-

illa, a rudimentary maxillary fossa, and tooth rows limited to

served on the original specimen and appears to be enhanced

eight teeth. The tooth rows were described as straight in P. ordo-

in their figures. The specimen was acid-etched, rather than

sensis by Russell and Zhao (1996) rather than concave as re-

mechanically prepared, and basal portions of the two most

ported in P. sinensis (Sereno 1990b). In P. sinensis, however, the

complete replacement crowns appear to have dissolved away.

curvature is often very subtle (Young 1958: fig. 57). Further-

The bulbous form of the primary ridge is based on two partial

more, it is clear that the tooth rows in all psittacosaurs are later-

crowns in the third and fifth alveoli. The more anterior of

ally concave, such as those in P. mongoliensis and the recently

the two preserves only the apical margin and cannot establish

described skull of P. sibiricus (Averianov et al. 2006: fig. 10A).

the form of the primary ridge. Two drawings of this crown

Xu and Zhao (1999: 80) added several other cranial features

edge have been published (Buffetaut and Suteethorn 1992:

to the diagnosis of P. ordosensis such as an ‘‘eminence on cau-

fig. 3B; Buffetaut and Suteethorn 2002: fig. 1b), neither of

dal frontal’’ and ‘‘posterodorsal corner of the skull depressed.’’

which appears to precisely match the available photograph of

The dorsal aspect of the skull, however, is not preserved in the

the crown (Buffetaut and Suteethorn 1992: fig. 2C). The latest

holotype specimen (Russell and Zhao 1996: fig. 5), and no

interpretation depicts the apical denticle above a widening

other specimens are cited or figured. Additional study of the

primary ridge (Buffetaut and Suteethorn 1992: fig. 3B; b), but

holotype and referred material may eventually show that di-

this drawing appears to show more of the crown than is ac-

agnostic features do exist, or, at least, a diagnostic combina-

tually preserved. The upper one-half of the more posterior

tion of features. For the time being, P. ordosensis is regarded as a

crown is preserved and has also been depicted in two ways.

nomen dubium.

The first drawing shows secondary ridges extending far-

Psittacosaurus sattayaraki. P. sattayaraki is based on a right

ther down the crown than the primary ridge (Buffetaut and

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 39

Suteethorn 1992: fig. 3B). The most recent interpretation shows an asymmetrical, primary ridge that broadens toward the middle of the crown, although much less than is typical in psittacosaurs, in which the ridge is very prominent and oc-

Cranial Morphology SKULL ORIENTATION AND PREORBITAL LENGTH

cupies the middle section of the crown. The broadening of

The shape and measurement of the skull in psittacosaurs are

this ridge, in addition, cannot be verified in the only available

very sensitive to the orientation of the cranium. If the dorsal

photograph (Buffetaut and Suteethorn 1992: fig. 3B). The pri-

skull roof is positioned so that it is horizontal, for example,

mary ridge may expand basally to some degree, although this

the anterior margin of the snout assumes a nearly vertical

should be documented with a properly lit, enlarged image or

orientation and appears particularly short in dorsal view.

stereophotograph.

In orthogonal views in this paper, the cranium is oriented

The ‘‘incipient’’ dentary flange, likewise, is not apparent

with the maxillary tooth row positioned along a horizontal.

in the only available photograph (Buffetaut and Suteethorn

Similarly, the lower jaw in disarticulation is oriented with the

1992: fig. 2E). The ventral contour of the dentary does not

dentary tooth row positioned along a horizontal. With the

match their drawing, which depicts a discrete ridge toward

cranium in this registration, the postorbital portion of the

the rear of the dentary ramus (Buffetaut and Suteethorn 1992:

skull roof in all species is tilted slightly posteroventrally in

fig. 3C). The psittacosaur dentary flange, furthermore, is not

lateral view, and the tapering anterior process of the nasal is

developed in this way, as a posteriorly located ridge that in-

exposed in dorsal view. From this horizontal axis, perpendicu-

creases in strength posteriorly. The flange is deepest anteriorly

lar lines are established to the anterior margin of the rostral

under the anterior end of the tooth row. In species with a

and the anterior rim of the orbit. The distance between these

reduced flange, such as P. sp., the flange is located more ante-

perpendiculars is a measure of preorbital length, which is com-

riorly and is swollen and strongest anteriorly.

pared to skull length, as measured from the anterior margin

Third, the features listed in the diagnosis of P. sattayaraki

of the snout to the posterior margin of the quadrate. Using

(incipient dentary flange, strongly convex alveolar edge, five

this posterior landmark for skull length reduces variation

denticles to each side of the apical denticle) are not unique

from crushing and breakage that often occurs along the pos-

and also vary during growth in psittacosaurs. Nothing in the

terior margin and corners of the skull. When measured consis-

additional defense of this species (Buffetaut and Suteethorn

tently in this manner, all psittacosaurs have a relatively short

2002) alters the fact that these poorly documented features are

preorbital skull segment that is less than 40% of skull length.

not unique. Post-hatching growth witnesses an increase in tooth and denticle counts, and the medial convexity of the tooth row in many psittacosaurs is often stronger than that shown in the Thai dentary.

ADULT SKULL SHAPE AND FORM The following is a brief summary of adult skull shape in psitta-

Fourth, the broad dorsoventral proportion of the anterior

cosaurs highlighting the most unusual and variable features.

end of Meckel’s canal and near vertical dentary symphysis

Various terms have been used to describe osteological features

are not present in any other psittacosaur species and do not

that are present primarily in psittacosaurs. A specialized termi-

closely resemble the condition in hatchlings (contra Buffetaut

nology, introduced in italics below, is developed to encourage

and Suteethorn 2002: 73). Although the specimen could have

terminological uniformity. The description is based on first-

been diagnosed on these grounds, its apparent immaturity

hand examination of P. sp., P. major, P. meileyingensis, P. mon-

and uncertain generic affinity mitigate against erecting a new

goliensis, P. neimongoliensis, P. sinensis and P. xinjiangensis (Se-

taxon. Xu and Zhao (1999) listed P. sattayaraki as a junior

reno 1987; Sereno and Zhao 1988; Sereno et al. 1988) as well as

synonym of P. mongoliensis without comment, although there

reference to descriptive accounts of P. lujiatunensis (Zhou et al.

appears to be no justification for such an attribution. Aver-

2006b), P. neimongoliensis (Russell and Zhao 1996), and P. sibi-

ianov et al. (2006) supported recognition as a new species of

ricus (Averianov et al. 2006).

Psittacosaurus but did not offer any unique features. Con-

General Skull Shape. The psittacosaur cranium, as its name

trary to Buffetaut and Suteethorn (1992) and Averianov et al.

suggests, resembles that in parrots in several regards (Zusi

(2006), the tooth row in P. sattayaraki is not straight but rather

1993). The bill-sheathed snout is short, deep, and narrow and

is gently medially convex in dorsal view as in all psittacosaurs

constructed primarily of the expansive posterolateral process

(see Buffetaut and Suteethorn 1992: fig. 2D). The basis for re-

of the premaxilla. The preorbital segment of the psittacosaur

ferral to Psittacosaurus rests entirely on the bulbous form of

skull is less than 40% of skull length, shorter than in nearly

the primary ridge in the dentary teeth, which is not well estab-

all other ornithischians. Preorbital length in P. mongoliensis is

lished. Ceratopsia, incertae sedis, may be the best tentative

approximately 37% of skull length, or slightly greater than

assignment for this fragmentary, albeit interesting, dentary

that in other species (30–35% of skull length). Because pre-

(Table 2.1).

orbital skull length increases somewhat during growth, spe-

40 sereno

cies comparisons are most useful when using the skulls of ma-

point on the posterior snout, the exact pattern at their junc-

ture individuals.

tion showing some variation. The premaxilla joins the maxilla

A proportionately deep median rostral bone caps the snout

along a rugose, akinetic suture that protrudes laterally as an

anteriorly. The snout sidewall lacks any sizeable openings,

anteroventrally sloping premaxilla-maxilla ridge (Fig. 2.2A).

given the high position of the external naris and absence of an

The attachment surface for the keratinous upper bill clearly

antorbital fenestra. The nasal, unlike that in any other dino-

extended beyond the rostral bone and onto the premaxilla.

saur, extends below the external naris, its anteroventral tips

The relatively small, subtriangular maxilla has a distinctive

resting against the dorsal end of the rostral bone. The inter-

subtriangular depression above the buccal emargination here

narial bar, as a consequence, is constructed from the nasal

termed the maxillary fossa (Figs. 2.2A, 2.3A, 2.7A; ‘‘secondary

alone, rather than a composite strut formed by processes of

depression,’’ Sereno 1990b: 583; ‘‘antorbital fossa,’’ Zhou et al.

the nasal and premaxilla. The sutureless internarial bar and

2006b: 105). The maxillary fossa is not a pneumatic depres-

rostral-nasal contact in psittacosaurs further strengthens the

sion and has no relation to the antorbital fossa (Sereno 1990b,

anterior margin of the snout, which is loaded by compressive

2000). In hatchling psittacosaurs, the maxillary fossa is pres-

forces from the upper bill.

ent (see below), although at no point is there any trace of

As in other ceratopsians, the skull roof is subtriangular in

an antorbital opening or any other connection between the

dorsal view, reaching its greatest breadth across the jugals just

maxillary fossa and the nasal cavity, a necessity were it to be

posterior to the orbit. In all species of Psittacosaurus, the ven-

considered a pneumatic structure. Of several neurovascular

tral margin of the skull narrows in width from the jugal horn

foramina that open within, or on the rim of, the maxillary

to the quadrate condyle (also Chaoyangsaurus; Zhao et al.

fossa, the largest is located near, or along the suture with the

1999). The skull in other neoceratopsians, in contrast, does

premaxilla. Here termed the anterolateral maxillary foramen,

not show similar narrowing along the ventral margin (e.g.,

it opens anterolaterally into an impressed vessel tract that

Yinlong, Xu et al. 2006; Liaoceratops, Xu et al. 2002). As in

passes anteroventrally to the edentulous margin of the pre-

all ceratopsians, the frontoparietal roof is relatively narrow,

maxilla (Fig. 2.2A). The ventral margin of the maxillary fossa

broadly exposing both the orbit and laterotemporal fenestra

forms the dorsal edge of the buccal, or cheek, emargination. A

in dorsal view.

characteristic process, here termed the maxillary protuberance

Dorsal Skull Roof. The psittacosaur rostral is broadly arched

(‘‘maxillary boss,’’ Sereno, 1987: 76; ‘‘maxillary process,’’ Se-

transversely and is relatively thin, despite a wedge-shaped ap-

reno et al. 1988; ‘‘protuberance,’’ Zhou et al. 2006b: 105) is

pearance in lateral view. Ventrally it occludes with the broadly

present on the posterior end of the rim of the cheek emargina-

arched edge of the predentary and lacks the derived anterior

tion near the maxilla-jugal suture. The maxillary fossa and

keel and recurved ventral tip that are present in neoceratop-

protuberance are poorly developed in some species (P. neimon-

sians. The palatal surface of the rostral is crescent-shaped and

goliensis, P. sinensis, P. sibiricus).

angles anteroventrally at approximately 60\ from the hori-

The buccal emargination, or cheek, is deeply inset. The lat-

zontal plane of the skull. A rounded, U-shaped attachment

eral wall of the cheek is nearly flat and inclined at approxi-

ridge occupies the middle one-third of this surface and extends

mately 45\ medioventrally. The opposing surface within the

between the palatal processes of the premaxillae (Fig. 2.5).

palate that borders the internal nares is nearly vertical as seen

This ridge, which is well exposed in P. mongoliensis, P. major

in palatal view (Fig. 2.4A). As a result, there is a dorsally thick-

and P. sp., presumably served to anchor the upper bill, rein-

ening wedge of bone above the maxillary tooth row, here re-

forcing the central portion of the rostral.

ferred to as the alveolar pedestal, which thins abruptly to a

The premaxilla is edentulous, expansive, and polygonal in

vertical plate in the region of the maxillary fossa.

shape, forming most of the vertical wall of the tall psitta-

One of the hallmarks of the skull of Psittacosaurus is the

ciform rostrum. The majority of the bone is composed of a

pyramidal jugal horn, which sometimes shows impressed vas-

hypertrophied posterolateral process, which in most ornithis-

cular grooves indicative of a keratinous sheath. Although the

chians is a small triangular process tapering between the max-

jugal horns in adult P. sinensis appear to be the most laterally

illa and nasal. The posteromedial or internarial process, in

prominent, the size and lateral prominence of the horns ap-

contrast, is reduced to a small spur that forms the antero-

pear to increase during growth in many species and are easily

ventral corner of the external naris. The premaxillary foramen

altered by postmortem distortion. Distinguishing species on

pierces the lateral wall near the anterior end of the premaxilla

the basis of jugal horn size, form, and orientation is fraught

well below the small external naris. The lateral wall of the

with uncertainty.

lacrimal canal is incompletely ossified in Psittacosaurus, leav-

The central body of the postorbital and the midpoint of the

ing an oval fenestra of the lacrimal canal between the premax-

postorbital bar are also raised as keratin-covered protuber-

illa and lacrimal, a condition unique to the genus. Premaxill-

ances, although neither reaching the size of the jugal horn. In

ary, maxillary, lacrimal, and jugal sutures converge toward a

most species, the protuberance on the central body of the

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 41

postorbital is developed only as a ridge, here termed the post-

rior extremity of the buccal emargination. Posteriorly the pre-

orbital crest, which extends posteriorly along the lateral side of

maxilla is rigidly sutured to the anteromedial process of the

the posterior process (Fig. 2.2A). In P. sibiricus, however, the

maxilla, which meets in the midline and excludes the premax-

portion of the protuberance on the central body is raised as a

illa from the footplate of the vomera and border of the inter-

blunt, dorsolaterally projecting postorbital horn (Averianov et

nal nares. The vomera are strongly arched dorsally in lateral

al. 2006). The second crest or horn, here termed the postorbital-

view such that the ventral edges are located just below the

jugal crest or horn, occurs more ventrally on the postorbital

ventral orbital margin as in Protoceratops (Sereno 2000).

bar. In P. mongoliensis and P. major it is developed only as the

Although it appears that psittacosaurs lack an interptery-

everted posterior edge of the ventral process of the postorbital.

goid vacuity, a gap is present separating most of the original,

In other species such as the new species from Inner Mongolia

or primary, palatal rami. The primary palatal ramus is exposed

and P. sinensis, it is raised as a smooth pyramidal horn mainly

only in dorsal view within the orbit, where the distal margins

on the postorbital or a more rounded horn with rugose texture

of the steeply angled palatal rami join in the midline, pinch-

split between the postorbital and jugal (Figs. 2.2B, 2.9).

ing the vomera anteriorly and the parasphenoid posteriorly. A

Finally, there is a quadratojugal protuberance on the central

pterygoid vacuity separates the posterior portions of the pri-

body of the quadratojugal that is best developed as a discrete

mary palatal rami but is obscured in ventral view by a second-

rugosity in P. meileyingensis. In most other species, this area is

ary, or neomorphic, lamina of the pterygoid. This secondary

raised only as a smooth eminence (Fig. 2.2A).

lamina contacts its opposite in the midline to form what has

Variably developed depressions, or fossae, besides the maxil-

been termed the basal plate, the median suture of which in

lary fossa are located near the orbit and include the postorbital-

adults is fused and raised into a low median crest (Sereno

jugal fossa just below the postorbital-jugal horn (Fig. 2.2B) and

1987). In ventral view of the palate, the vomera and the sec-

the jugal fossa located on the jugal below the orbital margin

ondary palatal laminae form the vaulted posterior palate,

(Fig. 2.3A).

which slopes posteroventrally at about 45\. The basal plate

A parietal-squamosal shelf projects posteriorly over the occi-

forms a secondary palatal surface ventral to the primary pala-

put as a transversely broad, anteroposteriorly narrow, hori-

tal rami, between which is a blind pocket of pyramidal shape.

zontal sheet. Often referred to as a ‘‘frill,’’ the shelf neither

Termed the pterygoid fossa, this pocket opens posteriorly be-

projects posterodorsally nor varies significantly in anteropos-

tween the basipterygoid articulations (Fig. 2.4B). Only psitta-

terior length, as occurs in neoceratopsians presumably in re-

cosaurs have this reinforced, akinetic palatal structure, which

sponse to a display function. Smooth depressions on the ven-

is designed to withstand unusually high bite force.

tral aspect of the shelf, here termed the occipital fossae, are

Braincase. The subrectangular occipital surface is divided in

present in all species to either side of the prominent keel-

the midline by a vertical keel-shaped nuchal crest, the ex-

shaped nuchal crest (Fig. 2.5). These fossae may have func-

panded dorsal end of which approaches but does not contact

tioned primarily as an expanded area of attachment for de-

the overlying parietal-squamosal shelf (Fig. 2.5). The paroc-

pressor mandibulae musculature.

cipital process extends laterally, expanding gradually to a sub-

Palate. The secondary palate, the majority of which is

quadrate distal end. There is no development of a ventral

formed by the premaxilla in all species (contra Zhou et al.

hook-shaped process for attachment of jaw-opening muscula-

2006b: fig. 4), is subcircular in ventral view and arches high

ture as in ornithopods. Rather that anchoring role may have

above the bony edge of the upper bill. In the midline, a large

been played by the parietosquamosal shelf, as indicated by the

slit-shaped incisive foramen opens onto the premaxillary pal-

occipital fossae on its ventral surface. These fossae, which are

ate from the nasal cavity (Fig. 2.4A; ‘‘median notch,’’ Sereno

invariably present in psittacosaurs and easily observed in the

1987: 75; ‘‘interpremaxillary foramen,’’ You et al. 2008: 185).

largest species (e.g., P. major), extend ventrally onto the par-

To each side of this foramen, the anteroventral edge of the

occipital processes (Fig. 2.5). The occipital condyle, basal tu-

premaxilla projects from the palatal surface to form what ap-

bera and basipterygoid processes lie roughly in the same hori-

pears to be a narrow triturating surface inset within the upper

zontal plane, the latter processes noticeably longer in P. major

bill for the lower bill (Fig. 2.4A). This edge is well preserved in

(Fig. 2.5).

P. sp., P. major, and P. mongoliensis (Fig. 2.4A, ts). Two large

The dorsal skull roof preserves shallow fossae for the fore-

foramina, here termed the medial and lateral premaxillary fora-

brain as well as two smaller oval fossae for the olfactory bulbs

mina (‘‘neurovascular canal,’’ You et al. 2008: 187), probably

(Fig. 2.11B). More anteriorly, there is a pair of shallow sulci on

provided the primary vascular support for the internal aspect

the ventral aspect of the nasals, which are clearly positioned

of the upper bill (Fig. 2.4A). They open anteriorly from the

anterior to the endocranial volume (Fig. 2.11B). The lateral

anterior end of the maxilla, the medial foramen is slit-shaped

nasal sulcus was misinterpreted recently as an enlarged fossa

and located medial to the bill margin and the lateral foramen

for an olfactory bulb (Zhou et al. 2007: fig. 2B).

is subcircular and located lateral to the bill margin at the ante-

42 sereno

Accessory Dermal Elements. The palpebrals in psittacosaurs

FIGURE 2.11. Psittacosaurus mongoliensis (PI 3779/20) skull table in (A) dorsal view; (B) ventral view. Cross-hatching indicates broken bone; dashed line indicates estimated edge. See text for abbreviations. Scale bar is 3 cm.

and other ceratopsians are subtriangular rather than elongate

Lower Jaw. The lower jaw fuses at the mandibular symphy-

as in most other ornithischians. In all species of Psittacosaurus,

sis with maturity, uniting the dentaries (including their tooth

in addition, the palpebrals project laterally from the antero-

rows) and the predentary as an akinetic unit. A pendant den-

dorsal corner of the orbit, extending laterally to a width equal

tary flange (‘‘ventral flange,’’ Sereno 1987: 134; ‘‘dentary

to that of the posterior skull table (Fig. 2.12). In basal neo-

flange,’’ Zhou et al. 2006b: 110; ‘‘ventrolateral flange,’’ You et

ceratopsians, in contrast, the palpebrals curve posteriorly.

al. 2008: 192) projects ventrolaterally from the dentary ramus

A sclerotic ring is composed of a series of thin, oval overlap-

in P. lujiatunensis, P. meileyingensis, P. major, and P. sibiricus. The

ping plates that compose a ring of relatively modest size in

dentary flange is low in P. mongoliensis and represented only by

Psittacosaurus (Fig. 2.13). The circular sclerotic ring is rela-

rugosities in some other species. Both coronoid and intercoro-

tively small, the maximum outside diameter (18 mm) about

noid ossifications are present, the latter a delicate strap-shaped

one-half of the maximum diameter of the orbit (38 mm). Posi-

bone currently preserved only in P. mongoliensis (Fig. 2.14).

tive plates are located at the top and bottom of the ring, but

The jaw joint, which is similar in all species, allowed rotary as

only one negative plate is present on the anterior side of the

well as anteroposterior movement. Rounded, poorly divided

ring. Given the circular geometry of the sclerotic ring, the

quadrate condyles articulate against a broad, flat articular plat-

number of positive plates must equal the number of negative

form composed mostly of the articular (Fig. 2.14). A robust

plates. Another negative plate, therefore, must have been

retroarticular process extends posterior to the jaw articulation,

present on the damaged posterior side of the ring. Thus there

unlike the very short process in neoceratopsians.

are at least 15 plates in the sclerotic ring of P. mongoliensis, 14

Stapes and Hyoid. The slender columelliform stapes is pre-

of which are preserved, and at least 1 missing negative plate

served in P. mongoliensis, P. lujiatunensis (Zhou et al. 2006b),

(Sereno 1987).

and P. major (LH PV1). In the last individual the stapes is pre-

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 43

FIGURE 2.12. Psittacosaurus mongoliensis (IVPP V7668) left palpebral in (A) dorsal view; (B) ventral view. Grey tone indicates matrix. Scale bar is 1 cm.

FIGURE 2.13.

Psittacosaurus mongoliensis (AMNH 6254) right sclerotic ring in (A) lateral view; reconstruction of right sclerotic ring in (B) lateral view. See text for abbreviations. Dashed line indicates estimated edge. Scale bar is 5 mm.

served in natural articulation, the slender columnar shaft ex-

length and diverge posteriorly, the posteriormost tooth in

tending toward the otic notch between the quadrate head and

each tooth row separated from its opposite in the midline by

paroccipital process (Fig. 2.15). The slightly expanded distal

approximately three times the distance of the anteriormost

section of the ossified shaft nearly reaches the presumed posi-

pair (Figs. 2.4, 2.9B).

tion of the tympanum.

Enamel is always present on both sides of the crown (contra

A pair of rods identified as ceratohyals are preserved in

You et al. 2008: 194), although it is thicker on the lateral and

P. mongoliensis, P. sibiricus and P. sinensis (Colbert 1945, Sereno

medial aspects of maxillary and dentary crowns, respectively

1987, Averianov et al. 2006). Approximately one-third as

(Fig. 2.6). The most characteristic feature of the dentition is

long as the skull, two specimens preserve the ceratohyal be-

the bulbous primary ridge on the dentary crowns (Fig. 2.6C).

tween the mandibular rami, demonstrating that the flattened

Tooth-to-tooth wear facets, which are evident shortly after

end of the shaft is positioned anteriorly and the cylindrical

hatching, truncate upper and lower crowns at approximately

end posteriorly (P. mongoliensis PI 3779/17; P. sinensis IVPP

20\ from the vertical. There are clear traces of crown-to-

V738). The anterior end is transversely compressed and the

crown wear in several species that indicate that the power

shaft curves posterolaterally toward the mandibular rami, a

stroke is nearly straight and draws the mandibles posterodor-

configuration similar to that in other ornithischians such

sally at 30\ from the horizontal (Fig. 2.16). Tooth wear in psit-

as Edmontosaurus (Versluys 1923) and Corythosaurus (Ostrom

tacosaurs thus is neither orthal nor propalinal (contra Nor-

1961).

man and Weishampel 1991: fig. 11). Rather psittacosaurs

Dentition. The psittacosaur dentition is limited to a dozen or

employed a unique masticatory cycle involving posteriorly

less teeth in the maxillary and dentary that remain struc-

divergent, akinetic upper and lower jaws and an isognathus

turally simple, with one functional and one replacement

posterodorsal power stoke. The resulting high-angle, tooth-

tooth per alveolus. Tooth number increases with age, ranging

to-tooth wear facets resemble those in other ornithischians

in adults between 10 and 12 for large species and 9 and 10 for

that have kinetic upper and lower tooth rows and an orthal

smaller species. The tooth rows of all species are equal in

power stroke.

44 sereno

FIGURE 2.14. Psittacosaurus mongoliensis (AMNH 6534) right lower jaw in medial view. See text for abbreviations. Cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bar is 3 cm.

smallest individual (AMNH 6535; Figs. 2.17–2.19) was identified as the partial skeleton of a ‘‘tiny dinosaur’’ by Walter Granger in his field notes in 1922 and later described as a psittacosaur in passing by Andrews (1932: 223). The abundance and similar size of much of the postcranial material suggests that a nest of hatchlings may have been inadvertently collected, which would have been the first to have been encountered by paleontologists. The juvenile material was first studied by Coombs (1980, 1982), who described the two smallest skulls (AMNH 6535, 6536) as well as postcranial material assigned arbitrarily to the larger specimen (AMNH 6536) (Figs. 2.17–2.22). By that time, Psittacosaurus mongoliensis (AMNH 6254) right stapes in (A) presumptive lateral view; (B) presumptive ventral view with three cross sections of the stapedial shaft at four times magnification. Cross-hatching indicates bone section. Scale bar is 1 cm. FIGURE 2.15.

the postcranial bones belonging to the smallest skull (AMNH 6535) had become disassociated from the skull (Fig. 2.17). Subsequently this partial, articulated skeleton was located and reunited with the skull by the author (Sereno 1987). On the basis of the skull alone, Coombs (1980: 380, fig. 2) estimated its body length between 230 and 250 mm with a skull approxi-

NEONATE SKULL TRANSFORMATION

mately 50% of its trunk length. The postcranial skeleton preserves most of an articulated trunk and pertains to a hatchling

Material. In 1922 the Third Asiatic Expedition collected a

that was probably less than 200 mm in length. Its skull length

number of immature individuals along with the adult holo-

is approximately 70% of the estimated length of the trunk

typic skeleton of Psittacosaurus mongoliensis (Osborn 1923,

(Fig. 2.17A).

1924) from the Khukhtek Formation (= Oshih or Ondai Sair

Now fully prepared, the two smaller skulls are figured in

Formations, Tevsh svita) on the northern flank of Arts Bogd in

detail (Figs. 2.18–2.22). The smallest skull (AMNH 6535) has a

an east-west graben known as Oshih (locally as ‘‘Osh’’; Rougier

length of 27.5 mm, or about 20% of adult skull length. The

et al. 2001). The juvenile material includes three partial skulls

skull is preserved without distortion or flattening. The left

of different age (AMNH 6535, 6536, 6540), several partial skel-

stapes is preserved in two pieces on the left side of the brain-

etons as well as disarticulated postcrania (Sereno 1987). The

case (Fig. 2.20A). Most of the rostral, all of the predentary, and

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 45

FIGURE 2.16. Psittacosaurus mongoliensis (PI 698/1-3). (A) Posterior portion of the right dentary with posteriormost eight tooth positions in lateral view; (B) close-up of four worn teeth (shaded in A) showing mesowear in lateral view. Cross-hatching indicates broken bone; arrows show the angle of passage of individual teeth in the maxillary tooth row across the dentary tooth row. Scale bar is 1 cm.

both postorbitals and squamosals are eroded away. There are

1982). Rounded anterodorsal and posterodorsal skull margins

no palpebrals or sclerotic rings, although these are present in

become squarer as the skull table flattens. Preorbital skull

nestlings that are slightly larger (skull length approximately

length increases from approximately 30% of total skull length

35 mm; Meng et al. 2004) and so were probably lost to erosion.

in the smallest individual to 35% in the slightly larger skull,

The second smallest skull has a length of about 40 mm, or

with most adult skulls still slightly longer at about 37%. Thus

about 30% of adult skull length, and is smaller than juveniles

most of the lengthening of the preorbital portion of the skull

found recently in association (skull length approximately 57

occurs at very small size. The laterotemporal fenestra increases

mm; Qi et al. 2007). The skull is flattened exposing the dorsal

in anteroposterior length from about 45% of its height in

skull roof in right lateral view (Figs. 2.21B, 2.22B). The right

the smallest skull to about 70% in an adult. The parietal-

palpebral and left sclerotic ring are preserved (AMNH 6536;

squamosal shelf is not well preserved in the hatchling skulls,

Fig. 2.22). The palpebral is subtriangular rather than elongate

so how it changes with growth is uncertain.

as in many other ornithischians; Coombs (1982: fig. 3) mis-

Other changes include the absence or rudimentary form of

takenly interpreted a section of the pterygoid as the posteri-

what may be interpreted as secondary sexual characteristics,

orly elongate shaft of the palpebral.

which here include the various horns, rugose eminences and

Both skulls of P. mongoliensis are associated with the anterior-

upturned edges that appear to be ornamental rather than serv-

most cervical vertebrae and were presumably originally articu-

ing as sites for muscle attachment. The jugal horn and man-

lated with postcranial skeletons. The trunk of the smaller in-

dibular flange are rudimentary, and there is no development

dividual is preserved. The association of postcrania with the

of the maxillary protuberance or upturned lip of the prefron-

larger individual cannot be established, although the presence

tal. Other features that change with growth but also have been

of multiple individuals of comparable size suggests that nest-

cited as differences among species include relative preorbital

lings may have been collected.

skull length, relative size of the occipital condyle, relative

General Skull Shape. Skull form and sutural pattern in even

position of the posterior margin of the secondary palate, and

the smallest individual are remarkably similar to that of an

number of teeth in the maxillary and dentary tooth rows.

adult. General shape changes in the skull during growth in-

Demonstrating maturity is important, especially when these

clude many of those expected in any diapsid, such as reduc-

features are invoked as species differentia.

tion in the size of the orbit and braincase and development of

Dorsal Skull Roof. In both skulls the premaxilla has a large

a rim on the supratemporal fossa and sagittal crest (Coombs

foramen near its anterior margin and distinct depressions pos-

46 sereno

FIGURE 2.17. Psittacosaurus mongoliensis (AMNH 6535) hatchling skull and partial skeleton in (A) left lateral view; partial postcranial skeleton in (B) right lateral view. See text for abbreviations. Cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bars are (A) 2 cm; (B) 1 cm.

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 47

FIGURE 2.18.

Psittacosaurus mongoliensis (AMNH 6535) stereopairs of hatchling cranium in (A) right lateral view; (B) ventral view. Scale bar is 1 cm. Reproduced in color on the insert.

terior to the rostral and ventral to the prefrontal as in the adult

rate shaft arches posteriorly to its head, the maxillary and

(Figs. 2.19A, B, 2.22). The premaxilla-maxilla suture is ele-

dentary buccal emargination and coronoid process are well

vated as a ridge with a large vascular foramen at its antero-

developed, and a well-developed mandibular flange is present

ventral extremity. A maxillary fossa with foramina is located

in the smaller hatchling skull (Coombs 1982; Fig. 2.19F). On

posteroventral to the ridge. An antorbital fenestra, thus, is

the left side, the lacrimal canal is exposed as often occurs in

absent in the smallest individual, and the external fossa on the

adults. In dorsal view the orbit and laterotemporal fenestra are

maxilla must be a secondary feature unique to Psittacosaurus.

exposed lateral to the skull table. In ventral view, the maxil-

It is not a pneumatic depression, judging from the absence of

lary tooth rows are gently laterally concave to either side of a

any communication with the nasal passage and presence of

strongly arched palatal septum, and a subrectangular mandib-

foramina and surface texture in adults. As will be explored

ular flange projects ventrolaterally posterior to the maxillary

below, to the contrary, it appears to be an attachment fea-

tooth row. The quadrate condyle also can be seen to angle

ture related to the unusually expanded jaw musculature in

anteromedially rather than having a transverse orientation.

psittacosaurs.

The predentary-dentary suture is open in both skulls (contra

In both skulls the free ventral margin of the premaxilla sep-

Coombs 1982), fusing later in growth in adult mandibles.

arates the rostral and maxilla (contra Coombs 1982: 100,

Palate. The palate shows a number of changes during

fig. 7A, B), and the sutures of the premaxilla, maxilla, lacrimal,

growth. The premaxillary palate, which is well preserved and

and jugal converge on a point as in the adult. In AMNH 6540,

exposed on the smallest skull (Figs. 2.19F, 2.20B), becomes

the bony platform for the bill and the vascular foramina and

more deeply vaulted with maturity. The posterior edge of the

impressed channels supplying the keratinous bill are already

secondary palate, located anterior to the tooth rows in the

established as in the adult. The sutures on the sidewall of the

smallest skull, extends farther posteriorly to a position even

snout and vascular supply to the bill are similar to that of

with the tooth rows in adults. The unusual structure of the

an adult.

posterior palate, in which a neomorphic process of the ptery-

In lateral view of the smallest skull, the jaw articulation is

goid extends to the midline as a horizontal basal plate, is in

located below the level of the maxillary tooth row, the quad-

place in the smallest hatchling. The basal plate is preserved as a

48 sereno

Psittacosaurus mongoliensis (AMNH 6535). Hatchling cranium in (A) left lateral view; (B) right lateral view; (E) dorsal view; (F) ventral view. Left lower jaw in (C) lateral view; (D) medial view. See text for abbreviations. Grey tone indicates matrix; cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bar is 1 cm. FIGURE 2.19.

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 49

50 sereno

FIGURE 2.20.

Facing above. Psittacosaurus mongoliensis (AMNH 6535) hatchling cranium in (A) posterodorsolateral view showing palate and an enlarged view of the stapedial footplate; (B) posteroventrolateral view of braincase. See text for abbreviations. Grey tone indicates matrix; cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bar is 1 cm. FIGURE 2.21.

Facing below. Psittacosaurus mongoliensis (AMNH 6536) stereopairs of hatchling skull and anterior cervical vertebrae in (A) left lateral view; (B) right lateral view. Scale bar is 1 cm. Reproduced in color on the insert. FIGURE 2.22.

Psittacosaurus mongoliensis (AMNH 6536) hatchling skull and anterior cervical vertebrae in (A) left lateral view; (B) right lateral view. See text for abbreviations. Grey tone indicates matrix; cross-hatching indicates broken bone; dashed line indicates estimated edge. Scale bar is 1 cm.

horizontal shelf between the pendant, well-developed man-

vided posterior margin, the remnants of the bifid ventral pro-

dibular rami of the pterygoids, and it encloses a pterygoid

cess in neornithischians. The lateral process in the juvenile

fossa under the tent-shaped palatal rami of the pterygoids

is poorly developed, whereas in the adult it extends pos-

(Figs. 2.4B, 2.20B).

teriorly as a raised edge to join the lateral margin of the den-

Braincase. The braincase decreases in size with growth. In

tary buccal emargination. A mandibular flange is present on

both skulls, the basioccipital forms the majority of a small

the dentary of the larger hatchling skull and increases in

occipital condyle. Apparently unlike Protoceratops (Brown and

prominence, extending posteriorly as a ridge onto the angular

Schlaikjer 1940), however, the exoccipitals form the lateral

in the adult. The external mandibular fenestra decreases in

corners of the condyle even in the smallest skull (Fig. 2.20B).

size, and the surangular decreases in depth compared to the

The basal tubera are developed as parasagittal ridges that do

angular.

not extend toward the midline. The pedicel of the exoccipital

Stapes. The stapes is preserved in the smallest skull, the foot-

faces posterolaterally in the smallest skull and reorients to

plate broken away from most of the shaft, at least a section of

face laterally in the adult (Fig. 2.20B). The exoccipital and

which is ossified (Fig. 2.20A). The relative diameter of the foot-

opisthotic are fully coossified, which must occur prior to

plate of the stapes decreases with growth, and the shaft be-

hatching.

comes relatively much longer in the adult (AMNH 6254).

Lower Jaw. In the lower jaw, the predentary is well preserved,

Dentition. Maxillary and dentary tooth count also increases

uniting the ends of the dentaries in the slightly larger juvenile

from 5 in AMNH 6535 to 7 in AMNH 6536 to 10–12 teeth

(AMNH 6540; Fig. 2.22). The predentary and dentary fuse late

in adults (Sereno 1987). The teeth of both hatchling skulls

in ontogeny but can remain suturally distinct or disarticulated

exhibit wear facets, as noticed by Coombs (1982). Well-

in subadults. The bone is initially crescentic in dorsal view

developed, high-angle wear facets truncate several crown tips

but becomes more strongly U-shaped in the adult. In antero-

as in adults. As mentioned above, many aspects related to

ventral view, the predentary is subquadrate with a weakly di-

mastication are in place in the smallest hatchling skull, in-

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 51

cluding the inset of the tooth row, the presence of both ante-

characters that characterized, or varied among, the four spe-

rior bill-bearing bones, and the position and form of the lower

cies of Psittacosaurus then known (P. mongoliensis, P. meileyin-

jaw joint (Fig. 2.20). These hatchlings, thus, were not only

gensis, P. xinjiangensis, P. sinensis). The present review recog-

precocial feeders (Coombs 1982), but the masticatory appara-

nizes nine species using approximately 40 cranial characters.

tus appears to have sliced vegetation in a manner identical to

Psittacosaurus remains the most speciose dinosaurian genus,

that in the adult.

despite inflation of species numbers with taxa that now appear poorly justified. Differentiation among several species of Psittacosaurus is subtle and requires continuing review of cur-

Discussion and Conclusions

rent collections, which contain many new specimens. Such

TAXONOMY AND SPECIES RELATIONSHIPS

review must carefully consider the maturity of specimens under comparison and use dental features that are understood in

Psittacosaurus. The lone genus Psittacosaurus is characterized

the context of variation along a tooth row.

by 21 synapomorphies, about 40% of which are new to this

A Basal Species. The interrelationships of psittacosaurs were

analysis (Table 2.2). Preorbital skull length is very short, less

first addressed by Sereno (1987: table 19) using six characters

than 40% skull length (as measured from the tip of the pre-

scored in the four species that were regarded as valid at that

maxilla to the posterior aspect of the quadrate condyle at

time (P. mongoliensis, P. meileyingensis, P. xinjiangensis, P. sinen-

the jaw articulation). Basal neoceratopsians such as Liaocera-

sis). Only one of the six characters was scored as informative

tops, Archaeoceratops, and Protoceratops have preorbital lengths

(lateral projection of the jugal horn in P. xinjiangensis and

ranging from about 45 to 55% (Sereno 2000; Xu et al. 2002;

P. sinensis), a character subject to considerable developmental

You and Dodson 2003). Although Yinlong appears to be par-

and individual variation as well as postmortem distortion.

ticularly short-snouted (about 24% skull length), the holotype

Subsequently, Russell and Zhao (1996), Xu (1997), and Averia-

and only known skull is from a juvenile, and its particularly

nov et al. (2006) have presented quantitative cladistic analyses

short snout is attributable to immaturity (Xu et al. 2006).

of psittacosaur relationships, using 13, 17, and 31 characters,

The tall snout, its sutural pattern, the dorsal position of the external naris, and the unusual structure of the inter-

respectively. Recently You et al. (2008) added P. lujiatunensis and P. major to the data matrix in Averianov et al. (2006).

narial bar are all features unique to psittacosaurs among dino-

There is little consensus between these analyses regarding

saurs (Table 2.2, characters 2–7). Three other features, the

character data or results. Reanalysis of the data matrix of Rus-

premaxilla-maxilla ridge, the maxillary protuberance, and the

sell and Zhao (1996) shows that 12 of 13 characters are infor-

construction of the supratemporal bar, may well be attach-

mative, which yield 9 most parsimonious trees (23 steps; CI =

ment sites for enhanced jaw musculature (Table 2.2, charac-

0.61; RI = 0.50). A strict consensus tree has one internal node

ters 8, 9, 12, 13).

uniting P. ordosensis, P. neimongoliensis and P. sinensis on the

A unique fenestra is nearly always present over the lacrimal

basis of the relative width of the postorbital region of the skull

canal, an opening that is unrelated to the antorbital fossa or

(broad versus narrow), a relative-linear character that is diffi-

fenestra, which are absent (Table 2.2, characters 10, 11). The

cult to evaluate as stated (missing relative measure; Sereno

posterior palate is reinforced with an additional bony plate to

2007). With one additional step in length, all phylogenetic

resist transverse compression as well as a pendant mandibu-

structure within Psittacosaurus breaks down (Fig. 2.23A). The

lar process to provide attachment for enhanced pterygoideus

analysis of Xu (1997) is based on Russell and Zhao (1996),

musculature (Table 2.2, characters 14, 15). Those muscles may

although he excluded the new species P. neimongoliensis and

have pulled the lower jaw forward, taking advantage of the

P. ordosensis and included P. mazongshanensis, a species the

flat jaw articulation (Table 2.2, character 16).

validity of which has been questioned above (Fig. 2.23B). Xu

The palpebral projects strongly laterally in all species, a de-

(1997) added several characters, although the data matrix in

rived feature of unknown function (Table 2.2, character 17).

some cases does not match the number of characters or states

The predentary and dentary are firmly united, reinforcing the

in the character list as mentioned by Averianov et al. (2006).

rounded bill margin at the anterior end of the lower jaw and

Reanalysis of the data matrix in Averianov et al. (2006)

providing direct attachment for jaw musculature (Table 2.2,

shows that 30 of 31 characters are informative (Fig. 2.23C),

characters 18, 20). The jaw articulation is flat for the most

which yield seven most parsimonious trees as reported (58

part, allowing posterior displacement of the lower tooth row

steps; CI = 0.54; RI = 0.58). Although three internal nodes are

during jaw closure (Table 2.2, characters 16, 21). Finally the

present as indicated by Averianov et al. (2006), this data has

dentary teeth have a bulbous primary ridge, the shape and

little more phylogenetic structure than that in Russell and

prominence of which characterize the genus (Table 2.2, char-

Zhao (1996; Fig. 2.23A). With one additional step in length, all

acter 22).

phylogenetic structure within Psittacosaurus breaks down (Fig.

Psittacosaur species. Sereno (1987) listed some 35 cranial

52 sereno

2.23C). Likewise, when a single species with the most missing

Table 2.2. Cranial Characters with Derived States in Well-Known Species of Psittacosaurus Synapomorphy 1. Preorbital skull length: more (0), or less than (1), 40% of skull length

Original author

1

2

3

4

5

6

Sereno 1987

1

1

1

1

1

0

Sereno 1987

1

1

1

1

1

0

Sereno 1987

1

1

1

1

1

0

(rostral to quadrate condyle). 2. External naris, ventral margin, location relative to the orbital ventral margin: ventral (0); dorsal (1). 3. Anterior tip of the nasal internarial process, location: dorsal (0), or ventral (1), to the external naris. 4. Rostral-nasal contact: absent (0); present (1).

Sereno 1987

1

1

1

1

1

0

5. Premaxilla, dorsolateral process, maximum width: less (0), or

Sereno 1987

1

1

1

1

1

0

6. Premaxilla-prefrontal contact: absent (0); present (1).

this paper

1

1

1

1

1

0

7. Premaxilla-jugal approximation or contact: absent (0); present (1).

Sereno 1987

1

1

1

1

1

0

8. Maxillary fossa: absent (0); present (1).

this paper

1

1

1

1

1

0

9. Maxillary protuberance: absent (0); present (1).

Sereno 1987

1

1

1

1

1

0

10. Lacrimal canal fenestra: absent (0); present (1).

Sereno 1987

1

1

1

1

1

0

11. Antorbital fenestra and fossa: present (0); absent (1).

Sereno 1987

1

1

1

1

1

0

12. Postorbital posterior process, extension along supratemporal bar:

this paper

1

1

1

1

1

0

this paper

1

1

1

1

1

0

14. Pterygoid basal plate (secondary lamina): absent (0); present (1).

this paper

1

1

1

1

1

0

15. Pterygoid mandibular ramus, length: short process (0); pendant

Sereno 1987

1

1

1

1

1

0

subequal to (1), dorsoventral orbital diameter.

partial (0); complete (1). 13. Squamosal, end of anterior process, position on postorbital: lateral (0); dorsal (1).

process projecting into the adductor fossa (1). 16. Quadrate medial condyle: convex (0); planar (1).

this paper

1

1

1

1

1

0

17. Palpebral, orientation of posterior (medial) margin: posterolateral

this paper

1

1

?

1

1

0

this paper

1

1

?

1

1

0

this paper

1

1

1

1

1

0

20. Dentary flange: absent (0); present (1).

Sereno 1987

1

1

1

1

1

0

21. Articular, surface for quadrate condyles: concave (0); planar (1).

this paper

1

1

1

1

1

0

22. Dentary teeth, primary ridge, shape: narrow, smooth (0); cone-

Sereno 1987

1

1

1

1

1

0

(0); transverse (laterally divergent) (1). 18. Predentary ventral process(es), shape: longer than broad (0); broader than long (1). 19. Predentary anterior margin, shape (dorsal view): V-shaped (0); semicircular (1).

shaped (approximately 1/3 crown face) with secondary ridging. Note: The best known species of Psittacosaurus as well as the basal neoceratopsian Liaoceratops yanzigouesis (Xu et al. 2002) are scored. The basal neoceratopsian Chaoyangsaurus youngi (Zhao et al. 1999) also exhibits the derived state of characters 1 and 9, rendering the outgroup condition ambiguous for Psittacosaurus. Species abbreviations, 1: P. mongoliensis; 2: P. lujiatunensis; 3: P. major; 4: P. sp.; 5: P. sinensis; 6: Liaoceratops yanzigouesis.

data (P. xinjiangensis) is removed and the data reanalyzed, all

Formation (Xu and Wang 1998), with P. lujiatunensis. That

phylogenetic structure within Psittacosaurus breaks down.

replacement should have been further explained, as approxi-

When the species of questionable validity is removed (P. ma-

mately two-thirds of the character states for P. lujiatunensis

zongshanensis), only one subclade remains (P. mongoliensis,

differ from those given for P. sp. by Averianov et al. (2006). Yu

P. meileyingensis, P. xinjiangensis). This trio of species, however,

et al. (2008: fig. 6) presented a single fully resolved tree of

breaks down with one additional step in length. When the

58 steps, although reanalysis of their data yields 3 trees at that

unnamed species from the Yixian Formation (Psittacosaurus

length (Fig. 2.23D, consensus tree). As with Averianov et al.

sp.; Xu and Wang 1998) is removed, only a single pair of spe-

(2006), only one pair of species remains (P. sibiricus, P. sinensis)

cies remain (P. sibiricus, P. sinensis). In this case, two additional

when considering trees one step longer, and this species pair

steps in tree length are required before all phylogenetic struc-

breaks down with two additional steps in tree length. The present phylogenetic assessment is based on a smaller

ture breaks down. Recently, You et al. (2008) used the data matrix of Averianov

set of characters in what are viewed as the least controversial

et al. (2006), adding P. major and replacing ‘‘Psittacosaurus sp.,’’

species (Fig. 2.23E). Only 10 cranial characters are scored in 8

which is based on an incomplete specimen from the Yixian

species (Table 2.3). Using Yinlong, Liaoceratops, and Protocera-

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 53

Table 2.3. Variable Features among Species of Psittacosaurus Autapomorphy 1. Skull length relative to trunk length: approximately 30% (0); 40–45% (1). 2. Preorbital skull length relative to total skull length: 40–36% (0); 35–30% (1).

Author

1

2

3

4

5

6

7

8

9

A

0

1

1

1

1

1

1

1

1

modified

0

1

1

1

1

1

?

1

1

from B 3. External mandibular fenestra: present (0); absent (1).

B

0

1

1

1

1

1

?

1

1

4. Dentary flange, prominence: rugosity (0); low crest (1); prominent flange with

B

1

2

2

2

1

0

?

0

0

anterior corner (2). 5. Jugal fossa: absent (0); present (1).

C

0

0

1

1

0

0

?

0

0

6. Postorbital-jugal ornamentation: ridge (0); horn (1).

B

0

0

0

0

1

1

?

1

1

7. Maxillary crowns, maximum width relative to height: 75% (0); subequal (1).

A

0

0

0

0

0

1

1

1

1

8. Premaxilla-jugal contact: absent (0); present (1).

B

0

0

0

0

?

1

?

0

1

9. Supratemporal bar, orientation (dorsal view): parasagittal (0); posteriorly divergent at

B

0

0

0

0

0

0

?

1

1

A

0

0

0

0

0

0

?

1

1

an angle of approximately 15\ (1). 10. Lower jaw length: long, predentary opposes rostral and premaxilla (0); short, predentary opposes premaxilla and anterior maxillary teeth (1). Note: Original author abbreviations for characters that unite species groups, A: this paper; B: Sereno (1987); C: Zhou et al. (2006b). Character 4 is ordered. Species abbreviations, 1: P. mongoliensis; 2: P. meileyingensis; 3: P. lujiatunensis; 4: P. major; 5: P. sibiricus; 6: P. sp.; 7: P. xinjiangensis; 8: P. neimongoliensis; 9: P. sinensis.

tops as outgroups and removing the poorly known species

skull—when measured in adult skulls with uniform registra-

P. xinjiangensis (which can only be scored for two of the char-

tion of the cranium as discussed above (see discussion, general

acters), a parsimony analysis yields 3 trees of 11 steps (CI =

adult skull shape)—is approximately 30–35% of skull length

0.91; RI = 0.93). P. mongoliensis is positioned as the sister taxon

in all psittacosaur species except P. mongoliensis, where it aver-

to other psittacosaurs.

ages about 37%. Preorbital length in both Chaoyangsaurus and

One character that supports this conclusion, skull length

Yinlong, however, is short (approximately 30%) as well. Other

relative to trunk length, is new to the analysis. In P. mongolien-

basal neoceratopsians such as Archaeoceratops (You and Dod-

sis the skull is proportionately larger than in other psittaco-

son 2003), in contrast, have preorbital lengths approximately

saur species, varying between about 40–45% of trunk length

40% of skull length. The basal condition for Ceratopsia thus is

(Table 2.3, character 1). Trunk length in this regard equals the

not settled.

length of the presacral vertebral column or, if the vertebrae are

Lastly, the external mandibular fenestra is open in P. mongo-

not exposed, the length from the back of the skull roof to the

liensis but closed in other species (Table 2.3, character 3).

middle of the preacetabular process of the ilium. These pro-

P. mongoliensis is the only psittacosaur with a substantial ex-

portions are known in most species, with proportions for

ternal mandibular fenestra in fully mature individuals (e.g.,

P. lujiatunensis inferred from other specimens from the Lujia-

AMNH 6234). The basal neoceratopsian Yinlong (Xu et al.

tun Beds (Xu and Wang 1998; Mayr et al. 2002). Skull length

2006) also has an open fenestra, although this varies among

in P. mongoliensis, in contrast, is only approximately 30% of

basal neoceratopsians (e.g., small in Chaoyangsaurus and

trunk length, as best preserved in the holotype skeleton (Os-

closed in some other neoceratopsians).

born 1924; Sereno 1990b). The major question here involves

Two Clades. Some character evidence suggests that psittaco-

the outgroup condition, because neoceratopsians also have

saur species other than P. mongoliensis can be divided into two

skulls that are 40% or more of trunk length (Sereno et al.

groups, here termed the ‘‘P. major’’ and ‘‘P. sinensis’’ clades,

2007). The Late Jurassic ceratopsian Yinlong downsi (Xu et al.

which is maintained with one additional step in tree length

2006), however, resembles P. mongoliensis with a skull length

(Fig. 2.23E). The P. major-clade is characterized by hypertro-

only approximately 30% of trunk length. Because the holo-

phy of the dentary flange with development of a marked ante-

type of Y. downsi is a subadult, furthermore, its skull propor-

rior corner and a fossa centered on the jugal (Table 2.3, char-

tion is likely to decrease slightly with maturity. More remote

acters 4 and 5). This psittacosaur subclade, however, breaks

nonceratopsian ornithischians have skull lengths comparable

down with a single additional step in tree length.

to P. mongoliensis and Y. downsi, which are usually less than 30% and always less than 40% of trunk length.

The P. sinensis-clade is characterized by a postorbital-jugal horn (Table 2.3, character 6). Species in this group that are

Another feature suggesting that P. mongoliensis is basal to

more advanced than P. sibiricus have sutural contact between

other species is the preorbital length of the skull (Table 2.3,

the premaxilla and jugal and have reduced the dentary flange

character 2). The length of the preorbital segment of the

to a ridge (Table 2.3, characters 4 and 7). Finally, P. neimongo-

54 sereno

FIGURE 2.23. Previous phylogenetic hypotheses for psittacosaur species. (A) Consensus tree based on data in Russell and Zhao (1996); (B) tree given by Xu (1997); (C) consensus tree based on data in Averianov et al. (2006); (D) consensus tree based on data in You et al. (2008); (E) present phylogenetic hypothesis for psittacosaur species. Dashed lines indicate loss of resolution when considering trees one step longer than minimum length.

liensis and P. sinensis have posteriorly divergent supratem-

1996: fig. 1B), the lower jaw was shifted forward relative to the

poral bars (diverging posteriorly at an angle of 30\ from an

cranium, such that the quadrate condyle is positioned over

apex a short distance anterior to the rostral) and a lower jaw

the retroarticular process. When the jaw joint is realigned,

that is noticeably shorter than the upper jaw (Table 2.3, char-

P. neimongoliensis also has the very short lower jaw. In other

acters 8 and 9). The short lower jaw is not an artifact of preser-

species such as P. mongoliensis, it is impossible to slide the

vation or movement relative to the cranium. Its shorter length

lower jaws posteriorly into this position.

is preserved in natural articulation in the holotypic skull of

Ceratopsian Roots. There has yet to be found a truly transi-

P. sinensis (Fig. 2.8). The short length of the lower jaw posi-

tional form to Psittacosaurus that would help to polarize char-

tions the predentary posterior to the rostral in opposition to

acters within the genus, which is restricted to a relatively nar-

the premaxilla and anteriormost maxillary teeth. In the origi-

row temporal interval within the Early Cretaceous. Recently

nal reconstruction of P. neimongoliensis (Russell and Zhao

other ceratopsians have been discovered in beds of similar

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 55

age (Xu et al. 2002) and in earlier horizons of Late Jurassic age (Xu et al. 2006). Most notable among these finds are Chaoyangsaurus (Zhao et al. 1999) and Yinlong (Xu et al. 2006), which have been placed in some arrangements outside other ceratopsians including psittacosaurs (Xu et al. 2002, 2006). In skull structure and masticatory function, however, the psittacosaur pattern described in this account differs in many ways from these ceratopsians, which are characterized by a narrow hooked upper beak, deep supratemporal bar, markedly divergent tooth rows, and very short retroarticular process. These neoceratopsian features may point to an early divergence from psittacosaurs during the Jurassic, a hypothesis that awaits a rigorous test from new character data on early ceratopsians. Acknowledgments

I thank J. Tanner for the pencil drawings in Figs. 2.10, 2.11, and 2.15, and C. Abraczinskas for labeling these figures and for drafting all other figures. W. Amaral prepared the juvenile specimens of P. mongoliensis, and R. Masek and E. Dong prepared the adult remains of P. sp. and P. major. For access to specimens, I thank E. Gaffney (AMNH), X. Xu and X. Zhao (IVPP), S. Zhen and C. Rao (BNHM), L. Tan (LH), and S. Kurzanov and T. Tumanova (PI). For access to images of basal ceratopsians, I thank P. Makovicky. For comments on the manuscript, I thank P. Makovicky and M. Ryan. For support of the research, I thank the National Science Foundation (Doctoral Dissertation Improvement Grant), the American Museum of Natural History, the National Geographic Society, the David and Lucile Packard Foundation, and the University of Chicago. References Cited Andrews, R. C. 1932. The New Conquest of Central Asia: A Narrative of the Explorations of the Central Asiatic Expeditions in Mongolia and China, 1921–1930. New York: American Museum of Natural History. Averianov, A. O., A. V. Voronkevich, S. V. Leshchinskiy, and A. V. Fayngertz. 2006. A ceratopsian dinosaur Psittacosaurus sibiricus from the Early Cretaceous of west Siberia, Russia and its phylogenetic relationships. Journal of Systematic Palaeontology 4: 359–395. Brinkman, D. B., D. A. Eberth, M. J. Ryan, and P. Chen. 2001. The occurrence of Psittacosaurus xinjiangensis Sereno and Chow, 1988 in the Urho area, Junggar Basin, Xinjiang, People’s Republic of China. Canadian Journal of Earth Sciences 38: 1781– 1786. Brown, B., and E. M. Schlaikjer. 1940. The structure and relationships of Protoceratops. Annals of the New York Academy of Sciences 40: 133–266. Buffetaut, E., N. Sattayarak, and V. Suteethorn. 1989. A psittacosaurid from the Cretaceous of Thailand and its implications for the palaeogeographical history of Asia. Terra Nova 1: 370–373.

56 sereno

Buffetaut, E., and V. Suteethorn. 1992. A new species of the ornithischian dinosaur Psittacosaurus from the Early Cretaceous of Thailand. Palaeontology 35: 801–812. ———. 2002. Remarks on Psittacosaurus sattayaraki Buffetaut & Suteethorn, 1992, a ceratopsian dinosaur from the Lower Cretaceous of Thailand. Oryctos 4: 71–73. Cheng, Z. 1983. Reptilia (Chapter 7). In The Mesozoic Stratigraphy and Paleontology of Guyang Coal-bearing Basin Nei Monggol Autonomous Region, China, pp. 123–136. Beijing: Geology Press. Colbert, E. H. 1945. The hyoid bones in Protoceratops and Psittacosaurus. American Museum Novitates 1301: 1–10. Coombs, W. P., Jr. 1980. Juvenile ceratopsians from Mongolia— the smallest known dinosaur specimens. Nature 283: 380–381. ———. 1982. Juvenile specimens of the ornithischian dinosaur Psittacosaurus. Palaeontology 25: 89–107. Dong, Z. 1973. Dinosaurs from Wuerho. Reports of paleontological expedition to Sinkiang II. Pterosaurian fauna from Wuerho, Sinkiang. Memoirs of the Institute of Vertebrate Paleontology and Paleoanthropology 11: 45–52. [In Chinese.] ———. 1993. A new species of stegosaur (Dinosauria) from the Ordos Basin, Inner Mongolia, People’s Republic of China. Canadian Journal of Earth Sciences 30: 2174–2176. Gradstein, F. M., J. G. Ogg, and A. G. Smith, eds. 2004. A Geologic Time Scale 2004. Cambridge: Cambridge University Press. He, H. Y., X. L. Wang, Z. H. Zhou, F. Jin, F. Wang, L. K. Yang, X. Ding, A. Boven, and R. X. Zhu. 2006. 40Ar/ 39Ar dating of Lujiatun Bed ( Jehol Group) in Liaoning, northeastern China. Geophysical Research Letters 33. Lü, J., Y. Kobayashi, Y. N. Lee, and Q. Ji. 2007. A new Psittacosaurus (Dinosauria: Ceratopsia) specimen from the Yixian Formation of western Liaoning, China: The first pathological psittacosaurid. Cretaceous Research 28: 272–276. Lucas, S. G. 2006. The Psittacosaurus biochron, Early Cretaceous of Asia. Cretaceous Research 27: 189–198. Marsh, O. C. 1889. The skull of the gigantic Ceratopsidae. American Journal of Science, Series 3, 38: 501–506. ———. 1890. Description of new Dinosaurian reptiles. American Journal of Science, Series 3, 39: 82–86. Mayr, E., E. G. Linsley, and R. L. Usinger. 1953. Methods and Principles of Systematic Zoology. New York: McGraw-Hill Book Co. Mayr, G., S. D. Peters, G. Plodowski, and O. Vogel. 2002. Bristlelike integumentary structures at the tail of the horned dinosaur Psittacosaurus. Naturwissenschaften 89: 361–365. Meng, Q., J. Liu, D. J. Varricchio, T. Huang, and C. Gao. 2004. Parental care in an ornithischian dinosaur. Nature 431: 145–146. Norman, D. B., and D. B. Weishampel. 1991. Feeding mechanisms in some small herbivorous dinosaurs: Processes and patterns. In J. M. V. Rayner and R. J. Wootton, eds., Biomechanics in Evolution, pp. 161–181. Cambridge: Cambridge University Press. Osborn, H. F. 1923. Two Lower Cretaceous dinosaurs from Mongolia. American Museum Novitates 95: 1–10. ———. 1924. Psittacosaurus and Protiguanodon: Two Lower Cretaceous iguanodonts from Mongolia. American Museum Novitates 127: 1–16. Ostrom, J. H. 1961. Cranial morphology of the hadrosaurian di-

nosaurs of North America. Bulletin of the American Museum of Natural History 122: 33–186. Owen, R. 1842. Report on British fossil reptiles. Part II. Report of the British Association for the Advancement of Science 1841: 60– 204. Qi, Z., P. M. Barrett, and D. A. Eberth. 2007. Social behaviour and mass mortality in the basal ceratopsian dinosaur Psittacosaurus (Early Cretaceous, People’s Republic of China). Palaeontology 50: 1023–1029. Rougier, G. W., M. J. Novacek, M. C. McKenna, and J. R. Wible. 2001. Gobiconodonts from the Early Cretaceous of Oshih (Ashile), Mongolia. American Museum Novitates 3348: 1–30. Rozhdestvensky, A. K. 1955. New data on psittacosaurs— Cretaceous ornithopods. In Questions on the Geology of Asia. Vol. 2: Izdatel’stvo, pp. 783–788. Moscow: Akademii Nauk SSSR. [In Russian.] ———. 1960. Locality of Lower Cretaceous dinosaurs in the Kuzbass. Paleontologicheskii Zhurnal 2: 165. [In Russian.] ———. 1977. The study of dinosaurs in Asia. Journal of the Palaeontological Society of India 20: 102–119. Russell, D. A., and X. Zhao. 1996. New psittacosaur occurrences in Inner Mongolia. Canadian Journal of Earth Sciences 33: 637–648. Seeley, H. G. 1888. The classification of the Dinosauria. Report of the British Association for the Advancement of Science 1887: 698–699. Sereno, P. C. 1987. The ornithischian dinosaur Psittacosaurus from the Lower Cretaceous of Asia and the relationships of the Ceratopsia. Ph.D. diss., Columbia University, New York. ———. 1990a. New data on parrot-beaked dinosaurs (Psittacosaurus). In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 203–210. Cambridge: Cambridge University Press. ———. 1990b. Psittacosauridae. In D. B. Weishampel, P. Dodson and H. Osmólska, eds., The Dinosauria, pp. 579–592. Berkeley: University of California Press. ———. 1990c. Clades and grades in dinosaur systematics. In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 9–20. Cambridge: Cambridge University Press. ———. 2000. The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia. In M. Benton, E. Kurochkin, M. Shishkin, and D. Unwin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. 480–516. Cambridge: Cambridge University Press. ———. 2005. Stem Archosauria—TaxonSearch. http:// www.taxonsearch.org/dev/file—home.php (version 1, 2005 November 7), Chicago. ———. 2007. Logical basis for morphological characters in phylogenetics. Cladistics 23: 565–587. Sereno, P. C., S. Chao, Z. Cheng, and C. Rao. 1988. Psittacosaurus meileyingensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northeastern China. Journal of Vertebrate Paleontology 8: 366–377. Sereno, P. C., S. McAllister, and S. L. Brusatte. 2005. TaxonSearch: A relational database for suprageneric taxa and phylogenetic definitions. PhyloInformatics 8: 1–20. Sereno, P. C., and S. Zhao [Chao]. 1988. Psittacosaurus xin-

jiangensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northwestern China. Journal of Vertebrate Paleontology 8: 353–365. Sereno, P. C., X.-J. Zhao, L. Brown, and L. Tan. 2007. New psittacosaurid highlights skull enlargement in horned dinosaurs. Acta Palaeontologica Polonica 52: 275–284. ———. In review. New psittacosaurid from the Bayan Gobi Formation of western Inner Mongolia, China. Proceedings of the Royal Society of London. Steel, R. 1969. Ornithischia. Handbuch der Paläoherpetologie 15: 1–84. Versluys, J. 1923. Der Schädel des Skelettes von Trachodon annectens im Senckenberg-Museum. Abhandlungen der Senckenburg Naturforsh. Ges. 38: 1–19. Xu, X. 1997. A new psittacosaur (Psittacosaurus mazongshanensis sp. nov.) from Mazongshan Area, Gansu Province, China. In Z. Dong, ed., Sino-Japanese Silk Road Dinosaur Expedition, pp. 48–67. Beijing: China Ocean Press. Xu, X., C. A. Forster, J. M. Clark, and J. Mo. 2006. A basal ceratopsian with transitional features from the Late Jurassic of northwestern China. Proceedings of the Royal Society B: Biological Sciences 273: 2135–2140. Xu, X., P. J. Makovicky, X. Wang, M. A. Norell, and H. You. 2002. A ceratopsian dinosaur from China and the early evolution of Ceratopsia. Nature 416: 314–317. Xu, X., and M. A. Norell. 2006. Non-avian dinosaur fossils from the Lower Cretaceous Jehol Group of western Liaoning, China. Geological Journal 41: 419. Xu, X., and X. Wang. 1998. New psittacosaur (Ornithischia, Ceratopsia) occurrence from the Yixian Formation of Liaoning, China and its stratigraphical significance. Vertebrata PalAsiatica 1998: 147–158. Xu, X., and X. Zhao. 1999. Psittacosaur fossils and their stratigraphical implications. In Y. Wang and T. Deng, eds., Proceedings of the Seventh Annual Meeting of the Chinese Society of Vertebrate Paleontology, pp. 75–80. Beijing: China Ocean Press. You, H., and P. Dodson. 2003. Redescription of the neoceratopsian dinosaur Archaeoceratops and the early evolution of Neoceratopsia. Acta Palaeontologica Polonica 48: 261–272. ———. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 478–493. Berkeley: University of California Press. You, H., K. Tanoue, and P. Dodson. 2008. New data on cranial anatomy of the ceratopsian dinosaur Psittacosaurus major. Acta Palaeontologica Polonica 53: 183–196. You, H., and X. Xu. 2005. An adult specimen of Hongshanosaurus houi (Dinosauria: Psittacosauridae) from the Lower Cretaceous of western Liaoning Province, China. Acta Geologica Sinica 79: 168–173. You, H., X. Xu, and X. Wang. 2003. A new genus of Psittacosauridae (Dinosauria: Ornithopoda) and the origin and early evolution of marginocephalian dinosaurs. Acta Geologtica Sinica 77: 15–20. Young, C. C. 1931. On some new dinosaurs from western Suiyan, Inner Mongolia. Bulletin of the Geological Survey China 2: 159– 166.

Taxonomy, Cranial Morphology, and Relationships of Parrot-Beaked Dinosaurs 57

———. 1958. The dinosaurian remains of Laiyang, Shantung. Palaeontologia Sinica (C) 16: 53–138. Zhao [Chao], X. 1962. New species of Psittacosaurus from Laiyang, Shantung. Vertebrata PalAsiatica 6: 349–360. [In Chinese.] Zhao, X., Z. Cheng, and X. Xu. 1999. The earliest ceratopsian from the Tuchengzi Formation of Liaoning, China. Journal of Vertebrate Paleontology 19: 681–691. Zhou, C., K. Gao, X. Du, W. Qi, and S. Zhang. 2006a. Advances in the study of psittacosaurids and the application of CT Scan. Acta Scientiarum Naturalium Universitatis Pekinensis 42: 146–152. Zhou, C., K. Gao, R. C. Fox, and S. Chen. 2006b. A new species of Psittacosaurus (Dinosauria: Ceratopsia) from the Early Cretaceous Yixian Formation, Liaoning, China. Palaeoworld 15: 100– 114.

58 sereno

Zhou, C., K. Gao, R. C. Fox, and X. Du. 2007. Endocranial morphology of psittacosaurs (Dinosauria: Ceratopsia) based on CT scans of new fossils from the Lower Cretaceous, China. Palaeoworld 16: 285–293. Zhu, R., Y. Pan, R. Shi, Q. Liu, and D. Li. 2007. Palaeomagnetic and 40Ar/ 39Ar dating constraints on the age of the Jehol Biota and the duration of deposition of the Sihetun fossil-bearing lake sediments, northeast China. Cretaceous Research 28: 171– 176. Zusi, R. L. 1993. Patterns of diversity in the avian skull. In J. Hanken and B. K. Hall, eds., The Skull: Patterns of Structural and Systematic Diversity, pp. 391–437. Chicago: University of Chicago Press.

3 A New Species of Archaeoceratops (Dinosauria: Neoceratopsia) from the Early Cretaceous of the Mazongshan Area, Northwestern China H A I - L U Y O U , K Y O TA N O U E , A N D P E T E R D O D S O N

a new species of basal neoceratopsian dinosaur,

Basin in the Mazongshan area and is characterized by the co-

Archaeoceratops yujingziensis, is described. The

existence of basal representatives of the diverse clades that

specimen was collected from the Lower Cretaceous Xin-

would subsequently dominate Late Cretaceous dinosaur fau-

minpu Group of the Yujingzi Basin in the Mazongshan

nas. It includes two basal neoceratopsians, Archaeoceratops os-

area of northwestern China, and is represented by a

himai (Dong and Azuma 1997) and Auroraceratops rugosus (You

partial skull, right mandible, and partial postcranial

et al. 2005), the basal hadrosauroid Equijubus normani (You

skeleton. A. yujingziensis differs from the type species,

et al. 2003a), and the basal titanosauriform Gobititan shenz-

A. oshimai, in having a laterally deflected rostral end of

houensis (You et al. 2003b). Comparative study shows that,

the maxilla; striations on the enameled premaxillary

although closely related, all dinosaurs from the Xinminpu

teeth; absence of a primary ridge on the maxillary teeth;

Group in the Gongpoquan Basin are more derived than those

and a horizontal shelf on the dentary teeth. The discov-

from the Jehol Group in western Liaoning Province in their

ery of A. yujingziensis extends the geographic distribution

respective clades (Xu and Norell 2006). Thus, the Gongpo-

of the genus 100 km southeast, and is compatible with

quan dinosaur assemblage probably represents a later, more

the Early Cretaceous age designation for the dinosaur-

derived stage in the evolution of the Early Cretaceous Psit-

bearing beds of the Yujingzi Basin in the Mazongshan

tacosaurus fauna in northern China (You and Luo 2008). In 2002, a joint expedition of the Chinese Academy of Geo-

area.

logical Sciences and the Gansu Provincial Museum worked in

Introduction

the Yujingzi Basin in the Mazongshan area, about 100 km southeast of the Gongpoquan Basin (Fig. 3.1). Subsequent

The Lower Cretaceous Xinminpu Group of the Mazongshan

fieldwork in this area has been conducted since 2004 by the

area in northwestern China has yielded numerous dinosaurs

Fossil Research and Development Center of the Third Geology

mainly due to two projects: the Sino-Japanese Silk Road Dino-

and Mineral Resources Exploration Academy of Gansu Prov-

saur Expedition in 1992 (Dong and Azuma 1997), and the

ince. Abundant dinosaur material has been found, and a new

Sino-American Mazongshan Dinosaur Project during 1997–

genus of therizinosaurian dinosaur, Suzhousaurus megathe-

2000 (You 2002). A diverse dinosaur assemblage, includ-

rioide, has been studied (Li et al. 2007). Here, we describe a new

ing members of Theropoda, Sauropoda, Iguanodontoidea,

specimen of a basal neoceratopsian dinosaur found in 2002.

and Neoceratopsia, has been discovered in the Gongpoquan

Comparative study shows that it represents a new species of

59

FIGURE 3.1.

Map showing the localities of Archaeoceratops yujingziensis n. sp. (south) and Archaeoceratops oshimai (north).

Archaeoceratops. Therefore, this finding provides new informa-

lack of a primary ridge on the maxillary teeth; and a horizon-

tion on the anatomy of Archaeoceratops, extends its geographi-

tal shelf on the dentary teeth.

cal distribution, and confirms the Early Cretaceous age of the dinosaur-bearing beds in the Yujingzi Basin.

Locality and Horizon. Yujingzi Basin, Mazongshan area, northwestern China. Xinminpu (= ‘‘Xinminbao’’) Group, Early Cretaceous, ?Aptian-Albian (Tang et al. 2001).

Systematic Paleontology Description

Ornithischia Seeley 1888 Ceratopsia Marsh 1890

The specimen was not discovered in the field but came to light

Neoceratopsia Sereno 1986

during preparation of an iguanodontian dinosaur in the labo-

Archaeoceratops Dong and Azuma 1997

ratory. Fusion of the cranial sutures is suggestive of maturity,

Archaeoceratops yujingziensis n. sp.

but the unfused sacral centra and unfused neural arches are immature characters.

Type specimen. Chinese Academy of Geological SciencesInstitute of Geology: CAGS-IG-VD-003. The material includes

SKULL

a partial skull extending from the right premaxilla to jugal and mandibular process of the pterygoid, the distal part of the

The intact partial skull (75 mm long) consists of elements on

right quadrate, complete right mandible, a dorsal neural arch,

the right side, from the premaxilla to the middle of the jugal,

a sacral centrum, three proximal caudals, a partial right scap-

and part of the mandibular process of the pterygoid. The ven-

ula, both femora, two metatarsals, and three pedal phalanges.

tral borders of the antorbital fossa and the orbit, and also the

Etymology. ‘‘Yujingzi’’ (Chinese): name of the basin where

rostral end of the mandibular adductor fossa, are included

the specimen was discovered.

(Fig. 3.2).

Diagnosis. A. yujingziensis possesses four autapomorphies

The premaxilla measures 20 mm in length and 11 mm in

that do not exist in A. oshimai: a laterally deflected rostral end

height as preserved, but neither the rostral nor the dorsal bor-

of the maxilla, causing a spoon-shaped premaxillary beak

der are complete. There is a horizontal palatal projection. The

with the premaxillary teeth placed lateral to the line of the

premaxillary-maxillary suture can be traced on the side of the

maxillary teeth; striations on the enameled premaxillary

premaxilla just caudal to the second tooth. In palatal view it

teeth, perhaps combined with their ventrolateral orientation;

passes either through or immediately behind a small alveolus

60 you, tanoue, & dodson

Partial skull of Archaeoceratops yujingziensis n. sp. (CAGS-IG-VD-003) in (A, B) right lateral; (C, D) left lateral; (E, F) dorsal; and (G, H) ventral views. A, C, E, and G are photographs. B, D, F, and H are interpretive outlines. Scale bar is 2 cm. ect: ectopterygoid; j: jugal; mx: maxilla; pl: palatine; pmx: premaxilla; pt: pterygoid.

FIGURE 3.2.

for a vestigial third tooth. The line of the premaxillary teeth

nature of the premaxillary-maxillary suture rostral and dorsal

falls far lateral to the line of the maxillary tooth row (about

to the antorbital fossa remains unknown. The ventral border

10 mm), and the teeth are oriented somewhat ventrolaterally

of the antorbital fossa is very sharply defined and appears to

instead of straight ventrally.

slope rostroventrally; more typically in basal neoceratopsians

The right maxilla is largely complete ventrally, from the

the ventral edge of the antorbital fossa is horizontal. The max-

premaxilla to its caudal end, and contains the complete tooth

illary tooth row is strongly inset. A series of at least six shallow

row. However, it is incomplete dorsally, from the level of the

pits occurs on the ventrolaterally facing surface of the maxilla,

ventral border of the antorbital fossa. The relationships with

dorsal to the alveolar margin. A striking and perhaps autapo-

most adjacent bones cannot be determined. For example, the

morphic feature is the sharp rostrolateral bend or deflection of

A New Species of Archaeoceratops 61

the edentulous rostral part of the ventral border of the max-

MANDIBLE

illa. This results in an almost duck-beak appearance of the premaxilla. There is a maxillary contribution to the secondary

The right mandible is mostly preserved, missing only the tip

palate. Overlapping the caudal end of the maxilla is the ecto-

of the predentary, rostral end of the dentary tooth row, pre-

pterygoid, which contributes to the dorsolateral component

articular, and the region of dentary-surangular-angular junc-

of the pendant mandibular process of the pterygoid. The me-

tion (Fig. 3.3). It measures 98 mm in length as preserved, miss-

dial face of the maxilla is a vertically oriented planar surface.

ing only about 5 mm rostrally. Its reconstructed length is

There are six dental foramina on the medial surface situated

about the same as the femur. It is 25 mm wide and 43 mm

between the fourth and tenth maxillary teeth.

high. In lateral view, the ventral margin is convex ventrally,

The jugal overlaps the lateral side of the maxilla ventral to

and the pyramidal coronoid process stretches dorsally in the

the orbit and from the level of the mandibular adductor fossa

caudal half. Running across the base of the coronoid process

to the antorbital fossa. Due to the position of a crack, it cannot

there is a well-developed dorsally convex lateral ridge, whose

be determined if the jugal only closely approached the antor-

dorsalmost point is at the level of the dorsal end of the tooth

bital fossa or actually reached its caudal border. The preserved

row. This ridge is also well developed in A. oshimai (You and

fragment of jugal includes part of the ventral orbital rim; it

Dodson 2003). The existence of an external mandibular fora-

extends a short distance caudally opposite the rostral end of

men is unknown due to limits of preservation.

the mandibular adductor fossa. The caudal part shows a pus-

The predentary has a rostrodorsally elevated tip, a broad,

tulose texture typical of the jugals of Archaeoceratops and of

beveled triturating surface, and a very long ventral process.

other basal neoceratopsians. The jugal is contacted medially

The preserved predentary measures 19 mm in height and 32

by the ectopterygoid, which forms the rostral border of the

mm along the ventral margin. The lateral margins diverge at

mandibular adductor fossa.

an angle of about 55\ in dorsal view, which is larger than that

The rostrodorsally inclined palatine contacts the medial

of A. oshimai (approximately 45\). In lateral view, its dorsal

surface of the maxilla between the fifth and last maxillary

border is concave dorsally and the ventral margin is convex

tooth. It extends laterally 8 mm and may frame the channel

ventrally. The caudally stretching caudolateral process is very

from the antorbital fossa into the suborbital space. The surface

short—only 4 mm in length—compared to the ventral process.

of the maxilla dorsal to the rostral half of the tooth row and

The 4 mm long interdentary process tapers caudoventrally in

rostral to the palatine is generally flat and low. There is a sharp

dorsal view.

medial ridge marking the junction between the vertical me-

In lateral view, the dentary covers about half of the man-

dial face and the horizontal dorsal surface. Ten mm lateral to

dible. It contacts the predentary rostrally, the surangular

the medial ridge is a second low ridge that leads from the

caudodorsally, and the angular caudoventrally. Its caudal half

rostral border of the antorbital fossa into the floor of the sub-

contributes to the rostral half of the lateral ridge, and its

orbital space.

caudodorsal portion to the rostral half of coronoid process,

The pterygoid contributes the ventromedial part of the

including the apex. The dentary ramus is relatively low, with a

mandibular process, but only a portion of that is preserved. A

height of approximately 20 mm. The dentary is widest at the

slot along the decurved caudomedial edge of the mandibular

level of the penultimate tooth, measuring 21 mm. In dorsal

process represents the articular surface between the ectoptery-

view, there are two deep foramina along the lateral margin,

goid and the pterygoid.

lateral to the midpoint of the preserved dentary tooth row.

The banana-shaped ectopterygoid runs from the distal tip

There is also a smaller foramen lateral to the rostralmost tooth.

of the caudoventrally inclined mandibular process across

The caudal end of the dentary contributes to the rostral border

the caudal aspect of the maxilla, then bends laterally and

of the mandibular fossa.

slightly rostrally to contact the medial surface of the jugal.

The surangular is bounded by the dentary rostrally, the an-

The ectopterygoid would be expected to form part of the bor-

gular ventrally, and the articular caudomedially. It forms the

der of the palatine foramen, but due to the loss of its medial

caudal half of the coronoid process. Its dorsalmost end at the

portion in this specimen, the palatine formation cannot be

rostrodorsal corner is 2 mm lower than the apex of coronoid

recognized.

process. The caudodorsal margin stretches caudoventrally

The distal portion of the right quadrate is preserved. Its su-

from this point. The slope of the caudodorsal margin (about

ture to the quadratojugal, which is along its rostrolateral mar-

25\) is a little steeper than that of A. oshimai (about 20\). The

gin, is clear. Its medial edge is concave laterally, and the

caudal half of the lateral ridge runs along the caudolateral

pterygoid process should start right at the preserved dorso-

border of the surangular. At the caudal end of the lateral ridge,

medial corner. The ventral surface of the quadrate is long me-

a low protuberance extends laterally. Its caudomedial portion

diolaterally, and bears a median depression.

contributes to the lateral side of the lateral cotyle. The caudal

62 you, tanoue, & dodson

Right mandible of Archaeoceratops yujingziensis n. sp. (CAGS-IG-VD-003) in (A, B) right lateral; (C, D) left lateral; and (E, F) dorsal views. A, C, and E are photographs. B, D, and F are interpretive outlines. Scale bar is 2 cm. an: angular; ar: articular; c: coronoid; d: dentary; pd: predentary; sa: surangular; sp: splenial.

FIGURE 3.3.

end of the surangular composes the short retroarticular pro-

of the dentary and the angular. The rostrodorsal margin as-

cess along with the articular.

cends caudodorsally. The caudoventral margin composes the

The angular composes the caudoventral part of the mandible. It contacts the dentary rostrally, surangular dorsally,

ventral and possibly rostral border of the mandibular fossa. The ventral margin of the splenial is convex ventrally.

prearticular rostromedially, and articular caudodorsally. In

The articular is rectangular in dorsal view and is oblique

lateral view, it is about twice as large as the surangular. The

with the long axis directed caudomedially. In caudal view it is

ventral margin ascends caudodorsally at about 30\. Its caudal

8 mm high. The rostral three-quarters of its dorsal surface and

end reaches almost to the caudal edge of the mandible.

the surangular compose most of the articular surface to re-

The coronoid contributes to the medial side of the coronoid

ceive the quadrate. Both lateral and medial cotyles are shal-

process. Its ventral end is missing. The preserved coronoid is

low. Caudal to the articular surface, the 4 mm long retroarticu-

12 mm wide and 15 mm high. It widens dorsally and slightly

lar process exhibits a shallow depression for the mandibular

caudally.

depressor muscle. The rostral side of the articular contributes

A partial intercoronoid is preserved. It is best observed in

to the caudal border of the mandibular fossa.

ventral view through the opening on the lateral surface of the mandible. The intercoronoid is composed of a thin strip between the coronoid and the surangular. It is 6 mm long along

DENTITION

the shaft and 3 mm wide as preserved. The intercoronoid ex-

There are two stout, simple, peg-like premaxillary teeth with

tends caudodorsally along the caudoventral margin of the

enamel on both surfaces (Fig. 3.4A). The first tooth is the more

coronoid. It also tapers caudodorsally.

slender. In labial view, the first premaxillary tooth is 6.2 mm

The splenial lacks its rostral, dorsal, and caudal ends. The

long and its mesiodistal and labiolingual diameters are 3.1

preserved portion is 48 mm long. It covers the medial surface

mm and 2.4 mm, respectively. The second premaxillary tooth

A New Species of Archaeoceratops 63

FIGURE 3.4.

Dentition of Archaeoceratops yujingziensis n.sp. (CAGS-IGVD-003). (A) Right premaxillary teeth in lingual view; (B) right maxillary tooth row in labial view; (C) right dentary tooth row in lingual view. Arrow indicates denticles. Scale bars are 5 mm.

is 3.6 mm long in labial view and its mesiodistal and labio-

lingual surface. The mesial side of the crown from the dorsal-

lingual diameters are 3.6 mm and 2.7 mm, respectively. The

most point is about half the width of distal side. As a result, the

crowns are laterally compressed, with striations on the labial

crown is strongly asymmetrical. Two to three denticles are

and lingual surfaces and fine denticles (8 per mm on the distal

present along the mesial side of the crown, and five to six are

carina of the first tooth) on the mesial and distal carinae. Pre-

present along the distal side.

maxillary tooth denticles are also reported in Liaoceratops and Yamaceratops (Makovicky and Norell 2006). The columnar root tapers slightly towards the base of the crown. In ventral view, the premaxillary tooth is labiolingually symmetrical.

AXIAL SKELETON A single dorsal neural arch is preserved (Fig. 3.5A, B). Its ven-

Twelve alveoli comprise the 37 mm long maxillary tooth

tral border, which roofs the neural canal, is transversely wide,

row. Nine teeth are present, including the first through sixth

surrounding a relatively large neural canal. The prezygapo-

and ninth through eleventh teeth. Three of the teeth are pre-

physeal surfaces face dorsomedially, while the postzygapo-

served erupting (Fig. 3.4B). In ventral view, the tooth row is

physeal surfaces face ventrolaterally. The diapophyses are di-

lenticular, with its width increasing from the rostral end to the

rected dorsolaterally, as well as caudally.

middle, and decreasing caudally from the midpoint. The sec-

A single sacral centrum is preserved (Fig. 3.5C–E). It is prob-

ond, fourth, and fifth teeth show a steep wear facet. In labial

ably the first sacral centrum because it shares some features

view, the maxillary tooth crown is asymmetrical in that the

with the first sacral centrum of A. oshimai (You and Dodson

mesial side from the ventralmost point is about twice as wide

2003). These include a deep groove along the midline of its

as the distal side. The labial surface of the crown bears low

dorsal surface, a prominent articular surface for the first sacral

ridges. Maxillary teeth of A. oshimai have a low primary ridge,

rib along its caudolateral corner, and a smooth ventral surface

but it is not present in this specimen (You and Dodson 2003).

without a shallow groove.

There are two to three denticles both along mesial and distal sides of the crown.

Three articulated proximal caudal vertebrae are preserved (Fig. 3.5F–I). Only the first caudal includes a centrum. The

There are 10 dentary teeth preserved, including the caudal-

central articular surfaces are flat, wider than high, and con-

most tooth of the tooth row(Fig. 3.4C). The preserved dentary

strict ventrally. The ventral surfaces are smooth, and the artic-

tooth row is 34 mm long. The dentary tooth row is inset me-

ular surfaces for chevrons are not prominent. The articular

dial to the coronoid process as in A. oshimai (You and Dodson

surfaces for the transverse processes are large and situated

2003). Although the medial margin of the dentary is lingually

across the centro-neural arch suture. The zygapophyseal sur-

convex in dorsal view, dentary teeth are aligned in an almost

faces face more mediolaterally than dorsoventrally. The neu-

straight line. The teeth are closely packed in contact with each

ral spines are tall, about twice the height of the centra, and are

other. The preserved rostralmost tooth lacks its rostral half. All

directed more dorsally than caudally.

teeth except for the caudalmost tooth have steep wear facets, as in the maxillary teeth. At least two erupting teeth can be observed on the medial side of dentary, caudal to the third and

APPENDICULAR SKELETON

sixth from the last tooth. As in maxillary teeth, dentary tooth

A 38 mm long and 11 mm wide strap-like bone is identified as

crowns lack primary ridges and show only low ridges on the

a portion of the right scapula (Fig. 3.6A, B). Its lateral surface is

64 you, tanoue, & dodson

FIGURE 3.5.

Axial skeleton of Archaeoceratops yujingziensis n. sp. (CAGS-IGVD-003). Dorsal neural arch in (A) dorsal and (B) ventral views. First sacral centrum in (C) dorsal view; (D) ventral view; and (E) cranial view. Three proximal caudal vertebrae in (F) left lateral view; (G) dorsal view; (H) proximal view; and (I) distal view. Scale bar is 1 cm.

Discussion

smooth, while the medial surface bears a ridge along the ventral half, and becomes more prominent proximately. Both femora are almost completely preserved (Fig. 3.6C–H).

Archaeoceratops is one of the best-preserved Early Cretaceous

The entire length of the left femur is 102 mm, while that of

basal neoceratopsians (Xu et al. 2002; You and Dodson 2004;

the right one is 99 mm due to the weathering of its distal end.

Makovicky and Norell 2006), and includes the type species,

In caudal view and oriented vertically, the proximal half is

A. oshimai, which is represented by both a holotype (a nearly

twisted medially. The femoral head is more completely pre-

complete skull and jaws, partial vertebral column, and partial

served in the right femur than in the left. In proximal view,

pelvis) and a paratype (a partial vertebral column including

the head is axially compressed, projecting medially with a

a nearly complete tail, a partial pelvis, fragmentary hind

length about twice its width. The head is separated from the

limb bones, and a complete pes) from the late Early Creta-

greater trochanter by a constriction formed by troughs on

ceous Gongpoquan Basin of the Mazongshan area (Dong and

both the cranial and caudal surfaces of the proximal portion

Azuma 1997; You and Dodson 2003). It differs from other

of the element. In proximal view, the greater trochanter is

neoceratopsians in having a modest bumpy ornamentation

axially expanded, making the proximal end of the femur ap-

covering much of the lateral surface of the jugal, an exca-

pear as a capital ‘‘T,’’ whereas the cranial trochanter appears

vation on the lateral surface of the ischiadic peduncle, and

curved like a quotation mark.

strongly reduced shaft and proximal end of metatarsal I (You

The distal ends of right metatarsal III and IV are preserved

and Dodson 2003).

(Fig. 3.6I, J). The distal surface of metatarsal III is rectangular,

CAGS-IG-VD-003 can be assigned to Archaeoceratops based

while that of metatarsal IV is quasi-square. Three articulated

on the general similarity between them, as well as the great

pedal phalanges are identified as 1–3 of pedal digit IV (Fig.

differences between it and other basal neoceratopsians. For

3.6K). Their lengths decrease distally, from 17 mm, to 12 mm,

example, CAGS-IG-VD-003 differs from Liaoceratops in having

and to 8 mm, respectively.

a shelf on the surangular overhanging the angular and a verti-

A New Species of Archaeoceratops 65

FIGURE 3.6.

Appendicular skeleton of Archaeoceratops yujingziensis n. sp. (CAGS-IG-VD-003). A portion of right scapula in (A) medial view; and (B) lateral view. Left femur in (C) proximal view; (D) distal view; (E) lateral view; (F) caudal view; (G) medial view; and (H) cranial view. (I) istal end of right metatarsal III in dorsal view; ( J) distal end of right metatarsal IV in dorsal view; (K) phalanges 1–3 of pedal digit IV in dorsal view. Scale bars are 1 cm.

cal tab on the surangular, lateral to the glenoid cotyle (char-

yujingziensis is smaller still. Compared to A. oshimai, it is only

acters 84 and 86 of Makovicky and Norell 2006), from Au-

550 mm long and 230 mm at the hips. We found no convinc-

roraceratops in having a long lateral ridge on the surangular

ing characters to suggest the specimen is juvenile.

reaching the dentary, and from Yamaceratops in having a short retroarticular process.

Including A. yujingziensis, four species of basal ceratopsians have been discovered in the Mazongshan area, with Archaeo-

However, CAGS-IG-VD-003 possesses four autapomorphies

ceratops oshimai and Auroraceratops rugosus from the Gong-

that do not exist in A. oshimai: a laterally deflected rostral end

poquan Basin, and Psittacosaurus mazongshanensis from the

of the maxilla, resulting in a spoon-shaped premaxillary beak

Suanjingzi Basin (100 km north of Yujingzi Basin and east of

with premaxillary teeth located lateral to the line of the max-

the Gongpoquan Basin; Xu 1997) The widespread distribu-

illary teeth; striations on the enameled premaxillary teeth,

tion of basal ceratopsians in the Mazongshan area is compat-

perhaps combined with their ventrolateral orientation; ab-

ible with the late Early Cretaceous age designation for these

sence of a primary ridge on the maxillary teeth; and a horizon-

dinosaur-bearing beds.

tal shelf on the dentary teeth. Accordingly, we assign it to a new species: A. yujingziensis. Archaeoceratops oshimai is a very small animal. As recon-

Conclusions

structed by Dong and Azuma (1997), the type specimen mea-

CAGS-IG-VD-003 is assigned to a new and second species of

sures about 800 mm in length and 330 mm at the hips. Scaling

Archaeoceratops: A. yujingziensis. It differs from the type spe-

from data in You and Dodson (2003), the paratype would

cies, A. oshimai, in having a spoon-shaped premaxillary beak,

measure about 670 mm in length and 280 mm at the hip. A.

striations on the enameled premaxillary teeth, absence of a

66 you, tanoue, & dodson

primary ridge on the maxillary teeth, and a horizontal shelf on the dentary teeth. The occurrence of A. yujingziensis extends the paleogeographic distribution of the genus 100 km southeast, and is compatible with the late Early Cretaceous age designation for the dinosaur-bearing bed of the Yujingzi Basin in the Mazongshan area. Acknowledgments

We are grateful to Ji Qiang and Zhang Xing for supporting the project, Zhang Yu-Qing for preparing the specimen, and Xu Xing for arranging to access specimens at Institute of Vertebrate Paleontology and Paleoanthropology in Beijing. Review comments from David Eberth and James Kirkland improved the manuscript and are greatly appreciated. Funding was provided by the Ministry of Science and Technology of China (973 Project: 2006CB701405), the National Natural Science Foundation of China (40672007), and the Hundred Talents Project of Ministry of Land and Resources of China to You Hai-Lu. Kyo Tanoue was funded by Summer Research Stipends in Paleontology (University of Pennsylvania), School of Arts and Sciences Dissertation Research Fellowship (University of Pennsylvania), and Jurassic Foundation Research Grant. Peter Dodson thanks his chairman, Dr. Narayan Avadhani, for support. References Cited Dong, Z.-M., and Y. Azuma. 1997. On a primitive neoceratopsian from the Early Cretaceous of China. In Z.-M. Dong, ed., SinoJapanese Silk Road Dinosaur Expedition, pp. 68–89. Beijing: China Ocean Press. Li, D.-Q., C. Peng, H.-L. You, M. C. Lamanna, J. D. Harris, K. J. Lacovara, and J.-P. Zhang. 2007. A large therizinosauroid (Dinosauria: Theropoda) from the Early Cretaceous of northwestern China. Acta Geologica Sinica 81: 539–549. Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Marsh, O. C. 1890. Description of new Dinosaurian reptiles. American Journal of Science, Series 3, 39: 82–86. Seeley, H. G. 1888. On the classification of the fossil animals

commonly named Dinosauria. Proceedings of the Royal Society of London 43: 165–171. Sereno, P. C. 1986. Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). National Geographic Research 2: 234–256. Tang, F., Z.-X. Luo, Z.-H. Zhou, H.-L. You, J. A. Georgi, Z.-L. Tang, and X.-Z. Wang. 2001. Biostratigraphy and palaeoenvironment of the dinosaur-bearing sediments in Lower Cretaceous of Mazongshan area, Gansu Province, China. Cretaceous Research 22: 115–129. Xu, X. 1997. A new psittacosaur (Psittacosaurus mazongshanensis sp. nov.) from Mazongshan area, Gansu Province, China. In Z.-M. Dong, ed., Sino-Japanese Silk Road Dinosaur Expedition, pp. 48–67. Beijing: China Ocean Press. Xu, X., P. J. Makovicky, X. L. Wang, M. A. Norell, and H.-L. You. 2002. A ceratopsian dinosaur from China and the early evolution of Ceratopsia. Nature 416: 314–317. Xu, X., and M. A. Norell. 2006. Non-avian dinosaur fossils from the Lower Cretaceous Jehol Group of western Liaoning, China. Geological Journal 41: 419–437. You, H.-L. 2002. Mazongshan dinosaur assemblage from the late Early Cretaceous of northwest China. Ph.D. diss., University of Pennsylvania, Philadelphia. You, H.-L., and P. Dodson. 2003. Redescription of neoceratopsian dinosaur Archaeoceratops and early evolution of Neoceratopsia. Acta Palaeontologica Polonica 48: 261–272. ———. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press. You, H.-L., D.-Q. Li, Q. Ji, M. C. Lamanna, and P. Dodson. 2005. On a new genus of basal neoceratopsian dinosaur from the Early Cretaceous of Gansu Province, China. Acta Geologica Sinica 79: 593–597. You, H.-L., and Z.-X. Luo. 2008. Dinosaurs from the Lower Cretaceous Gongpoquan Basin in Jiuquan Area, Gansu Province, China. Acta Geologica Sinica 82: 139–144. You, H.-L., Z.-X. Luo, N. H. Shubin, L. M. Witmer, Z.-L. Tang, and F. Tang. 2003a. The earliest-known duck-billed dinosaur from deposits of late Early Cretaceous age in northwest China and hadrosaur evolution. Cretaceous Research 24: 347–355. You, H.-L., F. Tang, and Z.-X. Luo. 2003b. A new basal titanosaur (Dinosauria: Sauropoda) from the Early Cretaceous of China. Acta Geologica Sinica 77: 424–429.

A New Species of Archaeoceratops 67

4 A Redescription of the Montanoceratops cerorhynchus Holotype with a Review of Referred Material PETER J. MAKOVICKY

the basal neoceratopsian Montanoceratops cerorhyn-

of equivalent age and geographic distributions, and the

chus has historically been interpreted as a close relative

former view of these animals as a relict holdover from

of Ceratopsidae by some authors, based on several shared

older faunas can no longer be supported.

synapomorphies such as the presence of a nasal horn and a reduced laterotemporal fenestra. Only fragments of the skull were recovered with the holotype specimen,

Introduction

however, and were embedded in a plaster reconstruction

‘‘Leptoceratops’’ cerorhynchus (Brown and Schlaikjer 1942) was

that hindered detailed examination of these bones. Dis-

described based on a partial skeleton (AMNH 5464) compris-

mantling of this cranial reconstruction has led to re-

ing a few skull parts and large portions of the postcranium

identification of some skull parts, especially those pur-

from the St. Mary River Formation (Campanian) of Montana.

portedly exhibiting ceratopsid traits. The nasal horn is

This specimen was originally described as a new species of the

here re-identified as a jugal horn, and several previously

‘‘protoceratopsid’’ Leptoceratops. Shortly before its description,

undescribed or unrecognized cranial elements, such as a

the holotype material was assembled into a display mount

premaxilla, are also figured and described here. The pre-

with the cranial parts embedded in a large plaster skull re-

served cranial parts exhibit a suite of derived characters

construction (Brown and Schlaikjer 1942). Missing postcra-

shared with other members of Leptoceratopsidae, and

nial bones were apparently modeled on available parts of Lep-

phylogenetic analysis posits Montanoceratops cerorhyn-

toceratops gracilis (AMNH 5205, 5208), while missing cranial

chus as a leptoceratopsid more derived than Cerasinops,

bones seem to have been modeled on Protoceratops (e.g., pre-

and sister to a more exclusive clade of leptoceratopsids

maxillary teeth) or ceratopsids (e.g., enlarged, round naris).

with highly curved mandibles. A review of specimens re-

Sternberg (1951) observed that AMNH 5464 was distinct

ferred to Montanoceratops over the last decade finds weak

from the exquisite specimens of Leptoceratops gracilis (CMN

support for assignment to the genus for two of the speci-

8887–8889) that he had collected from the Maastrichtian

mens (MOR 542; AMNH 5244), but not in the case of a

Scollard Formation and that were far more complete than the

third specimen (RTMP 88.11.1). Recent discoveries

type material (Brown 1914). He therefore erected the new

indicate that leptoceratopsid diversity in Campanian-

genus Montanoceratops for the type of ‘’’Leptoceratops’’ cero-

Maastrichtian rocks of western North America ap-

rhynchus. The characters that Sternberg (1951) used to dis-

proaches that observed in centrosaurine ceratopsians

tinguish Montanoceratops from Leptoceratops are all cranial,

68

and therefore predicated on Brown’s identifications of these

cus on aspects of the anatomy that have been either misin-

partial elements prior to their incorporation into the skull

terpreted, overlooked, or not illustrated previously, and are

mount.

meant to supplement rather than supplant the excellent illustrations in Brown and Schlaikjer (1942).

Starting with Brown and Schlaikjer (1942), many authors have viewed Montanoceratops as being more derived than Lep-

Institutional Abbreviations. AMNH: American Museum of

toceratops and other known members of the paraphyletic fam-

Natural History, New York; CMN: Canadian Museum of Na-

ily Protoceratopsidae (Colbert 1948; Maryanska ´ and Osmól-

ture, Ottawa; FMNH: Field Museum of Natural History, Chi-

ska 1975; Ostrom 1966), a fact also reflected in several cladistic

cago; IGM: Institute of Geology, Ulan Bator; IVPP: Institute

analyses over the past 20 years, which have posited Montano-

of Vertebrate Paleontology and Paleoanthropology, Beijing;

ceratops cerorhynchus as the sister taxon to Ceratopsidae (Chin-

MOR: Museum of the Rockies, Bozeman; PIN: Paleontological

nery and Weishampel 1998; Sereno 1984, 1986, 2000). All

Institute of the Russian Academy of Sciences, Moscow; TCM:

characters cited in support of this hypothesis, such as pres-

Children’s Museum of Indianapolis, Indianapolis; TMP: Royal

ence of a nasal horn and an enlarged, circular naris, are identi-

Tyrrell Museum of Palaeontology, Drumheller.

cal to the distinguishing traits identified by Sternberg (1951),

Anatomical Abbreviations. 4–16: vertebral number; adb:

and therefore subject to the accuracy of Brown and Schlaik-

bulge on coronoid process base for adductor insertion; ann:

jer’s (1942) cranial reconstruction.

notch for tip of angular; as: articular depression for surangu-

Until recently, Montanoceratops cerorhynchus was only known

lar; cpn: coronoid process notch; cvr: caudoventral ridge of

from the holotype specimen. This situation has changed over

squamosal; d: dentary; epj: epijugal; ifs: interfrontal sutural

the last few years, starting with Chinnery and Weishampel’s

surface; itf: infratemporal fenestra; lma: lateral articulation for

(1998) description of a partial juvenile skeleton (MOR 542)

maxilla; mma: medial articulation for maxilla; nr: narial rim;

found close to the holotype locality, which they referred to

ns: neural spine; or: orbital rim; par: palatal ridge for articula-

Montanoceratops cerorhynchus. Another relatively complete

tion with maxilla; po: postorbital; pos: supratemporal shelf on

skeleton, RTMP 82.11.1, from the Crowsnest Pass area of

postorbital; pp: parapophysis; prd?: possible predentary frag-

Alberta was also referred to Montanoceratops sp. by Ryan and

ment; qj: quadratojugal; qja: articulation for quadratojugal;

Currie (1998), and this specimen was later used for scor-

qu: quadrate; san: surangular; sp: splenial; sq: squamosal; stf:

ing character information and briefly discussed by Sereno

supratemporal fossa margin on frontal; vp: ventral process of

(2000). Finally, Makovicky (2001) referred an isolated brain-

coronoid.

case (AMNH 5244) from the Horseshoe Canyon Formation to Montanoceratops, based on comparisons with the MOR 542 specimen. The recent discovery and description of several

Systematic Paleontology

basal neoceratopsian taxa from Asia (Makovicky and Norell

Ornithischia Seeley 1888

2006; Xu et al. 2002; You et al. 2005) and North America

Ceratopsia Marsh 1890

(Chinnery 2004; Chinnery and Horner 2007) has significantly

Leptoceratopsidae Nopcsa 1923

shifted our understanding of character distributions among

Montanoceratops cerorhynchus Brown and Schlaikjer 1942

basal neoceratopsians, however, and reevaluation of the referred material is warranted in light of this new information.

Material. AMNH 5464, holotype of Montanoceratops cero-

Recent dismantling of the holotype specimen allows for

rhynchus (Brown and Schlaikjer 1942). The holotype includes

a reappraisal of some of Brown and Schlaikjer’s (1942) ele-

parts of many cranial elements, including the right premax-

mental identifications as well as reconsideration of character

illa, sections of the fused right and left jugals and epijugals,

evidence for the phylogenetic position of Montanoceratops

the right quadratojugal, parts of the right postorbital, both

cerorhynchus within Neoceratopsia. Phylogenetic analyses em-

squamosals, fragments of both frontals, parts of the left den-

ploying observations on the holotype material after it was

tary, surangular and coronoid, and distal fragments of both

dismantled have recovered a significantly more basal posi-

splenials. Brown and Schlaikjer (1942) also note the presence

tion for Montanoceratops as a close relative of Leptoceratops

of a section of the maxilla, although I was not able to identify

(Makovicky 2001, 2002; Makovicky and Norell 2006; Xu et al.

this piece. The postcranial skeleton of AMNH 5464 is com-

2002). In light of these re-identifications of elements, the

prised of most of the vertebral column except for the distal

holotype is redescribed here with a focus on previously mis-

caudal region. A full complement of ribs and a number of

identified and unidentified elements. I also present observa-

chevrons were also recovered. Most of the pectoral girdle ex-

tions on specimens referred to Montanoceratops in recent liter-

cept for the right coracoid and the forelimbs are missing, but

ature, and evaluate the validity of these referrals given new

representative elements from at least one side of the body are

information on the holotype of Montanoceratops and recently

known for the pelvic girdle and hindlimbs. The only missing

described leptoceratopsid taxa. The illustrations herein fo-

hindlimb elements are the distal tarsals and metatarsals.

A Redescription of the Montanoceratops cerorhynchus Holotype 69

Revised Diagnosis. Montanoceratops cerorhynchus is a lepto-

CMN 8888 and 8889 in fact have elliptical narial openings just

ceratopsid neoceratopsian characterized by two definitive and

like other basal neoceratopsians. Only the caudal end of the

three possible autapomorphies (see analysis below). A low,

ventral premaxillary margin is preserved in AMNH 5464. It is

caudoventrally bowed ridge along the caudal margin of the

thick and rugose as in Xuanhuaceratops (Zhao et al. 2006), Lep-

squamosal and a midline depression on the occipital face

toceratops (CMN 8889) and Udanoceratops (PIN 3907/11). The

of the supraoccipital (Makovicky 2001) are unique to Mon-

buccal margin curves gently upward as it approaches the con-

tanoceratops. An acute coronoid process notch (Chinnery

tact with the maxilla, suggesting that a shallow notch was

2004) as opposed to the wide one as observed in Leptoceratops,

present between the buccal margin of the maxilla and pre-

Prenoceratops, and Udanoceratops may also be an autapomor-

maxilla as in Leptoceratops (CMN 8887, 8888), in which the

phy of this taxon, but the lack of a notch in outgroups to these

premaxillary buccal margin is convex. Due to breakage the

four taxa complicates this. Montanoceratops is also distin-

absence of premaxillary teeth cannot be completely estab-

guished by having hyperelongate mid-caudal neural spines

lished, but there is no trace of alveoli in the preserved sub-

that are more than four times taller than their respective cen-

narial body of the premaxilla. In all neoceratopsian taxa

tra (Ostrom 1978). The mid-caudal spine height in the closely

with premaxillary dentition, the teeth are located close to the

related Cerasinops is unknown, however, which could affect

caudal end of the premaxillary body, which is the region pre-

the utility of this diagnostic trait in the future. A straight ven-

served in AMNH 5464. Therefore, it is almost certain that pre-

tral margin on the dentary is an autapomorphy of Montano-

maxillary teeth were absent in Montanoceratops as in Prenocera-

ceratops under some optimizations (i.e., ACCTRAN), but is

tops, Leptoceratops, and Udanoceratops, but not Cerasinops. The

complicated by presence of this state in a number of lepto-

medial face of the premaxilla bears a pair of horizontal ridges

ceratopsid outgroups. Montanoceratops can furthermore be dis-

separated by a shallow sulcus for articulation with the rostral

tinguished from other leptoceratopsids based on the following

prong of the maxilla. As in Protoceratops (FMNH PR 14064), a

character distributions: it differs from Cerasinops in lacking

small vascular foramen exits below the lower of the two pala-

premaxillary teeth (a derived trait) and a frontoparietal de-

tal ridges.

pression (a plesiomorphic trait); and it differs from Leptocera-

Prefrontal. Only a small fragment of the left prefrontal is

tops in lacking a flat shelf on the postorbital rim of the supra-

preserved, exhibiting the orbital rim and the attachment area

temporal fenestra (plesiomorphic trait in Montanoceratops).

for the palpebral. The orbital rim above the palpebral fossa is

Locality and Horizon. The specimen was collected from the

marked by rugose ornamentation similar to that on the orbital

St. Mary River Formation (latest Campanian—Early Maas-

margin of the postorbital. Chinnery and Weishampel (1998)

trichtian), three miles west of Buffalo Lake, Glacier County,

reported that the palpebral attachment area was rugose in

Montana, in 1916 by Barnum Brown and Peter Kaisen.

MOR 542, and slight rugosities are observed in other basal neoceratopsians such as Yamaceratops (IGM 100/1315), but

DESCRIPTION OF THE HOLOTYPE SPECIMEN

the degree observed in AMNH 5464 is unusual. The difference between MOR 542 and AMNH 5464 suggests that the development of circumorbital rugosity may increase with age.

Brown and Schlaikjer (1942) provided a detailed description

Jugal. The element misidentified as the right nasal by Brown

of the holotype specimen, most of which is accurate. Here I

and Schlaikjer (1942) is actually the left jugal (Fig. 4.1C, D),

focus on redescribing and illustrating those elements misiden-

which appears to be short, deep, and triangular, like those of

tified in the original description, and discussing anatomical

other leptoceratopsids including Leptoceratops (NMC 8889),

features relevant to diagnosing Montanoceratops and neocera-

Prenoceratops (Chinnery 2004), and Udanoceratops (PIN 3907/

topsian phylogeny in light of recent discoveries.

11). The structure that Brown and Schlaikjer (1942) inter-

Premaxilla. A broken right premaxilla (Fig. 4.1A, B) was ap-

preted as a nasal horncore is actually the dorsal part of the

parently not recognized by Brown and Schlaikjer (1942). It is

epijugal, which is partially co-ossified with the caudodorsal

missing the internarial process and the rostral and much of

edge of the jugal crest as in other basal neoceratopsians more

the ventral margins are also missing. The base of the subnarial

derived than Liaoceratops (Makovicky and Norell 2006). A

process is preserved and reveals a relatively straight, long, and

small section of the ventral rim of the jugal is preserved just

oblique lower narial rim, indicating that the naris was ellip-

in front of the jugal crest and was originally interpreted

tical as in all other basal neoceratopsians. You and Dodson

as the dorsal rim of an enlarged, rounded naris by Brown

(2003, 2004) have claimed that the naris of Leptoceratops is

and Schlaikjer (1942). It curves caudoventrally, suggesting a

circular and used this character to unite leptoceratopsids with

slightly recurved ventral tip to the jugal horn as is observed

ceratopids, but their inference appears to be based on an in-

in Prenoceratops (Chinnery 2004), Cerasinops (Chinnery and

accurate reconstruction of CMN 8888 (Russell 1970). Both

Horner 2007), and Udanoceratops (PIN 3907/11). The broken

70 makovicky

reception of the medial jugal process of the maxilla. These two surfaces are almost directly in line with one another, indicating a planar caudal end to the maxilla with little mediolateral separation between jugal articulations, as is indeed observed in the maxilla of MOR 542. This contrasts with the condition in Prenoceratops and Udanoceratops, in which the lateral jugal process of the maxilla is projected far lateral to the tooth row and the medial jugal process (Chinnery 2004). The right jugal (Fig. 4.1E–G) is less complete than the left, and its lighter hue and weathered appearance indicate that it was probably collected on the surface. It is more extensively co-ossified with the epijugal and is also fused to the right quadratojugal. It is likely that the difference in degree of fusion between left and right sides contributed to Brown and Schlaikjer’s (1942) misidentification of the left element. Only the caudodorsal edge of the bone between the epijugal and quadratojugal is preserved. Epijugal. The right epijugal is more completely preserved than the left. It is a massive, crescentic element, triangular in coronal section, and it wraps along the full length of the jugal crest to form the jugal horn. The apex of the epijugal is better preserved on the left partial element and the jugal horn was shaped as an elongate, caudodorsal ridge as in Yamaceratops and Leptoceratops, rather than being pointed as in many ceratopsids and juvenile to subadult specimens of Protoceratops (AMNH 6429, 6466). The epijugal is rugose and pitted on both the rostral and caudal faces, especially near its preserved apex, and this ornamentation is indicative of a keratinous cover in life. The suture between the jugal and epijugal can still be traced along its ventral half on the right side and remains Facial elements of Montanoceratops cerorhynchus (AMNH 5464). Right premaxilla in (A) lateral and (B) medial views; left jugal-epijugal complex in (C) lateral and (D) medial views; right jugal-epijugal-quadratojugal complex in (E) lateral, (F) medial, and (G) oblique caudomedial views. Scale bars are 2 cm.

FIGURE 4.1.

open in this region on the broken left side, but it is completely obliterated along the dorsal section of the lateral face on both sides, suggesting that fusion progressed from dorsal to ventral along the suture. Quadratojugal. The right quadratojugal (Fig. 4.1E–G) adheres to the medial face of the jugal and is also fused to the caudoventral edge of the epijugal. It is a tall, mediolaterally

left epijugal reveals a groove coursing along the caudal edge of

flattened element that is wider ventrally than dorsally and

the jugal, which corresponds to the neurovascular trace iden-

is triangular in coronal section with the apex directed ros-

tified by Chinnery (2004) on the jugal of Prenoceratops.

trally. This morphology compares well to the quadratojugal of

The medial face of the left jugal bears a large but narrow

Udanoceratops (Kurzanov 1992) and differs from the medio-

articular surface for the quadratojugal, which extends along

laterally wide, wedge-shaped quadratojugal of Protoceratops.

the caudodorsal edge of the element adjacent to the epijugal.

The shape of the right quadratojugal conforms exactly to the

This articular surface is wide ventrally, but narrows as it ex-

quadratojugal sutural surface preserved on the left jugal. Dis-

tends dorsally along the border of the laterotemporal fenestra

tally, the quadratojugal extends a short distance below the

and terminates at the midlevel of the preserved section of the

end of the epijugal as in Cerasinops (Chinnery and Horner

epijugal. Two articular surfaces for reception of the maxilla are

2007). The medial surface of the massive ventral end is rugose

present at the rostral end of the element. At the rostral tip of

and covered with irregular ridges and grooves for sutural con-

the bone is a flat smooth surface inset from the external sur-

tact with the quadrate. The rostroventral edge feathers to a

face for reception of the lateral jugal process of the maxilla.

thin edge, and there is no rostral prong of the quadratojugal

Just ventral and medial to this sutural surface is a deep slot for

extending along the ventral edge of the jugal as occurs in the

A Redescription of the Montanoceratops cerorhynchus Holotype 71

Makovicky (2001) referred to Montanoceratops, in spite of the fact that all other neurocranial elements are solidly fused and have obliterated sutures. A similar pattern of cohesion between neurocranial elements is observed in the smaller and presumably much younger MOR 542, in which all elements save the orbitosphenoid, frontals and parietal appear to be conjoined though not fully fused. Postorbital. Two parts of the right postorbital (Fig. 4.2C, D) are preserved as indicated in Brown and Schlaikjer (1942: fig. 1). The larger more dorsal fragment forms a thick caudal wall to the orbit, and has a pronounced rugosity laterally above the caudodorsal corner of the orbit. A wide, rounded shelf borders the supratemporal fenestra medially as in other leptoceratopsids (Fig. 4.2D), but is not as developed as the wide flat surface perpendicular to the lateral face of the postorbital in Leptoceratops (CMN 8889). The postorbital of other neoceratopsians lacks a wide supratemporal rim in dorsal view, being formed as a thin dorsoventral strap instead. The medial face of the postorbital bears a shallow pit for the postorbital process of the laterosphenoid, which appears proportionately smaller than in the holotype of Leptoceratops (AMNH 5205). A short section of the tip of the rostral process of the squamosal is preserved in articulation with a notch on the caudal section of the postorbital (Fig. 4.2C). Ventrally the tip of the jugal is preserved embedded in a notch in the larger Caudal skull bones of Montanoceratops cerorhynchus (AMNH 5464). (A) Left squamosal in lateral view; (B) right squamosal in lateral view; right postorbital in (C) lateral and (D) dorsal views; right frontal in (E) medial and (F) dorsal views. Scale bars are 2 cm.

FIGURE 4.2.

postorbital fragment, unlike in Leptoceratops and Protoceratops in which the jugal underlies the ventral process of the postorbital. The second, smaller fragment represents a more ventral section of the orbital rim (Fig. 4.2C). It continues the mediolaterally wide postorbital section of the orbit, and a piece of the jugal remains sutured to it. The wide orbital shelf attenuates rather abruptly ventrally below which the postor-

quadratojugal of Asian protoceratopsids (Makovicky 2002). Frontals. The preserved frontal fragments (Fig. 4.2E, F) are

bital would have continued as a short process on the external face of the jugal.

relatively massive, but differ from those of Cerasinops in the

Squamosal. Both squamosals are represented, although the

absence of a pronounced frontoparietal depression (Brown

left is far more complete than the right (Fig. 4.2A, B). The

and Schlaikjer 1942; Chinnery and Horner 2007). The frontals

squamosal is rostrocaudally short, but dorsoventrally deep as

appear to be thickened just dorsal to the exit of the olfactory

in Leptoceratops (CMN 8888, 8889). The dorsal edge curves

nerve (Fig. 4.2E). Brown and Schlaikjer (1942) noted the sa-

caudodorsally in a sharp arc that resembles the squamosal

lient features of the frontal, and correct orientation of the

of Cerasinops (Chinnery and Horner 2007) more closely than

preserved laterosphenoid and postorbital sutures indicates a

that of either Leptoceratops or Prenoceratops (Chinnery 2004).

relatively wide and dorsally concave skull roof (Fig. 4.2E, F).

There is no postquadratic process as observed in coronosaurs.

The margin of the supratemporal fenestra is formed as a slight,

The caudal surface of the squamosal forms an acute corner at

laterally concave ridge on the dorsal surface of the frontal

the caudal end of the supratemporal fenestra. In lateral aspect,

(Fig. 4.2F), which is markedly different from the deep fossa on

a low but distinct ridge is expressed along the edge of the

the caudal surface of the frontal in Protoceratops and Bagacera-

element. This ridge, which is unique to the holotype speci-

tops. Leptoceratops (CMN 8889) has a similarly wide skull roof

men of Montanoceratops, is gently bowed caudally and defines

with a shallowly incised supratemporal fossa.

a slight depression on the lateral surface of the squamosal. The

The largest right frontal fragment reveals open sutures for

thin bone rostral to the ridge forms a low bulge on the left

the left frontal (Fig. 4.2E), the postorbital and the parietal.

squamosal, but this is not observed on the right side, and

It is noteworthy that the frontal sutures remain open on

appears to be a crushing artifact.

a large and presumably adult braincase (AMNH 5244) that

72 makovicky

The short, deep postorbital process of the squamosal is

surrounding the splenial symphyseal process is observed in Yamaceratops (IGM 100/1315). The caudal end of the mandibular ramus is incised by a small notch, which marks the end of the ventral edge of the dentary as in MOR 542 and Leptoceratops (AMNH 5205). This appears to be corroborated by a piece of bone that lies between the dentary and the splenial in this region (Fig. 4.3B) and probably represents the rostral terminus of the predentary. If this bone fragment is correctly identified, it would imply a relatively short mandible as in other leptoceratopsids, although it is less massive than in either Prenoceratops, Leptoceratops (AMNH 5025, CMN 8889), or Udanoceratops (Kurzanov 1992, PIN 3907/11). The preserved portion of the coronoid process (Fig. 4.4A) is relatively low and robust, although the distal tip is missing. The external surface is striated and rugose where the mandibular adductors would have inserted. The caudal margin is deeply indented by a deep, narrow notch (’coronoid process notch’ of Chinnery 2004) extending along the medial surface of the coronoid process, a feature also present in MOR 542. By contrast the coronoid process notch is more apical in position and open in Leptoceratops (AMNH 5205), Prenoceratops

Left dentary and splenial of Montanoceratops cerorhynchus (AMNH 5464) in (A) lateral, (B) medial, and (C) ventral views. Scale bar is 2 cm.

FIGURE 4.3.

(Chinnery 2004), and Udanoceratops (Kurzanov 1992). A large, rugose bulge or eminence, probably for insertion of an adductor slip, lies just below the distal end of the coronoid process as in Leptoceratops (AMNH 5205; CMN 8889) and Prenoceratops

deeply bifurcated as in other leptoceratopsid taxa and basal

(Chinnery 2004).

neoceratopsians such as Liaoceratops (Xu et al. 2002). Taken

Splenial. Parts of the ventral edge of the left splenial adhere

together with the dimensions of the jugal and postorbital, the

to the ventral edge of the left dentary (Fig. 4.3B, C), and an-

proportions of the squamosal indicate a relatively small in-

other large, flat piece of bone compares favorably with the

fratemporal fenestra that is taller than wide. Although several

splenial of MOR 542 and could be part of the right splenial.

authors viewed the reconstructed dimensions of the infratem-

The ventral edge of the splenial is rugose along the border of

poral fenestra of Montanoceratops as a trait uniting it with cera-

the left mandible (Fig. 4.3B). Rostral to the kink in the ventral

topsids, a proportionately small fenestra that is taller than

mandibular border, the splenial is slightly mediolaterally ex-

wide is present in a number of basal neoceratopsians includ-

panded where it adheres to the outwardly curving dentary

ing Liaoceratops (Xu et al. 2002), Yamaceratops (Makovicky and

(Fig. 4.3C). At the rostral end of this expansion, the tip of the

Norell 2006), Leptoceratops (CMN 8888), and Auroraceratops

abbreviated symphyseal process is observed. Unlike the condi-

(You et al. 2005).

tion observed in most other ceratopsians, in which the sym-

Dentary. The preserved elements of the left dentary (Fig.

physeal process is clearly delimited from the main body of the

4.3A–C) were briefly reviewed by Brown and Schlaikjer (1942).

splenial as a long, thin rod, it appears to be short, deep, and

Only the distal, symphyseal region and the coronoid process

tapering in AMNH 5464, and lacks the mediolateral expan-

of the left dentary are preserved in AMNH 5464. The elements

sion observed in Leptoceratops (NMC 8889). Surangular. The rostral tip of the surangular is preserved

display the same weathered appearance as the larger jugal-

in articulation with the coronoid process of the dentary. As

epijugal and appear to have been collected on the surface. The dentigerous portion of the dentary has a straight ven-

in Udanoceratops and Leptoceratops, the surangular bears a

tral margin in lateral view (Fig. 4.3A), which ends at a slight

pointed rostral prong that fits in between the coronoid pro-

rugose expansion in ventral view. Rostral to the straight lower

cess and the coronoid bone (Fig. 4.4A). The main part of the

edge, the dentary ventral margin is kinked to form a small

surangular is not preserved, but the surangular of a referred

‘‘chin,’’ and the symphyseal portion of the dentary ascends at

specimen (MOR 542) displays a low lateral shelf, which is less

a low angle as in MOR 542. In ventral view the lower border of

prominent than that of Protoceratops. Surprisingly, such a shelf

the dentary curves laterally and then turns medially again to

is absent in Leptoceratops (CMN 8889), in which the surangu-

form a shallow depression (Fig. 4.3C) for the splenial sym-

lar is laterally bowed, and in Prenoceratops which has a flat

physeal process. A similar curvature in the rostroventral rim

external surface on the surangular (Chinnery 2004). By con-

A Redescription of the Montanoceratops cerorhynchus Holotype 73

(CMN 8889). Laterally, the coronoid bears a dorsoventrally concave depression near the caudal border for articulation with the surangular (Fig. 4.4B). Axial Skeleton. The syncervical of AMNH 5464 is composed of three (Brown and Schlaikjer 1942; Tsuihiji and Makovicky 2007), not four (Lull 1933), vertebrae, and the high degree of fusion between the atlantal, axial, and third cervical elements suggests that AMNH 5464 was a mature animal. In Leptoceratops fusion between components progresses with ontogeny and full fusion is only seen in old individuals (Sternberg 1951; NMC 8889). The hypocentrum (atlas intercentrum) forms the majority of the cup for the occipital condyle, and is fused dorsally to the odontoid (atlas centrum) and caudally to the coossified axis intercentrum. The neurapophyses are fused with both the atlas intercentrum and odontoid proximally, and with axial neural arch distally. Confusion of the number of elements in the ceratopsian atlas can probably be attributed to the failure to recognize the expansion and co-ossification of the odontoid with the atlas intercentrum and the progressive migration of the atlantal neurapophyses (Tsuihiji and Makovicky 2007). The axis centrum is keeled ventrally, and fused to both the atlas cranially and the centrum of the third cervical caudally. The diapophyses and parapophyses are well developed, and the axial rib appears to have been fused to the axis, although it is not fully preserved. The axial neural spine is enlarged and triangular in outline. It is fused to the prezygapophyses and neural spine base of the third cervical, but the latter structure is free for most of its length and is lower than the axial neural Montanoceratops cerorhynchus (AMNH 5464). (A) left mandibular coronoid process; (B) left coronoid bone. Both in lateral view. Scale bar is 2 cm.

FIGURE 4.4.

spine. The centrum of the third cervical is sharply keeled as are the remainder of the cervicals, and it is taller than wide in caudal view. The remainder of the cervical vertebrae and the cranial dorsals (Fig. 4.5) are slightly keeled as in Psittacosaurus, Lepto-

trast, the holotype of Udanoceratops tschizhovi (PIN 3907/11)

ceratops, and other basal ceratopsians, but differ from the con-

bears a well-developed lateral ridge. Development of the sur-

dition in Protoceratops and ceratopsids (Brown and Schlaikjer

angular ridge varies with ontogenetic age in Protoceratops

1942). Chinnery and Weishampel (1998) described centra of

(pers. obs.).

the fourth and fifth cervicals of MOR 542 as markedly shorter

Prearticular. A splint of bone squeezed between the dentary

than those of more caudal elements, a condition now also

and the splenial may represent the rostral process of the pre-

known in Cerasinops (Chinnery and Horner 2007). Although

articular. It appears to reach far forward as in MOR 542, Baga-

they (Chinnery and Weishampel 1998) claimed this trait to be

ceratops, and ceratopsids. It is thicker proximally than distally,

present in the holotype, this does not seem to be the case (Fig.

where it becomes slightly expanded dorsoventrally.

4.5). Parallax in the photo published by Brown and Schlaikjer

Coronoid. The left coronoid is partially preserved (Fig. 4.4B),

(1942) appears to give the impression that the cranial cervicals

missing only its ventral extension and the dorsal edge. As in

are short in this specimen. The neural spines are relatively low

many basal neoceratopsians, it covered most of the internal

and thin on the cervicals as in MOR 542, but become longer

surface of the coronoid process. This differs from the cera-

and more robust in the direction of the dorsal series. The

topsid condition where the coronoid is restricted to the caudal

transverse processes of the first two free vertebrae point

half of the coronoid process. A part of the coronoid may have

slightly ventral to horizontal and become dorsolaterally re-

been exposed in lateral view, dorsal to the surangular and cau-

oriented throughout the cervical series, a character that may

dal to the coronoid process of the dentary as in Leptoceratops

distinguish Montanoceratops from other basal neoceratop-

74 makovicky

FIGURE 4.5.

Presacral vertebrae (4–15) of Montanoceratops cerorhynchus (AMNH 5464) in right lateral view. Scale bar is 2 cm.

sians. The parapophyses display a marked dorsal shift in posi-

A number of cervical and most thoracic ribs are preserved.

tion from the region of the neurocentral suture to the base of

The cervical ribs are relatively short and roughly Y-shaped as

the transverse process between the ninth and tenth presacrals,

in other neoceratopsians. The cranial thoracic ribs have the

respectively (Fig. 4.5). The transverse processes are progres-

capitula and tubercula set perpendicular to each other, but in

sively longer up to the 15th presacral, and then decrease again

more caudal ribs the capitular process becomes shorter and

in length as they approach the sacrum. The postzygapophyses

the two facets move closer to each other. The last pair of pre-

are separated throughout the dorsal column unlike in cera-

sacral ribs is single headed. Coracoid. Although Brown and Schlaikjer (1942) stated that

topsids, where the postzygapophyses are fused along the mid-

they recovered no pectoral girdle elements, a poorly preserved

line in the caudal trunk vertebrae. The sacrum consists of eight vertebrae that have fused zyg-

left coracoid is present. Most of the border of the element is

apophyses and centra, although contacts between centra are

lost, but the glenoid region is preserved. The elliptical cora-

still evident in the last three sacrals. The first four sacrals have

coid foramen pierces the body of the coracoid rostroventral to

neural spines in contact and possibly partially fused although

the glenoid.

no single spine is completely preserved. The second through

Pelvis. The pelvic girdle is comprised of both ilia, the left

fifth sacral ribs fuse distally to form a sacricostal yoke that

pubis, and both ischia, of which the right is incomplete. The

braces the acetabular portion of the ilium and the pubic and

ilia are similar to those of Protoceratops except for being pro-

ischiadic peduncles.

portionately slightly lower in depth. Their cranial ends are

The tail is not complete and only 15 vertebrae are pre-

slightly everted and each bears a shelf ventrally for the femoral

served, but they are representative of all but the distalmost

abductors. The postacetabular portion of the ilium is rela-

part of the tail. The proximal caudals have centra that are

tively flat and vertical. The pubic peduncle is relatively slender

circular in cross section, and bear moderately tall neural

and points cranioventrally. It is braced medially by the ribs of

spines. The first free caudal has very short caudal ribs and a

the second and third sacral vertebrae. The ischiadic peduncle

broad neural spine, but in the following vertebrae the ribs are

is much stouter than the pubic peduncle and the lateral sur-

longer and the neural spines become long and rod-like. The

face is formed as a massive, convex antitrochanter as in other

neural spines also become nearly vertical in the middle of

neoceratopsians.

the tail where they are tallest. The distalmost centra may

The pubis has a short finger-like prepubic process similar to

not have borne arches as in Leptoceratops (Sternberg 1951).

that of Protoceratops although more slender in its propor-

Extreme elongation of the midtail neural spines is more

tions. Only the base of the postpubic process is preserved. It is

pronounced in Montanoceratops than in either Leptoceratops

oval in cross section and comparison with MOR 542 (Chin-

(Sternberg 1951), or Udanoceratops (PIN 3907/11), but may be

nery and Weishampel 1998) indicates that the whole length of

comparable to the neural spines of Protoceratops (Brown and

the process was oval to circular in cross section unlike the

Schlaikjer 1942). Pronounced chevron facets are present distal

mediolaterally flattened postpubic process of Protoceratops

to the first few caudals.

and ceratopsids.

A Redescription of the Montanoceratops cerorhynchus Holotype 75

Hindlimbs. Both femora are preserved albeit crushed, and

noid process notch is placed closer to the apex of the coronoid

the left is missing the medial distal condyle. The femoral head

process, and is also more open with nearly perpendicular

is inclined dorsomedially and the greater trochanter is low,

edges. A distinct notch is not present in Yamaceratops, Archae-

unlike the tall trochanter of ceratopsids (Hatcher et al. 1907).

oceratops, and Liaoceratops, and its presence is a further syn-

The fourth trochanter is relatively large and thickened with a

apomorphy of leptoceratopsids. Despite the little overlap in

pendant lower edge, and a rugose and wide caudal edge. In

cranial material between the holotype and MOR 542, two

contrast to this, the fourth trochanter of ceratopsids and pro-

morphological traits of the lower jaw that are unique among

toceratopsids is more reduced and does not bear a pendant

known leptoceratopsids support referral of the latter speci-

ventral process.

men to Montanoceratops.

The right tibia, fibula, and astragalus are preserved. The ele-

In the postcranium, Chinnery and Weishampel (1998)

ments are similar to those of Leptoceratops and more massive

identified the short length of the fourth and fifth cervical ver-

than the corresponding elements in Protoceratops. The astraga-

tebrae and the everted dorsal rim of the ilium as traits that

lus is partially co-ossified with the tibia suggesting a mature

distinguish AMNH 5464 and MOR 542 from other ceratopsian

age for this individual.

taxa. As discussed above, everted ilia are more widely dis-

Several pedal elements are preserved comprising one subter-

tributed including in Leptoceratops (NMC 8887, 8888), Ar-

minal phalanx and three unguals. The subterminal phalanx

chaeoceratops (IVPP v 11114), and Protoceratops (AMNH 6471).

has well-developed ginglymoid articular surfaces, and is prim-

Abbreviated centra on the fourth and fifth cervicals are also

itive relative to the short flat phalanges of ceratopsids. All

present in Cerasinops. Furthermore, the fourth and fifth cervi-

three unguals are long and claw-like and display a moderate

cals of AMNH 5464 are not abbreviated as previously consid-

curvature in lateral view. A heavily abraded ceratopsian pha-

ered (Chinnery and Weishampel 1998), so there may be some

lanx is present in the surface material collected alongside the

ontogenetic variation in this character.

holotype. It is larger than any of the Montanoceratops material

AMNH 5244. Makovicky (2001) referred an isolated cera-

and has markedly different proportions, being both wider and

topsian braincase (AMNH 5244) from the Horseshoe Canyon

taller than long, and having almost flat, non-ginglymoid af-

Formation near Drumheller to Montanoceratops based on de-

finities suggesting ceratopsid or ankylosaur affinities.

rived characters shared with the braincase of MOR 542. The braincases of AMNH 5244 and MOR 542 overlap almost com-

REMARKS ON SPECIMENS REFERRED TO MONTANOCERATOPS

pletely in terms of the preserved bones and are virtually identical in most respects, although the paroccipital processes are proportionately deeper in the larger specimen. Chief

MOR 542. Chinnery and Weishampel (1998) referred a well-

among the diagnostic characters uniting the two specimens

preserved partial, associated skeleton (MOR 542) from the

was the presence of a small teardrop-shaped depression on the

St. Mary River Formation to Montanoceratops based on a hand-

supraoccipital midline just dorsal to the foramen magnum.

ful of traits and general anatomical similarity in overlapping

This feature is subtler in the smaller and younger MOR 542,

regions of the skeleton, combined with geographic and strati-

mainly because the midline supraoccipital ridge that bifur-

graphic considerations. According to Chinnery and Weis-

cates around and defines the depression in AMNH 5244, is far

hampel (1998), MOR 542 was very likely discovered extremely

less developed in the juvenile braincase. A supraoccipital ridge

close to the holotype locality. Relatively little overlap is evi-

and depression is absent in Leptoceratops (CMN 8888), while

dent between the holotype and referred specimens in the cra-

the prominent ridge in the holotype of Cerasinsops (Chinnery

nium, however, and comparisons are restricted. A character

and Horner 2007) is undivided ventrally, thus precluding de-

that unites the two Montanoceratops specimens is possession of

velopment of a depression adjacent to the foramen magnum.

a straight dentary margin; the dentary is highly curved in

Other similarities that Makovicky (2001) noted between the

Leptoceratops, Prenoceratops, Cerasinops, and Udanoceratops.

two specimens such as caudoventrally curved basipterygoid

A straight dentary margin is thus unique to Montanocera-

processes are also now known to be present in Prenoceratops

tops, among leptoceratopsids, although its status as either

(Chinnery 2004), Cerasinops (Chinnery and Horner 2007), and

an autapomorphy or as a primitive state shared with non-

possibly in Leptoceratops (CMN 8888). Dissimilarities between

leptoceratopsid taxa remains ambiguous in both the phylo-

the specimens such as absence of an ossified orbitosphenoid,

genetic results of Chinnery and Horner (2007) and those dis-

and more slender paroccipital processes with a less developed

cussed herein.

articulation for the pterygoid wing of the quadrate in MOR

Another, previously unrecognized similarity between these

542 probably reflect its younger age (Makovicky 2001).

two specimens is the constricted nature of the coronoid pro-

TMP 82.11.1. Ryan and Currie (1998) listed a skeleton from

cess notch and its position along the caudal edge of the coro-

the Crowsnest Pass region of Alberta as cf. Montanoceratops,

noid process. In both Leptoceratops and Prenoceratops, the coro-

but did not provide details for this referral. In a subsequent

76 makovicky

FIGURE 4.6.

Plaster cast of left temporal region of TMP 82.11.1 in (A) lateral and (B) dorsal views. Note lack of a squamosal ridge compared to Fig. 4.2A, and the wide supratemporal fossa in (B). Scale bar is 2 cm.

phylogenetic analysis, Sereno (2000) cited this specimen as

juveniles. Nevertheless, both the postorbital and squamosal of

displaying a fenestrated frill supporting a phylogenetic posi-

TMP 82.11.1 compare favorably with the corresponding ele-

tion for Montanoceratops as a sister taxon to ceratopsids. TMP

ments in Prenoceratops (Chinnery 2004). Unlike Montanocera-

82.11.1 preserves a virtually complete postcranium, but only

tops and Cerasinops, which have a relatively steeply upturned

the left postorbital region of the skull is preserved, comprising

dorsal squamosal margin, the dorsal margin appears only

the left postorbital, squamosal, a small, lateral section of the

slightly inclined in both Prenoceratops (TCM 2001.96.4; Chin-

parietal frill, and the quadrate head (Fig. 4.6A, B). The dorsal

nery 2004) and Leptoceratops (CMN 8888). In summary, TMP

surface of the postorbital is recessed and forms part of the

82.11.1 represents an indeterminate leptoceratopsid, which

supratemporal fossa as in all leptoceratopsids, but differs from

cannot be referred to Cerasinops, Montanoceratops, or Lepto-

that of Leptoceratops (NMC 8889) in that this shelf is rather

ceratops, but which may have affinities with Prenoceratops.

narrow. The squamosal is typical of leptoceratopsids, being roughly triangular with a deep and bifurcated rostral pro-

Discussion

cess and no postquadratic process. It lacks the diagnostic ridge that bounds the caudal edge of the squamosal in AMNH 5464,

While there has historically been little question regarding the

thus precluding reference of TMP 82.11.1 to Montanoceratops.

validity of M. cerorhynchus as a distinct taxon, its diagnosis has

The preserved section of the parietal frill does not extend far

changed significantly in response to correct re-interpretation

beyond the caudal end of the squamosal and is solid. Not

of its anatomy and new discoveries. AMNH 5464 was prepared

enough of it is preserved to preclude the possibility of a small,

and exhibited as a species of Leptoceratops in 1935, but it was

more medially placed fenestra, but it is different from the

not formally described until several years later (Brown and

holotype of Cerasinops (Chinnery and Horner 2007), in which

Schlaikjer 1942). During the preparation of their descrip-

the fenestrae are broad and bordered by only a thin band of

tion, Brown and Schlaikjer (1942) reinterpreted a number of

bone along the caudal rim of the frill.

their original identifications of the cranial elements, which

The mid-caudal vertebrae bear tall, vertical neural spines as

prompted them to design a second cranial reconstruction and

in other leptoceratopsids. Ostrom (1978) used the larger ratio

also formed the basis of their taxonomic diagnosis. Unfortu-

of neural spine height to central height in Montanoceratops

nately a figure of the first mount is not available, so I cannot

to distinguish it from other leptoceratopsids. Unfortunately,

determine what their original interpretations of the misiden-

central height cannot be determined in TMP 82.11.1 for pres-

tified elements may have been. Montanoceratops cerorhynchus

ervational reasons. Chinnery and Weishampel (1998) diag-

was originally diagnosed by its an enlarged nasal horn and

nosed Montanoceratops as having very short centra on the

straight lower dentary margin (Brown and Schlaikjer 1942),

fourth and fifth presacral vertebrae, but the rostral cervical

but the former character is based on misidentification of the

series is not exposed in TMP 88.11.1, and this state may not

right jugal-epijugal, and a straight dentary margin is more

be present in the holotype as discussed above. Despite the

broadly distributed in neoceratopsians calling to doubt its sta-

paucity of cranial remains for TMP 88.11.1, the limited com-

tus as uniquely derived in Montanoceratops. Sternberg (1951)

parisons possible indicate that it cannot be referred to Mon-

employed the same characters to distinguish it generically

tanoceratops, Leptoceratops, or Cerasinops. Comparisons with

from Leptoceratops. Ostrom (1978) added the extreme elonga-

Prenoceratops are limited not only by available material but

tion of the mid-caudal neural spines of AMNH 5464 as another

also the fact that the Prenoceratops cranial remains are young

feature distinguishing Montanoceratops from Leptoceratops.

A Redescription of the Montanoceratops cerorhynchus Holotype 77

FIGURE 4.7.

Cladistic analyses. (A) Single most-parsimonious tree (tree length = 262, consistency index = 0.653, retention index = 0.787) resulting from parsimony analysis of the revised and expanded matrix of Makovicky and Norell (2006) (see Appendix 4.1); (B–D) three equally parsimonious resolutions of the relationships among specimens of Montanoceratopswithin Leptoceratopsidae, when individual specimens of this taxon are treated as separate terminals. Analyses conducted using PAUP 4.10b (Swofford 1998) with the following search parameters: branch and bound search, all characters unordered, all 0-length branches collapsed.

Based on new information from MOR 542, Chinnery and

ber of characters relevant to resolving their phylogenetic posi-

Weishampel (1998) rediagnosed Montanoceratops as having

tion. Several of the added characters are autapomorphies and

caudoventrally directed and curved basipterygoid processes, a

thus have no bearing on recovering nodes in the cladogram,

short basisphenoid, a lenticular antorbital fossa, and a buccal

but were included to assess whether they optimize exclusively

shelf on the dentary teeth. Several of these traits, however, are

to Montanoceratops in light of ample missing data in closely

now known to be present in other leptoceratopsids following

related taxa.

study of historic specimens (Chinnery 2004; Makovicky 2001)

Analysis of the revised dataset (Table 4.2) resulted in a single

and discoveries of new taxa (Chinnery 2004; Chinnery and

tree (Fig. 4.7A) in which Montanoceratops is recovered as a lep-

Horner 2007), and none of them is exclusive to Montano-

toceratopsid that is more closely related to Prenoceratops (Lep-

ceratops. A partially everted iliac blade is also mentioned as a

toceratops, Udanoceratops) than to Cerasinops. This result is

diagnostic feature of this taxon (Chinnery and Weishampel

identical to that recently published by Chinnery and Horner

1998), but occurs to a varying degree in a variety of basal neo-

(2007). Of the characters previously considered diagnostic for

ceratopsians including Archaeoceratops. From his study of

Montanoceratops, several were recovered as unique autapo-

Montanoceratops braincases, Makovicky (2001) identified the

morphies. These include the caudoventral ridge along the

presence of a shallow fossa on the caudal surface of the su-

squamosal border, and possession of a supraoccipital depres-

praoccipital as a diagnostic trait of Montanoceratops.

sion. The straight ventral border of the dentary is also diag-

I added several characters discussed above to the cladistic

nostic of Montanoceratops among leptoceratopsids, although

data matrix published by Makovicky and Norell (2006) in or-

it is ambiguously optimized as either plesiomorphic rela-

der to establish which of the proposed diagnostic traits of

tive to taxa with curved dentaries, or as uniquely derived in

Montanoceratops represent autapomorphies, and to determine

Montanoceratops. The extreme elongation of some of the mid-

what combination of other characters may distinguish it from

caudal neural spines identified by Ostrom (1978) also cur-

its close relatives (Appendix 4.1). I also updated that data set

rently optimizes as an autapomorphy, but is ultimately depen-

by adding the recently published taxa Yinlong (Xu et al. 2006)

dent on the condition in Cerasinops for which this trait cannot

and Cerasinops (Chinnery and Horner 2007), as well as a num-

currently be evaluated.

78 makovicky

Only a limited degree of overlap is apparent between the

possibility that it could represent yet another leptoceratopsid

holotype and referred specimens of Montanoceratops. For ex-

taxon cannot be dismissed. Taken together with evidence for a

ample, no informative overlap occurs between the holotype

wider Laurasian distribution of leptoceratopsids (Lindgren et

and AMNH 5244, which is referred to Montanoceratops mainly

al. 2007), and the highly specialized jaw and dental features in

because it shares a single autapomorphy with the other re-

these animals, the recently discovered diversity suggests that

ferred specimen MOR 542. In order to gauge how robust these

leptoceratopsids represent a successful if poorly sampled cera-

referrals are in light of the limited overlap between specimens,

topsian radiation existing in parallel with coronosaurs.

and the incomplete nature of many leptoceratopsid taxa, I ran a second analysis in which Montanoceratops was parsed into a

Acknowledgments

separate terminal for each of the three specimens. In this fash-

I am grateful to Carl Mehling and Chris Norris for arranging a

ion, the preserved characters in each specimen including au-

loan of the holotype, and to B. Chinnery and Dave Weis-

tapomorphies were allowed to interact with each other as well

hampel for the loan of MOR 542. Rick Edwards and John

as with other taxa. This analysis resulted in three most parsi-

Weinstein provided excellent specimen photographs. I thank

monious trees (Fig. 4.7B–D), which differ in the relative posi-

Jack Horner for inspiring this research by ‘‘busting’’ the plaster

tions of the holotype relative to the referred specimens. In

skull reconstruction of AMNH 5464. This research was sup-

one of these topologies the Montanoceratops specimens form

ported by funding from NSF EAR 0418648 to the author.

a clade, but in the other two they form a polytomy at a node between Cerasinops and the Prenoceratops (Leptoceratops,

Appendix 4.1

Udanoceratops) clade, which is consistent with all three speci-

The following characters are appended to the character list of

mens deriving from a single taxon, because in no trees do any

Makovicky and Norell (2006). Those marked with an asterisk

of the Montanoceratops specimens group with another taxon.

represent putative autapomorphies of Montanoceratops and

Nevertheless, this analysis indicates the referrals of MOR 542

were included to evaluate whether they optimize uniquely to

and AMNH 5244 are not robust, because none of the diagnos-

this terminal.

tic features listed above is present in all three specimens, and as a consequence, they do not cluster together in every most-

134) Postorbital lateral surface smooth (0) or strongly

parsimonious solution. It should be borne in mind that this

sculptured (1).

analysis was constructed in order to test whether the holotype

135) Angular lateral surface smooth (0) or strongly

and referred specimens of Montanoceratops group together,

sculptured (1).

and several terminals are conspecific.

136) Rostral end of quadratojugal contacting jugal

Until recently, many authors viewed leptoceratopsids as

undivided (0) or bifid around caudal end of

a relict lineage, whose coexistence with the abundant and seemingly more diverse ceratopsids represents an anomaly or

jugal (1). 137) Nasals flat or convex on midline (0) or with distinct

paradox (Sternberg 1951; Dodson and Currie 1990). This in-

nasal midline depression (1).

terpretation appears to be no longer current in view of the

138) Parietal roof flat or gently convex (0) or with sharp

recent discovery of new members of this clade. At least four

midline crest (1).

leptoceratopsid taxa occur in the Campanian-Maastrichtian

139) Surangular without a lateral process below glenoid

of Alberta, Montana, and Wyoming, with some temporal

(0) or surangular knob (1) (the homology of this

overlap between them. This is almost comparable to the diver-

process, seen in Psittacosaurus, with the lateral

sity of centrosaurines in the same or laterally equivalent beds

glenoid wall of neoceratopsians is unclear).

from the same region and time span (Dodson et al. 2004).

140) Caudal neural spine height less than four times

Moreover, the diversity of leptoceratopsids is not fully re-

height of associated centrum (0) or four or more

solved. An isolated leptoceratopsid jaw from the Dinosaur

times taller than centrum (1).*

Park Formation (Ryan and Currie 1998) cannot currently be

141) Ridge along the caudoventral edge of squamosal

assigned to any named taxon. It differs from mandibles of

absent (0) or present (1).*

Leptoceratops, Cerasinops, and Prenoceratops in having a gentle

142) Occipital surface of supraoccipital flat, convex, or

rather than pronounced curvature of the dentary lower mar-

with midline ridge (0) or with midline depression

gin, and possessing a well-developed ‘‘chin.’’ While similar to

along base of midline ridge (1).*

Montanoceratops in the latter feature and in having a deeper

143) Accessory antorbital fenestra between naris and

rostral than caudal end (Chinnery and Weishampel 1998), it

antorbital fenestra absent (0) or present (1).

differs from Montanoceratops in having a gently curved ventral

144) Radius without lateral and medial tuberosities along

edge, and in lacking a constricted coronoid process notch.

distal half of shaft (0) or tuberosities present (1).

Although TMP 82.11.1 may be referable to Prenoceratops, the

A Redescription of the Montanoceratops cerorhynchus Holotype 79

Table 4.2. Data matrix for parsimony analyses of relationships between basal ceratopsian genera. Combinatorial codings: a = (0/1); b = (1/2). For the overall analysis shown in Fig. 4.7A, the individual scores for the three specimens of Montanoceratops were merged into a single terminal using the appropriate command in MacClade 4.05 (Maddison and Maddison 1992). TAXON

10

20

30

40

50

Hypsilophodon

00000? ? ?00

0000000000

001000000-

0000000000

0000?00011

Stegoceras

00000? ? ?00

0000000000

000011000-

0? ?010010?

0000000111

Archaeoceratops

1100101111

011010000?

?010001110

01?110? ?01

1? ? ? ? ? ?11?

Asiaceratops

? ? ? ?1? ? ?1?

? ? ? ?1?0? ? ?

?01? ? ? ? ? ? ?

??????????

??????????

Bagaceratops

11001?1111

01?01?01?0

1010001111

0101010112

1100?111? ?

Chaoyangsaurus

110110001?

?00?120? ? ?

?0?1? ?010?

? ? ? ?0? ? ? ?0

? ? ? ?00001?

Centrosaurus

1110101111

0110011211

1000112111

1102110111

1211111100

Leptoceratops

1100111111

0110120000

1110001110

0110001101

100? ?111?1

‘‘Graciliceratops’’

1100? ? ? ? ? ?

? ? ? ? ? ? ? ? ?0

? ? ? ? ? ?1? ? ?

? ? ? ? ?10111

1? ? ? ? ? ? ? ? ?

AMNH 5464

1? ? ? ? ? ? ? ?1

? ? ? ?1? ? ? ? ?

? ? ? ? ? ?1?10

011? ?011?1

??????????

AMNH 5244

??????????

??????????

??????????

??????????

??????????

MOR 542

1? ? ? ? ? ? ? ? ?

? ? ? ? ? ? ? ? ?0

10100? ? ? ? ?

??????????

1? ? ? ? ? ? ? ? ?

Protoceratops

1100111111

0100120100

1010001111

0101010112

110011111?

Psit. mongol.

1101100011

1011220001

100100110?

0000000000

0010001111

Triceratops

1110111111

0010011211

1000112111

1102110111

1211111100

Udanoceratops

1? ?0111111

?0? ?1?00?0

?1? ? ? ?1110

? ? ? ? ? ? ? ? ?1

1? ? ? ? ? ? ? ? ?

Zuniceratops

? ? ? ? ? ? ?1? ?

? ? ? ? ? ? ?0? ?

?0? ? ? ?1? ? ?

1? ? ? ? ? ? ? ? ?

??????????

Liaoceratops

?100101111

1010100000

101100100-

0101001100

1?00000111 0?00? ? ? ?1?

Psit. sinensis

1101110011

101122000?

?00000110-

0000000000

Xuanhuaceratops

??????????

? ?0? ? ? ? ? ? ?

??????????

??????????

??????????

Yamaceratops

11? ?1010? ?

? ?1? ? ? ? ? ?0

1010001010

0101001101

1100000111

Prenoceratops

? ?0?1? ? ?1?

00101?000?

?0100?1110

011? ?011?1

1? ? ? ? ? ? ? ? ?

Cerasinops

1? ? ? ? ? ? ? ? ?

??????????

?0? ? ? ?10? ?

01? ? ?011?1

1? ? ? ? ? ? ? ? ?

Yinlong

01001000?0

?0?001000?

?00001000-

0100000000

0000?00?1?

TAXON

60

70

80

90

100

Hypsilophodon

00—000000

01?1000010

0000100010

0000000000

0000000000

Stegoceras

00—000000

0?00000?00

?000100100

000000?000

0000000000

Archaeoceratops

?10? ? ?1110

1?01001110

1?101010? ?

01?11100?1

0101000101

Asiaceratops

? ? ? ? ? ?11?0

0? ? ? ? ? ? ? ? ?

?1101? ?0? ?

? ? ? ?01? ? ? ?

? ?3-000?00

Bagaceratops

? ?11101110

1001001111

111110?011

011111?111

013-000201

Chaoyangsaurus

? ? ? ? ? ?01?0

? ? ? ? ? ? ?110

0?101001? ?

00000000?1

1110000000

Centrosaurus

2211?12110

1110112110

211110?010

121011000?

013-111201

Leptoceratops

01000011?1

0011010111

1210001010

0110110101

013-002211

‘‘Graciliceratops’’

? ? ?100?1? ?

??????????

? ?1?1? ?0? ?

01?111? ? ?1

? ? ? ?000201

AMNH 5464

0?0? ? ? ? ? ? ?

? ? ? ? ? ? ? ?1?

? ?1?10?010

0?1? ? ? ? ? ? ?

? ?3? ? ?22?1

AMNH 5244

0? ? ? ? ?1121

? ? ? ?010? ? ?

??????????

??????????

??????????

MOR 542

? ? ? ? ? ?1121

00?1010? ? ?

? ?1?101010

01?1110? ?1

01? ?002211

Protoceratops

1211a01110

1001001111

1111001011

0111110111

0111000201

Psit. mongol

0100100100

0001001000

0110-1000?

01?0000001

003-000000

Triceratops

2210112110

1110112110

211110?010

1210110001

013-111201 013-002211

Udanoceratops

??????????

? ? ? ? ? ? ? ?1?

?21?0? ?0?0

01?011? ?01

Zuniceratops

? ? ? ? ? ? ?1? ?

??????????

?11?1? ?0? ?

0? ? ? ? ? ? ? ? ?

? ? ?-0?1201

Liaoceratops

0101100110

10?10011?0

0110011000

01?0101001

0100000100

Psit. sinensis

010010? ? ? ?

? ? ?0? ? ?000

01101000? ?

01?000000?

?03-00?0?0

Xuanhuaceratops

??????????

??????????

?11010?1? ?

? ? ?0?000?1

111?000000

Yamaceratops

010? ? ?1110

10?100111?

11101010?0

01111110?1

01a0000100

Prenoceratops

? ?0? ? ?1121

0? ? ? ? ? ?11?

12100010? ?

011011? ? ?1

013-002211

Cerasinops

1101001111

00?1010? ? ?

? ?1?0? ? ? ? ?

0b1111? ? ?1

?11?00210?

Yinlong

00—110020

0? ?00?1010

01000000?0

0000000101

?100000000

80 makovicky

Table 4.2. Continued TAXON

110

120

130

140

147

Hypsilophodon Stegoceras

0000000000

0000000000

0000000000

0000000000

00000-0

000000? ?0?

0?10? ? ? ? ?0

0?0?0? ? ?01

00011?-0?0

?0000-?

Archaeoceratops

? ? ?0?1000?

?0100?0? ? ?

? ? ? ? ? ? ?00?

?000001100

000?0-0

Asiaceratops

0?001? ? ? ? ?

? ? ? ? ?1? ? ? ?

1? ? ? ? ? ? ? ? ?

? ?0? ?0? ?0?

???????

Bagaceratops

010011? ? ? ?

1? ? ? ? ? ? ? ? ?

? ? ? ? ? ?0? ? ?

? ? ?000010?

001?0-?

Chaoyangsaurus

0000000000

00? ? ? ? ? ? ? ?

??????????

? ? ?110? ?0?

? ? ? ?0-0

Centrosaurus

1101111112

1203101110

1011111111

1110000100

00010-1

Leptoceratops

0110110101

1101011101

1000000000

0000000100

0001100

‘‘Graciliceratops’’

0100? ? ? ? ? ?

? ?1? ?1? ? ?1

? ? ? ? ? ?0? ?0

000000?10?

00000-?

AMNH 5464

?11? ?10111

?1?201110?

?0? ? ?00000

000? ?0? ?01

1? ? ?110

AMNH 5244

??????????

??????????

??????????

? ? ? ? ? ? ?1? ?

?1? ? ? ? ?

MOR 542

0110?1? ? ? ?

? ?1? ?1? ?01

10?00?0000

000?0? ? ?0?

?1? ?110

Protoceratops

0100110111

1112011101

1000010000

001000?100

00000-1

Psit. mongol

0000000000

1000000001

1100000000

0000010110

00000-0

Triceratops

11011?1112

1203101110

1011111111

1110000100

00010-1 ? ?01? ? ?

Udanoceratops

011011? ? ? ?

? ?02011? ?1

100? ? ? ? ? ?0

? ? ? ?000100

Zuniceratops

01?01? ? ? ? ?

? ? ? ? ?0? ? ? ?

? ? ? ? ? ? ? ?1?

? ? ?0?00? ? ?

?01?0-?

Liaoceratops

?10001? ? ? ?

??????????

?0? ? ? ? ? ? ? ?

? ? ?000110?

000?0-?

Psit. sinensis

00?000? ?00

10?1000? ? ?

?10? ? ? ? ? ?0

?000010110

0?000-?

Xuanhuaceratops

00000000? ?

? ?0? ? ? ? ? ?1

10? ? ? ? ? ? ?0

?0?110? ? ? ?

? ? ? ? ?-0

Yamaceratops

010011? ? ? ?

? ?0? ?1? ? ? ?

??????????

?00000?10?

00? ?0-0

Prenoceratops

0110?1? ? ? ?

??????????

??????????

? ? ?0?00?00

0000100

Cerasinops

0?1? ? ? ?0?1

101b?1?1?1

100? ? ? ? ?10

00000? ?10?

00? ?0-0

Yinlong

0000?0000?

?0? ?00?0? ?

? ? ?00100?0

? ?01101110

000?0-?

145) Coronoid process notch (Chinnery 2004) along caudal edge of dentary coronoid process absent (0) or present (1). 146) Coronoid process notch wide (0) or constricted notch (1).* 147) Cervical centra with ventral keels (0) or some or all postaxial centra without keels (1). References Cited Brown, B. 1914. Leptoceratops, a new genus of Ceratopsia from the Edmonton Cretaceous of Alberta. Bulletin of the American Museum of Natural History 33: 567–580. Brown, B., and E. M. Schlaikjer. 1942. The skeleton of Leptoceratops with the description of a new species. American Museum Novitates 1169: 1–15. Chinnery, B. J. 2004. Description of Prenoceratops pieganensis gen. et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Chinnery, B. J., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Chinnery, B. J., and D. B. Weishampel. 1998. Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) and relationships among basal neoceratopsians. Journal of Vertebrate Paleontology 18: 569–585.

Colbert, E. H. 1948. Evolution of the horned dinosaurs. Evolution 2: 145–163. Dodson, P., and P. J. Currie. 1990. Neoceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 593–618. Berkeley: University of California Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–516. Berkeley: University of California Press. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U. S. Geological Survey Monograph XLIX: 1–300. Kurzanov, S. M. 1992. A gigantic protoceratopsid from the Upper Cretaceous of Mongolia. Paleontological Journal 26: 103–116. [Translated from Russian]. Lindgren, J., P. J. Currie, M. Siverson, J. Rees, P. Cederstrom, and F. Lindgren. 2007. The first neoceratopsian dinosaur remains from Europe. Palaeontology 50: 929–937. Lull, R. S. 1933. A Revision of the Ceratopsia or Horned Dinosaurs. New Haven: Yale University Press. Maddison, W. P., and D. R. Maddison. 1992. MacClade version 3; Analysis of Phylogeny and Character Evolution. Sunderland: Sinauer Associates, Inc. Makovicky, P. J. 2001. A Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) braincase from the Horseshoe Canyon Formation of Alberta. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontol-

A Redescription of the Montanoceratops cerorhynchus Holotype 81

ogy of Philip J. Currie, pp. 243–262. Bloomington: Indiana University Press. ———. 2002. Taxonomic revision and phylogenetic relationships of basal Neoceratopsia (Dinosauria: Ornithischia). Ph.D. diss., Columbia University, New York. Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Marsh, O. C. 1890. Description of new Dinosaurian reptiles. American Journal of Science, Series 3, 39: 82–86. Maryanska, ´ T., and H. Osmólska. 1975. Protoceratopsidae (Dinosauria) of Asia. Acta Palaeontologica Polonica 33: 133–182. Nopcsa, F. 1923. Die Familien der Reptilien. Fortschritte der Geologie und Paleontologie 2: 1–210. Ostrom, J. H. 1966. Functional morphology and evolution of the ceratopsian dinosaurs. Evolution 20: 290–308. ———. 1978. Leptoceratops gracilis from the ‘‘Lance’’ Formation of Wyoming. Journal of Paleontology 52(3): 69–704. Russell, D. A. 1970. A skeletal reconstruction of Leptoceratops gracilis from the Upper Edmonton Formation (Cretaceous) of Alberta. Canadian Journal of Earth Sciences 7: 181–184. Ryan, M. J., and P. J. Currie. 1998. First report of protoceratopsians (Neoceratopsia) from the Late Cretaceous Judith River Group, Alberta, Canada. Canadian Journal of Earth Sciences 35: 820–826. Seeley, H. G. 1888. On the classification of the fossil animals commonly named Dinosauria. Proceedings of the Royal Society of London 43: 165–171. Sereno, P. C. 1984. The phylogeny of Ornithischia: A reappraisal. In W.-E. Reif and F. Westphal, eds., Third Symposium on Mesozoic Terrestrial Ecosystems: Short Papers, pp. 219–227. Tubingen: Attempto Verlag. ———. 1986. Phylogeny of the bird-hipped dinosaurs (order Ornithischia). National Geographic Research 2: 234–256.

82 makovicky

———. 2000. The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia. In M. J. Benton, M. Shishkin, D. Unwin, and E. Kurochkin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. 480–516. New York: Cambridge University Press. Sternberg, C. M. 1951. Complete skeleton of Leptoceratops gracilis Brown from the Upper Edmonton Member on Red Deer River, Alberta. National Museum of Canada Bulletin, Annual Report 123: 225–255. Swofford, D. 1998. PAUP*. Phylogenetic analysis using parsimony (*and other methods), 4. Sunderland: Sinauer Associates. Tsuihiji, T., and P. J. Makovicky. 2007. Homology of the neoceratopsian cervical bar elements. Journal of Paleontology 81: 1132–1138. Xu, X., C. A. Forster, J. M. Clark, and J. Mo. 2006. A basal ceratopsian with transitional features from the Late Jurassic of northwestern China. Proceedings of the Royal Society B-Biological Sciences 273: 2135–2140. Xu, X., P. J. Makovicky, X. L. Wang, M. A. Norell, and H. L. You. 2002. A ceratopsian dinosaur from China and the early evolution of Ceratopsia. Nature 416: 314–317. You, H., and P. Dodson. 2003. Redescription of neoceratopsian dinosaur Archaeoceratops and early evolution of Neoceratopsia. Acta Palaeontologica Polonica 48: 261–272. ———. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press. You, H., D. Q. Li, Q. Ji, M. C. Lamanna, and P. Dodson. 2005. On a new genus of basal neoceratopsian dinosaur from the Early Cretaceous of Gansu Province, China. Acta Geologica Sinica 79: 593–597. Zhao, X. J., Z. W. Cheng, X. Xu, and P. J. Makovicky. 2006. A new ceratopsian from the Upper Jurassic Houcheng Formation of Hebei, China. Acta Geologica Sinica 80: 467–473.

5 First Basal Neoceratopsian from the Oldman Formation (Belly River Group), Southern Alberta T E T S U T O M I YA S H I TA , P H I L I P J . C U R R I E , A N D B R E N D A J . C H I N N E R Y- A L L G E I E R

an isolated frontal (TMP 87.89.8) referred to Pre-

ceratopsians (non-ceratopsid neoceratopsians) is surprisingly

noceratops sp. from the Oldman Formation of southern

poor (Ryan and Currie 1998; You and Dodson 2004). This con-

Alberta is a recent addition to the scarce record of North

trasts with the remarkable diversity of Asian basal neocera-

American basal neoceratopsians. TMP 87.89.9 shares the

topsians during this time (You and Dodson 2004; Makovicky

diagnostic characters of Prenoceratops, such as the trans-

and Norell 2006). Recent discoveries (Ryan and Currie 1998;

verse postorbital ridge and the deep frontal depression.

Chinnery 2004; Chinnery and Horner 2007) strongly suggest

The crista cranii coinciding with the interfrontal suture is

that the diversity of North American basal neoceratopsians is

identified as a new autapomorphy for the genus Pre-

greater than previously thought. Basal neoceratopsians have

noceratops. A unique combination of characters in TMP

been identified from the Milk River, Dinosaur Park, Horseshoe

87.89.8 includes the absence of the olfactory bulb im-

Canyon, St. Mary River, and Scollard formations in southern

pression, presence of the cerebral fossa, and a fossa that

Alberta, Canada, and from the Two Medicine and St. Mary

probably represents part of the nasal cavity. Although

River formations in northwestern Montana, in the United

this suggests that TMP 87.89.8 may represent a distinct

States (Brown 1914; Sternberg 1951; Chinnery and Weisham-

species of Prenoceratops, the possibility of ontogenetic

pel 1998; Ryan and Currie 1998; Makovicky 2001; Chinnery

variation cannot be ruled out. There were at least three

2004; Chinnery and Horner 2007). The chronological range of

distinct basal neoceratopsian taxa in the Campanian of

the basal neoceratopsians in Alberta thus spans from the latest

southern Alberta and northwestern Montana: Cerasinops,

Santonian to latest Maastrichtian (Ryan and Currie 1998), al-

Leptoceratops, and Prenoceratops. Prenoceratops is the sec-

though there are several long gaps within this range where

ond ornithischian genus exclusively occurring in the up-

basal neoceratopsian material has not been recovered. The

per Oldman and upper Two Medicine formations,

longest of these gaps occurs between the Milk River and Dino-

suggesting a faunal link due to the similar inland, rela-

saur Park formations. Although cf. Montanoceratops (TMP

tively dry environments.

82.11.1) was reportedly collected from the Belly River Group and could have potentially filled this gap, the specimen is

Introduction

likely to have been recovered from the Willow Creek Formation (a rationale for this interpretation in ‘‘Discussion’’).

Despite our understanding of the richness of Late Cretaceous

Except for excellent skeletons of Leptoceratops (Brown 1914;

dinosaur faunas, the North American record of basal neo-

Sternberg 1951) from the Scollard Formation and that of cf.

83

FIGURE 5.1.

Geologic correlation chart for Upper Cretaceous strata in southern Alberta and northwestern Montana. The grey shade indicates formations from which basal neoceratopsian material has been collected (sources cited in the text). Marine units are shaded black. Radiometric dates and nomenclature from Brinkman (2003), Eberth (2005), Eberth et al. (2001), Horner et al. (2001), Rogers (1997, 1998), and Rogers et al. (1993).

84 miyashita, currie, & chinnery-allgeier

Referred Specimen. TMP 87.89.8, an isolated right frontal. Locality. TMP 87.89.8 was surface collected from the upper Oldman Formation (Fig. 5.1) at Devil’s Coulee, southern Alberta (Fig. 5.2), close to the nesting site of Hypacrosaurus stebingeri (Horner and Currie 1994). The upper Oldman Formation exposed at Devil’s Coulee is dated at 75.05 Ma (Eberth and Deino 1992). The Devil’s Coulee locality is approximately correlated with the Lethbridge Coal Zone in the uppermost part of the Dinosaur Park Formation in Dinosaur Provincial Park, which marks an early westward transgression of the Bearpaw Sea in the southern plains of Alberta. Comments. The basal neoceratopsian affinity of TMP 87.89.9 is supported by the combination of the following characters: medially emarginated orbital rim of thick frontal; prefrontal separated by nasal from protruding anterior process of frontal; wide and flat orbital depression; and unfused frontal suture. TMP 87.89.8 is referred to Prenoceratops Chinnery 2004 based on the following diagnostic characters: crista cranii coalesced to the interfrontal suture; straight, transverse postorbital ridge; FIGURE 5.2. Locality map showing Devil’s Coulee, where TMP

87.89.8 was collected. The map is modified after Horner and Currie (1994).

and frontal depression excavated deeper than one-third the thickness of frontal (Chinnery 2004).

Description Montanoceratops (TMP 82.11.1; Ryan and Currie 1998) from

TMP 87.89.8 (Fig. 5.3) is an isolated right frontal. The anterior

the Willow Creek Formation, the basal neoceratopsian speci-

portion and much of the lateral margin were lost due to ero-

mens from Alberta are mostly fragmentary and often repre-

sion. As preserved, the specimen measures 98.4 mm in length.

sented only by teeth from microvertebrate localities.

The bone is deep medially, but becomes shallower anteriorly,

A basal neoceratopsian frontal (TMP 87.89.8) represents the

laterally, and posteriorly. The frontal of TMP 87.89.8, Lepto-

first record of this group from the Oldman Formation of Al-

ceratops CMN 8889, and UWGM 3992.05 (Fig. 5.4C, D) is

berta, and provides new information on the cranial anatomy

thicker and wider than it is in Prenoceratops (Fig. 5.4A, B).

of basal neoceratopsians.

Leptoceratops and the animal represented by TMP 87.89.8 are

Institutional Abbreviations. CMN: Canadian Museum of Na-

larger than Prenoceratops (ICM 2003.1.1; MNHCM field num-

ture, Ottawa; ICM: Children’s Museum of Indianapolis, Indi-

ber 61, Fig. 5.4A). This is consistent with an allometric trend in

anapolis; MNHCM: Mokpo Natural History and Culture Mu-

dinosaurs that the skull roof generally becomes wider and

seum, Mokpo; TMP: Royal Tyrrell Museum of Palaeontology,

thicker as body size increases. As in most ceratopsians (Dod-

Drumheller; UWGM: University of Wisconsin Geology Mu-

son et al. 2004), the frontal is not fused to its counterpart but

seum, Madison.

the medial frontal sutural surface is sculpted by numerous

Anatomical Abbreviations. cc: crista cranii; ci: impression of cerebrum; fd: frontal depression; if: interfrontal suture; ls:

grooves. A mid-height longitudinal groove separates this surface into upper and lower parts (Fig. 5.3D).

laterosphenoid suture; na: nasal contact; nc: nasal cavity; ob:

In dorsal view (Fig. 5.3A), a sutural surface at the anterior

olfactory bulb impression; or: orbital rim; os: orbitosphenoid

edge represents the posterior end of the nasal contact, suggest-

suture; pa: parietal suture; po: postorbital contact; pr: post-

ing that the nasal excluded the prefrontals from contacting

orbital ridge; sr: sagittal ridge.

each other. Most basal neoceratopsians share this character, whereas in at least some ceratopsids (e.g., Centrosaurus and

Systematic Paleontology

Styracosaurus; Sampson et al. 1997), the prefrontals meet each other at the midline and prevent frontal-nasals contact. The

Ornithischia Seeley 1888

orbital rim is gently emarginated as in other basal cera-

Ceratopsia Marsh 1890

topsians. The deeply sculpted grooves associated with vascu-

Neoceratopsia Sereno 1986

lar pits on the dorsal surface are similar to cranial ornamenta-

Prenoceratops Chinnery 2004

tion present on most ceratopsids. Leptoceratops (UWGM

Prenoceratops sp.

31992.05; Fig. 5.4C) and Prenoceratops (MNHCM field num-

First Basal Neoceratopsian from the Oldman Formation 85

FIGURE 5.3.

TMP 87.89.8 in (A) dorsal, (B) ventral, (C) right lateral, and (D) medial views. Scale bars are 5 cm.

86 miyashita, currie, & chinnery-allgeier

phology of the sagittal ridge, but the ridge is absent on Prenoceratops frontals (MNCHM field numbers 61, 65). The partially preserved parietal sutural surface indicates that the frontals were wedged between the parietals on the midline. The frontal-parietal suture is roughly transverse in other basal ceratopsians such as Protoceratops (Gregory and Mook 1925). In ventral view (Fig. 5.3B), the crista cranii delineates the wide orbital space medially and coalesces with the interfrontal suture. The right and left cristae would have met at the midline, constricting the dorsal sulcus for passage of the olfactory tract. The olfactory tract then would have exited anteriorly below this constriction. Prenoceratops frontals (MNHCM field numbers 61 and 65) show the same condition, but Leptoceratops (UWGM 31992.05; Fig. 5.4D) retains separation between the right and left cristae as in other dinosaurs. Ceratopsids have a transverse ridge connecting the cristae cranii at about the same position (Lehman 1989). A similar transverse ridge is also observed in hadrosaurids (Horner 1992; Evans 2006) and in basal euornithopods (Weishampel et al. 2003), in which the sphenethmoid suture extends onto the ridge. Such a ridge is absent in basal ornithopods (Sues 1980; Galton 1989, 1997) and pachycephalosaurs (TM pers. obs.). The connection between the crista cranii and the orbitosphenoid is substantial, occurring from the constriction between the cristae posterolaterally and extending to the laterosphenoid suture. Based on the position of this suture, the laterosphenoid of TMP 87.89.8 Prenoceratops frontals in (A) dorsal view (MNHCM field number 65 on left, MNHCM field number 61 on right) based on Chinnery (2004); (B) ventral view (field number 65). Leptoceratops frontals (UWGM 3992.05) in (C) dorsal and (D) ventral views based on Ott (2003). Scale bars are 2 cm.

FIGURE 5.4.

would have had an exceptionally long lateral process to contact the postorbital. There is no clear indication of the sphenethmoid contact, which has been observed on the transverse ridge between cristae cranii in euornithopods (Horner 1992; Weishampel et al. 2003; Evans 2006). The crista cranii of Leptoceratops have a rough texture along the anteromedial margin of the orbital depression that probably marks the spheneth-

bers 61 [Fig. 5.4A] and 65 [Fig. 5.4A]) have less conspicuous

moid suture (UWGM 31992.05; Fig. 5.4D).

pitting, with shallow parallel grooves occurring on the dorsal

A deep, wide depression occurs near the interfrontal suture

surface of the frontals. The postorbital contact is partially pre-

on the ventral surface at the broken anterior end (Fig. 5.3B)

served in TMP 87.89.8, and indicates that the postorbital over-

that most likely marks the posterior end of the nasal cavity.

lapped the frontal posteriorly as it does in other basal cera-

Leptoceratops has a shallow, wide trough for the nasal that ex-

topsians including Leptoceratops, Liaoceratops, Psittacosaurus,

tends to the ventral surface of the frontal (Ott 2003). The de-

and Yamaceratops (Makovicky and Norell 2006).

pression may be absent in Prenoceratops, although the corre-

The posterior one-third of the bone is marked by the deep

sponding area seems to be slightly depressed in two of the

frontal depression (Fig. 5.3A). As in Prenoceratops (Chinnery

three Prenoceratops frontal specimens (MNHCM field numbers

2004) and Cerasinops (Chinnery and Horner 2007), the frontal

48 and 61 [Fig. 5.4B]). The presumed nasal cavity depression is

depression is deeper than one-third of the total frontal thick-

much deeper in TMP 87.89.8 than it is in either Prenoceratops

ness. A sharply defined, straight transverse postorbital ridge

or Leptoceratops. One of the frontals of Prenoceratops (MNHCM

separates the frontal depression from the rest of the bone. This

field number 65, Fig. 5.4A) preserves a clear olfactory bulb

is diagnostic for Prenoceratops (Chinnery 2004), but is more

impression behind the nasal cavity (TM pers. obs.), but TMP

dorsally pronounced in TMP 87.89.8. In other basal cera-

87.89.8 has no recognizable feature in the same position. In

topsians, this postorbital ridge is gently curved. A low, narrow

TMP 87.89.8, there is a shallow, small depression near the

sagittal ridge separates the frontal depression from the mid-

interfrontal suture on the ventral surface at the posterior end

line at least anteriorly in TMP 87.89.8. Leptoceratops (CMN

of the specimen, which represents the cerebral fossa (Fig.

8889; UWGM 31992.05) shares with TMP 87.89.8 the mor-

5.3B). Since size and depth is much smaller than that expected

First Basal Neoceratopsian from the Oldman Formation 87

for a cerebrum from a basal neoceratopsian of this size, the

to Prenoceratops, Leptoceratops sp. is known from the Dinosaur

cerebrum must have been only in a partial contact with the

Park Formation (Ryan and Currie 1998) and Cerasinops from

frontal. Prenoceratops frontals do not form a distinct depres-

the Two Medicine Formation below the interval that pro-

sion for the cerebrum.

duced Prenoceratops (Chinnery and Horner 2007). Ryan and Currie (1998) also reported a basal neoceratopsian

Discussion

tentatively identified as Montanoceratops sp. (TMP 82.11.1) recovered from a massive yellow sandstone block located on the

TMP 87.89.8 shares with Prenoceratops the autapomorphies of

Oldman River (Tanke 2007). Examination of the specimen

(1) transversely straight postorbital ridge, (2) frontal depres-

suggests that the preserved matrix is more consistent with

sion that is deeper than one-third the thickness of the frontal,

sediments from the Maastrichtian-aged portion of the Willow

and (3) crista cranii that coincides with the interfrontal suture

Creek Formation (Fig. 5.1) that predominantly crops out

(Chinnery 2004). TMP 87.89.8 has four possible autapomor-

along the Oldman River, rather than from the Belly River

phic characters: a deeply sculpted dorsal surface of the frontal;

Group that is not widely exposed along the river (Wall and

a deep nasal cavity depression; an interfrontal suture that is

Rosene 1977). Grain size of the matrix associated with TMP

divided into upper and lower parts; and an unequivocal sagit-

82.11.1 is finer than the typical medium-to-fine sands in the

tal ridge along the interfrontal suture that would have sepa-

Belly River Group, and the matrix contains more cement than

rated right and left frontal depressions at least anteriorly. Un-

typical sandstone from the Belly River Group (TM unpub-

fortunately, these characters are not consistently present on

lished data). Therefore, we propose that TMP 82.11.1 most

the remaining North American neoceratopsians (Cerasinops,

likely came from the Maastrichtian-aged Willow Creek Forma-

Leptoceratops, Montanoceratops, and Zuniceratops) to facilitate a

tion and, therefore, can no longer be included in consider-

robust phylogenetic analysis. It is possible that these dif-

ation of the diversity of North American basal neoceratop-

ferences, in addition to the width and the thickness of the

sians from the Campanian.

frontals, are the result of individual variation, or are size-

Prenoceratops and the lambeosaurine Hypacrosaurus stebin-

related. The frontal of TMP 87.89.8 would have been at least

geri (Horner and Currie 1994) are the only two ornithischians

approximately 30% longer than the frontals of Prenoceratops

shared between the upper Oldman Formation (OF) of Alberta

(MNHCM field numbers 61 and 65; Fig. 5.4A, B) suggesting

and the upper Two Medicine Formation (TMF) of Montana.

that it may represent a large adult of Prenoceratops pieganensis.

These formations represent similarly well-drained, seasonally

Alternatively, TMP 87.89.8 may have come from a new species

dry floodplain environments (OF, D. A. Eberth pers. com.

of Prenoceratops. Until more material can be recovered from

2007; TMF, Rogers 1997). The type material of Prenoceratops

the Oldman Formation of Alberta, we conservatively refer

(ICM 2003.1.1) was recovered from a bonebed 50 m below the

TMP 87.89.8 to Prenoceratops sp.

Bearpaw Formation (Chinnery 2004), which places its strati-

Comparison of TMP 87.89.8 to other basal neoceratopsians

graphic position close to the bentonite dated at 74.3 Ma in the

from North America indicates that there is significant mor-

upper Two Medicine Formation (Horner et al. 1992). With the

phological variation in skull roof characters (e.g., shape of

radiometric dating of the Devil’s Coulee locality at 75.05 Ma

the crista cranii, depth of frontal depressions, and presence

(Eberth and Deino 1992), the type material of Prenoceratops is

or absence of nasal cavity, impression of olfactory bulb, and

less than 1.0 Ma younger than TMP 87.89.8. Presence of Preno-

cerebral fossa) within these taxa. These characters are candi-

ceratops and Hypacrosaurus stebingeri in the upper Two Medi-

dates for a refined phylogenetic analysis of ceratopsians in

cine formations suggests persistence of the Oldman Forma-

the future.

tion ornithischian fauna from Alberta in the inland setting of

Finally, it must be noted that the sculpturing on the dorsal

Montana up to the maximum westward transgression of the

frontal surface of Leptoceratops (UWGM 31992.05; Fig. 5.4C),

Bearpaw Sea. It is also possible that the younger occurrences of

Prenoceratops (MNCHM field number 61), and TMP 87.89.8

Prenoceratops and Hypacrosaurus stebingeri in the Two Medicine

implies the presence of a thick keratinous covering (Horner

Formation are a result of inland migration in response to the

and Marshall 2002) over at least this portion of the skull.

westward Bearpaw transgression, a hypothesis supported by the suggestions of a number of authors (e.g., Sternberg 1951;

CAMPANIAN BASAL NEOCERATOPSIAN DIVERSITY IN NORTH AMERICA

Ryan and Currie 1998; Ott 2006; You and Dodson 2004) who have attributed the rarity of basal neoceratopsians in western North America to their preference for an inland, drier habitat.

TMP 87.89.8 is the first basal neoceratopsian material from

The seasonally dry floodplain depositional setting of the

the Oldman Formation, helping to fill in one of the longest

strata at Devil’s Coulee (D.A. Eberth pers. comm. 2007) also

gaps in the basal neoceratopsian record in Alberta. In addition

supports this hypothesis.

88 miyashita, currie, & chinnery-allgeier

Acknowledgments

James Gardner and Brandon Strilisky (TMP) provided access to collections. We acknowledge that Nick Longrich (University of Calgary) independently came up with the identification of TMP 87.89.8. We also thank Lisa Buckley (Tumbler Ridge Dinosaur Center), Peter Dodson (University of Pennsylvania), Takuya Konishi (University of Alberta), Hans Larsson (McGill University), Michael Ryan (Cleveland Museum of Natural History), Darren Tanke (TMP), and François Therrien (TMP) for discussions. François Therrien and David Eberth (TMP) shared unpublished information on Devil’s Coulee. Don Henderson (TMP) and Don and Lorraine Woodruff (Drumheller) provided accommodations for TM. Reviews by David Evans (Royal Ontario Museum) and Michael Ryan improved clarity of the manuscript and led to refined assessment of the basal neoceratopsian diversity. T. Miyashita also acknowledges Federico Fanti (University of Bologna), Eva Koppelhus (University of Alberta), Kaoru Kitahara, Kazuhiro Magome, Jun-ichi and Kanae Miyashita, Hisao Nakagawa, Miyuki Tajima, and Mariko Takahashi (Tokyo) for their long-standing support. References Cited Brinkman, D. B. 2003. A review of nonmarine turtles from the Late Cretaceous of Alberta. Canadian Journal of Earth Sciences 40: 557–571. Brown, B. 1914. Leptoceratops, a new genus of Ceratopsia from the Edmonton Cretaceous of Alberta. Bulletin of the American Museum of Natural History 33: 567–580. Brown, B., and E. M. Schlaikjer. 1942. The structure and relationship of Protoceratops with a description of a new species. American Museum Novitates 1169: 1–15. Chinnery, B. 2004. Description of Prenoceratops pieganensis gen. et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Chinnery, B., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Chinnery, B., and D. B. Weishampel. 1998. Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) and relationships among basal neoceratopsians. Journal of Vertebrate Paleontology 18: 569–585. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A. 2005. The geology. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 54–82. Bloomington: Indiana University Press. Eberth, D. A., P. J. Currie, D. B. Brinkman, M. J. Ryan, D. R. Braman, J. D. Gardner, V. D. Lam, D. N. Spivak, and A. G. Neuman. 2001. Alberta’s dinosaurs and other fossil vertebrates:

Judith River and Edmonton Groups (CampanianMaastrichtian). In C. L. Hill, ed., Guidebook for the Field Trips of the Society of Vertebrate Paleontology 61st Annual Meeting: Mesozoic and Cenozoic Paleontology in the Western Plains and Rocky Mountains, pp. 47–75. Museum of the Rockies Occasional Paper 3. Eberth, D. A., and A. A. Deino. 1992. Geochronology of the Nonmarine Judith River Formation of Southern Alberta. Society of Economic Paleontologists and Mineralogists, 1992 Theme meeting, Mesozoic of the Western Interior. Evans, D. C. 2006. Nasal cavity homologies and cranial crest function in lambeosaurine dinosaurs. Paleobiology 32: 109– 125. Galton, P. M. 1989. Crania and endocranial casts from ornithopod dinosaurs of the families Dryosauridae and Hypsilophodontidae (Reptilia: Ornithischia). Geologica et Palaeontologica 23: 217–239. ———. 1997. Cranial anatomy of the basal hypsilophodontid dinosaur Thescelosaurus neglectus Gilmore (Ornithischia: Ornithopoda) from the Upper Cretaceous of North America. Revue Paleobiologie (Geneve) 16: 231–258. Gregory, W. K., and C. C. Mook. 1925. On Protoceratops, a primitive ceratopsian dinosaur from the Lower Cretaceous of Mongolia. American Museum Novitates 156: 1–9. Horner, J. R. 1992. Cranial morphology of Prosaurolophus (Ornithischia: Hadrosauridae) with descriptions of two new hadrosaurid species and an evaluation of hadrosaurid phylogenetic relationships. Museum of the Rockies Occasional Paper 2. Horner, J. R., and P. J. Currie. 1994. Embryonic and neonatal morphology and ontogeny of a new species of Hypacrosaurus (Ornithischia, Lambeosauridae) from Montana and Alberta. In K. Carpenter, K. F. Hirsch, and J. R. Horner, eds., Dinosaur Eggs and Babies, pp. 312–336. Cambridge: Cambridge University Press. Horner, J. R., and C. Marshall. 2002. Keratinous covered dinosaur skulls. Journal of Vertebrate Paleontology 22(3, Suppl.): 67A. Horner, J. R., J. G. Schmitt, F. Jackson, and R. Hanna. 2001. Bones and rocks of the Upper Cretaceous Two Medicine-Judith River clastic wedge complex, Montana. In C. L. Hill, ed., Guidebook for the Field Trips of the Society of Vertebrate paleontology 61st Annual Meeting: Mesozoic and Cenozoic Paleontology in the Western Plains and Rocky Mountains, pp. 3–13. Museum of the Rockies Occasional Paper 3. Horner, J. R., D. J. Varricchio, and M. B. Goodwin. 1992. Marine transgressions and the evolution of Cretaceous dinosaurs. Nature 358: 59–61. Lehman, T. M. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. Makovicky, P. J. 2001. A Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) braincase from the Horseshoe Canyon Formation of Alberta. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 243–262. Bloomington: Indiana University Press.

First Basal Neoceratopsian from the Oldman Formation 89

Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Marsh, O. C. 1890. A new family of horned dinosaurs from the Cretaceous. American Journal of Science 36: 477–478. Ott, C. J. 2003. Cranial anatomy and biogeography of the first Leptoceratops gracilis (Dinosauria: Ornithischia) specimens from the Hell Creek Formation, southeast Montana. M.Sc. thesis. University of Wisconsin-Madison, Madison. ———. 2006. Cranial anatomy and biogeography of the first Leptoceratops gracilis (Dinosauria: Ornithischia) specimens from the Hell Creek Formation, southeast Montana. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 213–233. Bloomington: Indiana University Press. Rogers, R. R. 1997. Two Medicine Formation. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, pp. 760–765. San Diego: Academic Press. ———. 1998. Sequence analysis of the Upper Cretaceous Two Medicine and Judith River formations, Montana: Nonmarine response to the Claggett and Bearpaw marine cycles. Journal of Sedimentary Research 68: 615–631. Rogers, R. R., C. C. Swisher, and J. R. Horner. 1993. 40Ar/ 39Ar age and correlation of the non-marine Two Medicine Formation (Upper Cretaceous), northwestern Montana. Canadian Journal of Earth Sciences 30: 1066–1075. Ryan, M. J., and P. J. Currie. 1998. First report of protoceratopsians (Neoceratopsia) from the Late Cretaceous Judith River Group, Alberta, Canada. Canadian Journal of Earth Sciences 35: 820–826. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Cera-

90 miyashita, currie, & chinnery-allgeier

topsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Seeley, H. G. 1888. On the classification of the fossil animals commonly called Dinosauria. Proceedings of the Royal Society of London 43: 165–171. Sereno, P. C. 1986. Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). National Geographic Research 2: 234–256. Sternberg, C. M. 1951. Complete skeleton of Leptoceratops gracilis Brown from the upper Edmonton Member on Red Deer River, Alberta. Annual Report of the National Museum for the Fiscal Year 1949–1950, Bulletin 123: 225–255. Sues, H.-D. 1980. Anatomy and relationships of a new hypsilophodontid dinosaur from the Lower Cretaceous of North America. Palaeontographica A 169: 51–72. Tanke, D. H. 2007. Ceratopsian discoveries and work in Alberta, Canada: Historical review and census. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs. CD ROM appendix. Drumheller: Royal Tyrrell Museum of Palaeontology. Wall, J. H., and R. K. Rosene. 1977. Upper Cretaceous stratigraphy and micropaleontology of the Crowsnest Pass-Waterton area, southern Alberta Foothills. Bulletin of Canadian Petroleum Geology 25: 842–867. Weishampel, D. B., C.-M. Jianu, Z. Csiki, and D. B. Norman. 2003. Osteology and phylogeny of Zalmoxes (n. g.), an unusual euornithopod dinosaur from the Latest Cretaceous of Romania. Journal of Systematic Paleontology 1: 65–123. You, H.-L., and P. Dodson. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press.

6 Zuniceratops christopheri: The North American Ceratopsid Sister Taxon Reconstructed on the Basis of New Data D O U G L A S G . W O L F E , J A M E S I . K I R K L A N D , D AV I D S M I T H , K A R E N P O O L E , B R E N D A C H I N N E R Y- A L L G E I E R , A N D A N D R E W M C D O N A L D

Zuniceratops christopheri Wolfe and Kirkland 1998 is

the 2000 paper include additional horncores, mandibles, a

a neoceratopsian dinosaur from the middle Turonian

complete jugal, braincase, frill fragments including the tee-

Moreno Hill Formation of west-central New Mexico. The

shaped proximal parietal bar, and several appendicular ele-

holotype is a partial skeleton (MSM P2101), with addi-

ments. These elements provide the evidence necessary to con-

tional bonebed material comprising at least seven partial

struct a more accurate skeletal reconstruction of Zuniceratops,

skeletons. The bonebed has yielded numerous disarticu-

included here, and provide an indication of ontogenetic and

lated cranial and postcranial elements, expanding our

individual variation for the taxon based on a minimum of

osteological knowledge of Zuniceratops and allowing exe-

seven Zuniceratops partial skeletons recovered from the bone-

cution of a full skeletal mount and reconstruction.

bed locality. Anatomical Abbreviations. aof: antorbital fenestra margin;

Introduction

bc: braincase—right lateral (upper) and dorsal (lower) views; bo: basiocciput; bsp: basisphenoid; cp: coronoid process; d:

The holotype (MSM P2101) of Zuniceratops christopheri (Wolfe

dentary in left lateral view; fm: foramen magnum; ju: jugal;

and Kirkland 1998), currently the oldest North American cera-

lac: lacrimal; lsp: laterosphenoid; mjc: maxillary-jugal contact

topsian possessing well-developed brow horns, was discov-

surface; mx: maxilla in left lateral view; n: naris opening; na:

ered as floated, but associated, skeletal fragments in 1996. Pub-

nasal in left lateral view; oc: occipital condyle; pa1: parietal

lication of a preliminary description in 1998 coincided with

fragment—rostral-dorsal view of proximal medial segment;

the discovery of a nearby bonebed that eventually yielded sev-

pa2: parietal fragment—rostral-dorsal view of marginal seg-

eral hundred unassociated but identifiable Zuniceratops ele-

ment; pa3: parietal fragment—rostral-dorsal view of marginal

ments, and many less diagnostic fragments, some of which

segment; pal: palpebral; par: paroccipital process; pd: preden-

were tentatively included in the 1998 description. Subsequent

tary in left lateral view; pmx: right premaxilla in medial view;

recovery and preparation of the bonebed specimens led to the

poh: left postorbital horncore in lateral view associated with

recognition (Wolfe 2000) that putative squamosal elements

jugal fragment and maxilla (MSM P2102); q/par: quadrate

attributed in 1998 to Zuniceratops were actually ischia of the

and paroccipital contact area; qu: right quadrate—caudal

therizinosaurid Nothronychus mckinleyi (Kirkland and Wolfe

view of broken medial margin; r: rostrum in medial view with

2001), and that Zuniceratops nasals had no evidence of a nasal

broken anterior and left lateral margins; so: supraoccipital

horn. Preparation and study of bonebed elements following

process; sq1: right squamosal fragment in medial view with

91

(A) Zuni Basin Field Locality with Zuniceratops skull reconstruction by Rob Gaston; (B) Partial stratigraphic section of the lower Moreno Hill Formation in the Two Rocks Balanced area with significant fossil occurrences indicated. Revised after Wolfe and Kirkland (1998).

FIGURE 6.1.

broken rostral and caudal margins; sq2: right squamosal frag-

microfauna, and dinosaur track impressions (Wolfe 2006).

ment in medial view with relatively complete rostral margin

The two Zuniceratops localities are at approximately the same

and broken caudal margin.

stratigraphic interval. The holotype specimen was collected from sideritic concretions weathered from dark-colored mud-

Stratigraphic Context and Occurrence

stones a few meters above the bonebed horizon, which lies at the pinched-out margin of a cross-bedded sandstone. Bone-

Zuniceratops specimens have been recovered from two locali-

bed specimens were mostly contained in sandy mudstone

ties approximately 750 m apart, within the middle third of the

containing shale partings and rip-up clasts and abundant car-

lower Moreno Hill Formation near the Arizona-New Mexico

bonized logs. Specimens described here are curated at the

border, in a 3 square km outcrop area referred to in field notes

Mesa Southwest Museum (MSM), Mesa, Arizona.

as ‘‘Two Rocks Balanced’’ (Fig. 6.1; Wolfe and Kirkland 1998;

The Zuniceratops holotype locality (MSM 98-65) contained

Wolfe 2000). Correlations using inter-tonguing stratigraphic

numerous fragments of large sideritic concretions containing

relations, ammonites, and radiometric dates indicate that the

well-indurated, permineralized bone. Mosaic cracking on the

lower Moreno Hill is mid- to late Turonian (approximately 90

bone surface suggests exposure and desiccation prior to burial

Ma) in age (Molenaar et al. 2002), with the fossil sites occur-

in an aquatic setting.

ring near the base of this interval. The bonebed discovered in

The holotype skeleton (MSM P2101) includes a postorbital

1998 yielded several hundred elements, including elements of

horn, partial dentary with intact and isolated teeth, partial

Zuniceratops, as well as the first described North American

braincase, coracoid, humerus, caudal centra and other identi-

therizinosaurid, Nothronychus mckinleyi (Kirkland and Wolfe

fiable fragments with no duplication of parts.

2001). Other localities within several hundred meters of these

The Zuniceratops bonebed locality (MSM 98-78) contained

sites have produced a new hadrosauroid (McDonald et al.

at least seven partial Zuniceratops skeletons based on dupli-

2006), two partial skeletons of a small tyrannosauroid coe-

cated elements (left dentaries). Some specimens, such as den-

lurosaur (Denton et al. 2004), abundant aquatic vertebrate

taries with teeth, horncores, and a complete left jugal, are

92 wolfe, kirkland, smith, poole, chinnery-allgeier, & mcdonald

FIGURE 6.2.

Cranial elements of Zuniceratops christopheri Wolfe and Kirkland 1998.

very well preserved despite disarticulation and apparent trans-

et al. 2004). Possible communal behavior in Zuniceratops is

port. There is an unaccounted-for bias toward the preserva-

implied by the co-association of individuals of various sizes/

tion of left-side skull elements. Other specimens are crushed

ages in a neoceratopsian bonebed. We interpret the Zunicera-

and some show signs of having rotted internally. Rounded

tops holotype site to be contemporaneous or slightly younger

bone clasts are common. Large, compressed, carbonized logs

than the bonebed site by at most hundreds or a few thousand

were intimately associated with the bones and commonly

years.

provided a parting surface between skeletal elements. Complementary elements of the associated Nothronychus skeleton were discovered several meters apart in the quarry and pro-

Description of Zuniceratops

vide some indication of the degree of disarticulation and

CRANIAL ELEMENTS

transport of bonebed elements. The bonebed site is interpreted to be the product of several partly disarticulated Zuni-

Nasal. The most complete Zuniceratops nasal (MSM P3197)

ceratops skeletons of various ages swept a short distance into a

(Fig. 6.2) is smooth and unfused, with no nasal ornamenta-

‘‘logjam’’ at the margin of a flood-stage stream channel (Wolfe

tion. Large partial specimens indicate that the nasal becomes

Zuniceratops christopheri 93

thicker by accumulating layered bone during growth. The narial opening bounded by the nasal and premaxilla is ovate in outline with no narial projection or accessory foramina associated with these elements. The nasal contributes approximately one-third of the dorsal and dorsocaudal narial border with the premaxilla forming the remainder of the border. No premaxillary teeth are present. The ventrocaudal margin of the premaxilla and nasal forms a large accessory antorbital fenestra between that margin and the ascending ramus of the maxilla. Postorbital. Zuniceratops postorbital horncores exhibit marked variation in size and curvature (Fig. 6.3). Zuniceratops horncores are flattened laterally, with vascular grooves on the bone surface. Broken horncores preserve an intracranial sinus at the base of the horn. The well-preserved supraorbital horncore of MSM P2102 (Fig. 6.2) includes the fused palpebral, lacrimal, and the proximal portion of the jugal, and articulates to an associated maxilla (MSM P2102) with teeth. Comparison of the holotype postorbital horncore (Fig. 6.4C–F) to those recovered from the bonebed suggests that the holotype was likely a juvenile or subadult specimen at the time of death. Robust, elongate postorbital horncores have been shown to be plesiomorphic for Ceratopsidae (Ryan 2007). Parietal. The preserved parietal fragments are thin and smooth-margined with no preserved exoccipitals (Fig. 6.2). Two partial proximal squamosals show no evidence of the ‘‘stepped’’ inflection along the squamosal surface noted among most centrosaurines (Avaceratops being the exception; Dodson 1986; Ryan 2007; Kirkland and DeBlieux this volume). A partial proximal parietal is notable for its inverted ‘‘T’’ shape, very similar in appearance to that seen in more basal neoceratopsians such as Protoceratops. The central parietal ramus is thin with a ‘‘T-shaped’’ cross section (Fig. 6.2).

Zuniceratops, postorbital horncore specimens from bonebed and holotype localities. MSM P3812 and MSM P2101 (holotype) are from right side shown in medial view; all others are in left side lateral view. MSM P4179 is a poorly preserved fragment. MSM P4181 was crushed and is shown partly reconstructed. MSM P2103 was crushed along proximal dorsal margin and orbital region. MSM P2102 is associated with maxilla and orbital region of jugal (not shown).

FIGURE 6.3.

Braincase. A well-preserved braincase (MSM P4182; Figs. 6.2, 6.5) from the bonebed has a robust spherical occipital condyle with a constricted neck as in more derived ceratopsids. Later-

than those of basal neoceratopsians and centrosaurines,

ally, the Zuniceratops braincase is long, low, and rectangular

which are generally taller in proportion to length.

compared to that of many ceratopsids. The braincase appears

Dentary. The dentary of Zuniceratops has a relatively ro-

more derived than those of basal neoceratopsians in having

bust, laterally off-set coronoid process with single-rooted

deep fossae ventral and lateral to the occipital condyle, well-

teeth (Fig. 6.2) extending to the caudal margin of the process.

developed basioccipital tubera, and dorsoventrally deep exoc-

In the largest dentaries, the dorsal portion of the coronoid

cipitals. The foramen magnum is roughly teardrop shaped,

becomes thickened and develops an incipient rostral projec-

distinct from the more rounded condition in basal neocera-

tion indicative of the more elaborated projection found in

topsians and the more ovate condition in ceratopsids, due to

ceratopsids. One dentary (MSM P3201) remains articulated

inclusion of the supraoccipital in the dorsal margin of the

with the surangular/articular complex, which is also shallow

foramen. Zuniceratops shares this condition with Avaceratops

but long; these components comprise approximately 20% of

(Dodson 1986; Dodson and Currie 1990; Penkalski and Dod-

the length of the 400 mm long specimen.

son 1999) and juvenile Triceratops (Goodwin et al. 2006;

Teeth. Zuniceratops teeth (Fig. 6.4A, B) range in number from

Horner and Goodwin 2006), but not juvenile Montanoceratops

about 17 to 22 in the smallest to largest dentaries, respectively.

(Chinnery and Weishampel 1998).

There are no more than two replacement tooth rows, and teeth

Maxilla. Zuniceratops maxillae are notably rectangular and

in the middle portion of the dental battery become markedly

long in lateral view, resembling those of chasmosaurines more

larger in larger specimens. Replacement teeth viewed in bro-

94 wolfe, kirkland, smith, poole, chinnery-allgeier, & mcdonald

Holotype horncore and teeth of Zuniceratops. (A) Maxillary tooth; (B) dentary tooth; horncore in (C) caudal, (D) medial, (E) left lateral, and (F) rostral views.

FIGURE 6.4.

FIGURE 6.5.

Zuniceratops braincase MSM P4182 in (A) right lateral, (B) left lateral, (C) cranial, and (D) caudal views.

ken jaws erupt between functional teeth creating a shelf below

femora, a tibia, a fibula, and a few isolated unguals and

the crown and a nearly double-rooted condition in the largest

metatarsals. Several dorsal and many caudal vertebrae were

loose teeth.

collected as was a partial sacrum. Caudal vertebrae possess notably tall neural spines comparable to those of basal neoceraPOSTCRANIA

Zuniceratops appendicular elements are rare. Well-preserved coracoids and scapulae indicate that the glenoid fossa in Zuni-

topsians such as Protoceratops (Dodson et al. 2004).

Discussion

ceratops was directed relatively ventrally, normal to the plane

The new Zuniceratops elements described here help eluci-

of the scapula. The rectangular scapulae are relatively robust,

date significant ontogenetic and individual variation in the

with a prominent ‘‘spine,’’ but the humerus is relatively slen-

jaws and horncores as previously noted in other ceratopsids

der compared to that of ceratopsids. One well-preserved is-

(Sampson et al. 1997). In relation to the larger specimens from

chium was found, and is robust and strongly curved cranially

the bonebed, the holotype exhibits a comparably smaller

as in chasmosaurines. Also collected were two partly crushed

brow horn relative to the orbit, a shorter dentary, and a less

Zuniceratops christopheri 95

FIGURE 6.6.

Zuniceratops reconstructions. (A) By Robert Gaston based on cast material and reconstruction of missing elements; (B) by Lukas Panzarin showing known skeletal elements (shaded).

robust coronoid process, and so resembles the juvenile con-

Neoceratopsian teeth from eastern (late Aptian) and west-

dition documented for Triceratops (Goodwin et al. 2006).

ern (Albian-Cenomanian) North America demonstrate that

Zuniceratops appears derived in having robust, elongate post-

neoceratopsians were well established in North America by

orbital horncores, a reinforced braincase, vertically shearing

the beginning of the late Cretaceous (Chinnery et al. 1998).

teeth, and other ceratopsid characters. The reconstructed

An indeterminate centrosaurine specimen from the Menefee

skull of Zuniceratops (Fig. 6.6) is long, low, and laterally com-

Formation of New Mexico extends the temporal ranges of

pressed with prominent brow horns, and superficially re-

Centrosaurinae and Ceratopsidae into the early Campanian

sembles a typical chasmosaurine skull. Several incipient cera-

(Williamson 1997). Zuniceratops is currently the oldest (Turo-

topsid characters (teeth becoming quite large and nearly

nian) ceratopsian in North America exhibiting elaborate post-

double-rooted, rostral expansion of the coronoid, intracranial

orbital horncores. New centrosaurines, such as the new Wah-

sinuses) are expressed only in the largest Zuniceratops speci-

weap Formation centrosaurine (Kirkland and DeBlieux this

mens suggesting that these characters have been retained by

volume) and Albertaceratops (Ryan 2007), indicate that large

ceratopsids. The strongly curved ischium and robust scapulae

brow horns are plesiomorphic for Ceratopsidae. The phylo-

are derived compared with those of basal neoceratopsians, but

genetic hypothesis of Wolfe and Kirkland (1998) suggesting

the appendicular hypertrophy present in larger ceratopsids is

that Zuniceratops is the sister group to the Ceratopsidae is sup-

not apparent in Zuniceratops. Plesiomorphic features include

ported by recent studies (Chinnery 2004; Dodson et al. 2004;

the single-rooted teeth and the simple fenestrate and un-

Makovicky and Norell 2006; Ryan 2007).

adorned parietal resembling that of Protoceratops. Accessory antorbital openings are known in Zuniceratops, Bagaceratops

Acknowledgments

(Maryanska ´ and Osmólska 1975), Magnirostris (You and Dong 2003), and the new Wahweap Formation centrosaurine (Kirk-

We would like to thank the members of the ‘‘Zuni Basin Pale-

land and DeBlieux this volume), indicating that the presence

ontological Project,’’ including Phil Platt, Sterling Nesbitt, and

of such an opening might be a synapomorphy of a clade in-

Michael Grenier. Harold and Phyllis Bollan, and their assis-

cluding some basal neoceratopsians and Ceratopsidae.

tants, are specially credited for thousands of hours of prepara-

The squamosals, although incomplete, do not appear to ex-

tion work. Rob Gaston (Gaston Designs) contributed research

pand or possess a stepped suture with the parietal as in most

casts and an excellent mount of the skeleton. Lukas Panzarin

centrosaurines (Ryan 2007).

is thanked for permission to use his superb skeletal recon-

96 wolfe, kirkland, smith, poole, chinnery-allgeier, & mcdonald

struction of Zuniceratops. Thomas Robira, Hazel Wolfe, Chris Wolfe, and Susan Bolander supervised field mapping, data management, and figures and maps used in this paper. Observations regarding the braincase are contributed, in part, from a senior thesis by Karen Poole at the University of Pennsylvania and Kent Sanders’s CT images made at the University of Utah. Several of the specimens described in this paper were collected under BLM Paleontological Resource Use Permit MSM-8172-RS-1A and with the assistance of our friends in the ranching community. Special thanks to Mike O’Neil, Patricia Hester, and Kevin Carson for assistance with BLM permitting and access. The Mesa Southwest Museum, Southwest Foundation, Southwest Paleontological Society, the Museum Guild, Dinamation International Society, and Discovery Communications are thanked for significant financial and volunteer contributions. We are also indebted to Tracy Ford, Louis Rey, Eric Baker, Carolyn Staehle, Greg Wenzel, Cliff Greene, Francois Gautier, and the other artists and sculptors who helped bring Zuniceratops to the public eye. We thank Philip Currie and Michael Ryan for reviewing an early draft of this paper. References Cited Chinnery, B. 2004. Description of Prenoceratops pieganensis gen. et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Chinnery, B., T. R. Lipka, J. I. Kirkland, J. M. Parrish, and M. K. Brett-Surman. 1998. Neoceratopsian teeth from the Lower to Middle Cretaceous of North America. New Mexico Museum of Natural History and Science Bulletin 14: 297–302. Chinnery, B., and D. B. Weishampel. 1998. Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) and relationships among basal neoceratopsians. Journal of Vertebrate Paleontology 18: 569–585. Denton, R., S. Nesbitt, D. G. Wolfe, and T. Holtz. 2004. A new small theropod dinosaur from the Moreno Hill Formation (Turonian, Upper Cretaceous) of New Mexico. Journal of Vertebrate Paleontology 24(3, Suppl.): 52A. Dodson, P. 1986. Avaceratops lammersi: A new ceratopsid from the Judith River Formation of Montana. Proceedings of Academy of Natural Sciences of Philadelphia 138: 305–317. Dodson, P., and P. J. Currie. 1990. Neoceratopsia. In D.B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 593–618. Berkeley: University of California Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian. 2006. The smallest known Triceratops skull: New observations on ceratopsid cranial anatomy and ontogeny. Journal of Vertebrate Paleontology 26: 103–112.

Horner, J. R., and M. B. Goodwin. 2006. Major cranial changes during Triceratops ontogeny. Proceedings of the Royal Society B 273: 2757–2761. Kirkland, J. I., and D. D. DeBlieux. 2010. New basal centrosaurine ceratopsian skulls from the Wahweap Formation (Middle Campanian), Grand Staircase–Escalante National Monument, southern Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 117–140. Bloomington: Indiana University Press. Kirkland, J. I., and D. G. Wolfe. 2001. First definitive therizinosaurid (Dinosauria: Theropoda) from North America. Journal of Vertebrate Paleontology 21: 410–414. Makovicky, P. J., and M. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria, Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Maryanska, ´ T., and H. Osmólska. 1975. Protoceratopsidae (Dinosauria) of Asia. Palaeontologica Polonica 33: 133–182. McDonald, A. T., D. G. Wolfe, and J. I. Kirkland. 2006. Preliminary observations on an iguanodontid dinosaur from the Zuni Basin, New Mexico. Late Cretaceous Vertebrates from the Western Interior. New Mexico Museum of Natural History and Science Bulletin 35: 277–279. Molenaar, C. M., W. A. Cobban, E. A. Merewether, C. L. Pilmore, D. G. Wolfe, and J. M. Holbrook. 2002. Regional stratigraphic cross section of Cretaceous rocks from east-central Arizona to the Oklahoma pan handle. U.S. Geological Survey Miscellaneous Field Studies Map MF-2382 (Sheets 1–3). Penkalski, P., and P. Dodson. 1999. The morphology and systematics of Avaceratops, a primitive horned dinosaur from the Judith River Formation (Late Campanian) of Montana, with the description of a second skull. Journal of Vertebrate Paleontology 19: 692–711. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, Southeastern Alberta. Journal of Paleontology 81: 376–396. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Williamson, T. E. 1997. A new Late Cretaceous (Early Campanian) vertebrate fauna from the Allison Member, Menefee Formation, San Juan Basin, New Mexico. New Mexico Museum of Natural History and Science Bulletin 11: 51–59. Wolfe, D. G. 2000. New information on the skull of Zuniceratops christopheri, a neoceratopsian dinosaur from the Cretaceous Moreno Hill Formation, New Mexico. In S. G. Lucas and A. B. Heckert, eds., Dinosaurs of New Mexico. New Mexico Museum of Natural History and Science Bulletin 17: 93–94. ———. 2006. Theropod dinosaur tracks from the Late Cretaceous (Turonian) Moreno Hill Formation of New Mexico. Late Cretaceous Vertebrates from the Western Interior. New Mexico Museum of Natural History and Science Bulletin 35: 115–118. Wolfe, D. G., S. Beekman, D. McGuiness, T. Robira, and R. Denton. 2004. Taphonomic characterization of a Zuniceratops bone

Zuniceratops christopheri 97

bed from the Middle Cretaceous (Turonian) Moreno Hill Formation. Journal of Vertebrate Paleontology 24(3, Suppl.): 131A. Wolfe, D. G., and J. Kirkland. 1998. Zuniceratops christopheri n. gen. & n. sp., a ceratopsian dinosaur from the Moreno Hill Formation (Cretaceous, Turonian) of west-central New Mexico.

New Mexico Museum of Natural History and Science Bulletin 14: 303–317. You, H., and Z. Dong. 2003. A new protoceratopsid (Dinosauria: Neoceratopsia) from the Late Cretaceous of Inner Mongolia, China. Acta Geologica Sinica 77: 299–303.

98 wolfe, kirkland, smith, poole, chinnery-allgeier, & mcdonald

7 Horned Dinosaurs (Ornithischia: Ceratopsidae) from the Upper Cretaceous (Campanian) Cerro del Pueblo Formation, Coahuila, Mexico MARK A. LOEWEN, SCOTT D. SAMPSON, ERIC K. LUND, ANDREW A. FARKE, MARTHA C. AGUILLÓN-MARTÍNEZ, CLAUDIO A. DE LEON, R U B É N A . R O D R Í G U E Z - D E L A R O S A , M I C H A E L A . G E T T Y, A N D D AV I D A . E B E R T H

a collaborative fieldwork project conducted in the

rhinoceratops) and large parietal fenestra (distinct from

Upper Cretaceous Cerro del Pueblo Formation, southern

Anchiceratops and Arrhinoceratops, yet shared with Chas-

Coahuila, Mexico, has yielded remains of at least two

mosaurus, Pentaceratops, and Agujaceratops). Phylogenetic

taxa of ceratopsid dinosaurs from three localities. Al-

analysis places Coahuilaceratops in an unresolved poly-

though fragmentary and incomplete, these ceratopsid

tomy with Anchiceratops, Arrhinoceratops, and other more

specimens provide important insights into the diversity

derived Maastrichtian chasmosaurines.

of dinosaurs in the southernmost portion of North America during the late Campanian. A partial squamosal with a stepped squamosal-parietal contact (CPC 279)

Introduction

represents the mostly southerly occurrence of a cen-

The horned dinosaur clade Ceratopsidae represents a mono-

trosaurine. An elongate, dorsoventrally compressed, and

phyletic radiation of Late Cretaceous megaherbivores. Evi-

laterally projecting supraorbital horncore with a ven-

dence of the 20 million year ceratopsid radiation is restricted

trally deflected tip (CPC 278) recovered from another lo-

almost exclusively to sediments deposited within the Western

cality may pertain either to Centrosaurinae or

Interior Basin (WIB) of North America. A series of dinosaur-

Chasmosaurinae.

bearing formations were deposited along the eastern margin

The third locality has produced multiple elements of a

of Laramidia, the western landmass produced when the epi-

chasmosaurine ceratopsid that are attributable to at least

eric Late Cretaceous Interior Seaway flooded the central re-

two individuals—a juvenile (CPC 277) and an adult (CPC

gion of North America. The ceratopsid record of the northern

276). The adult materials, postulated as pertaining to a

part of the WIB (mostly Alberta and Montana) is relatively

single individual, include several skull elements that per-

abundant compared to the south (Utah, Colorado, New Mex-

mit erection of a new taxon: Coahuilaceratops

ico, Texas, Mexico). For example, the ceratopsid clade (sub-

magnacuerna n. gen. and sp. Autapomorphies of

family) Centrosaurinae is known from dozens of skulls and

Coahuilaceratops include premaxillary narial strut with

skeletons in the north, many found within extensive mono-

twisted cross-sectional shape; and nasal bearing greatly

dominant bonebeds (Ryan 2007; Sampson and Loewen 2007,

thickened cross section. Other salient features of the new

this volume). In contrast, only a handful of isolated centro-

Cerro del Pueblo chasmosaurine include a caudally in-

saurine specimens have been recovered in the south (Heckert

clined narial strut (shared with Anchiceratops and Ar-

et al. 2003; Kirkland and DeBlieux 2006, 2007, this volume;

99

Sampson and Loewen 2007, this volume). A similar pattern

Cevallos-Ferriz 1994, 1998; Hernández et al. 1995; Rodríguez-

applies to Chasmosaurinae, although the southern record of

de la Rosa 1996; Rodríguez-de la Rosa et al. 1998; Kirkland et

these long-frilled ceratopsids is more extensive and includes

al. 2000; Kirkland and Aguillón-Martinez 2002; Rodríguez-de

several named taxa: Pentaceratops Osborn 1923, Agujacera-

la Rosa et al. 2002; Eberth et al. 2003, 2004; Gates et al. 2007).

tops Lucas et al. 2006, and Torosaurus Marsh 1891. Two of

Although dinosaurian materials have been known from the

these monospecific genera, Pentaceratops and Agujaceratops,

Cerro del Pueblo Formation for some time (see reviews in Her-

are of late Campanian age, and known only from the southern

nández et al. 1995 and Kirkland et al. 2000), ceratopsids have

region of the WIB (Forster et al. 1993; Lehman 1989, 1993,

remained enigmatic, restricted to a putative report (Murray

1998; Lucas et al. 2006). The genus Agujaceratops was recently

et al. 1960) of the nomen dubium centrosaurine Monoclonius

erected to replace the only southern species of Chasmosaurus

(Sampson et al. 1997). The materials associated with this re-

Lambe 1914, C. mariscalensis Lehman 1989 (Lucas et al. 2006).

port have yet to be relocated, and subsequent attempts to relo-

Late Campanian chasmosaurines in the north are currently

cate the site have also been unsuccessful ( J. Kirkland pers.

limited to three (or four, see below) taxa within the Dinosaur

com.). Although Janensch (1926) reported fragmentary cera-

Park Formation, and two from the Horseshoe Canyon Forma-

topsian material from the Maastrichtian of Coahuila and as-

tion of southern Alberta. The Dinosaur Park Formation in-

signed the material to Monoclonius, no ceratopsian materials

cludes three stratigraphically separated species of Chasmosau-

have been described or formally named from the Cerro del

rus (Ryan and Evans 2005): Chasmosaurus belli Lambe 1914,

Pueblo Formation.

C. russelli Sternberg 1940, and C. irvinensis Holmes et al. 2001.

An international collaborative involving three North Amer-

The Campanian portion (i.e., below the Drumheller Marine

ican institutions—the Utah Museum of Natural History (Salt

Tongue) of the Horseshoe Canyon Formation includes Anchi-

Lake City, Utah), the Secretaría de Educación y Cultura

ceratops ornatus Brown 1914 and Arrhinoceratops brachyops

through the Museo del Desierto (Saltillo, Coahuila), and the

Parks 1925. Langston (1959, 1975) refers specimens from

Royal Tyrrell Museum of Palaeontology (Drumheller, Alberta)

what is now recognized as the Dinosaur Park Formation near

—initiated the Parras Basin Dinosaur Project in 2002, with

Manyberries, Alberta, to Anchiceratops; however, these speci-

continuing fieldwork in 2003. The primary objectives of this

mens are fragmentary and a definitive assignment to Anchi-

project have been paleontological and geological exploration

ceratops is speculative (see discussion).

of the Late Cretaceous (Campanian) Cerro del Pueblo Forma-

Here we report on new ceratopsid materials recovered from

tion in the Parras Basin, southern Coahuila, Mexico. Results

the Upper Cretaceous (late Campanian) Cerro del Pueblo

of the project have revised the lithostratigraphy of the lower

Formation in southern Coahuila, Mexico. The Cerro del

Difunta Group and applied magnetostratigraphic analysis to

Pueblo Formation comprises the basalmost unit of the multi-

demonstrate that the Cerro del Pueblo Formation dates to the

kilometer-thick Difunta Group (Fig. 7.1), which was deposited

latest Campanian (Eberth et al. 2004). Remains of at least two

along with the underlying Parras Shale eastward and north-

species each of hadrosaur (e.g., the lambeosaurine Velafrons

ward of the Sierra Madre Oriental fold-and-thrust belt (Mur-

coahuilensis; Gates et al. 2007) and ceratopsid (horned) dino-

ray et al. 1962; Weide and Murray 1967; McBride et al. 1974;

saurs, as well as fish, turtles, crocodiles, lizards, snakes, and

Soegaard et al. 1997; Kirkland et al. 2000; Eberth et al. 2004).

mosasaurs, together with dinosaur eggshell and abundant

The Cerro del Pueblo Formation represents lower coastal plain

trackways, have also been described (Brinkman et al. 2002;

and shallow marine sediments that were subject to high-

Eberth et al. 2003, 2004; Gates et al. 2007; Rodríguez-de la

frequency sea-level fluctuations and coastal storm events

Rosa et al. 2003).

(Eberth et al. 2004). Westward thickening of the formation,

These expeditions also recovered specimens of at least two

from 162 m at Saltillo to 540 m at Porvenir de Jalpa (70 km west

ceratopsid taxa from three localities within the Cerro del

of Saltillo), is thought to represent an increased rate of sub-

Pueblo Formation (Eberth et al. 2003; Lund et al. 2007). These

sidence and accommodation in that direction (Eberth et al.

materials include representative materials of both Chasmo-

2004). Magnetostratigraphic and biogeographic data suggest

saurinae and Centrosaurinae, with the former being sufficient

that the sediments preserved in the Cerro del Pueblo Forma-

to erect a new taxon, the first chasmosaurine genus and species

tion were deposited between 72.5 Ma and 71.4 Ma, with a

described from Mexico. The Cerro del Pueblo ceratopsid lo-

maximum absolute age of 72.5 Ma for the formation (Eberth et

calities are located near the towns of Presa San Antonia and

al. 2004).

Porvenir de Jalpa, approximately 70 km west of Saltillo (Fig.

A combination of vertebrate, invertebrate, and paleobotani-

7.1). The contact between the Cerro del Pueblo Formation and

cal remains recovered from the Cerro del Pueblo Formation

the underlying Parras Shale is not exposed in the area; how-

has dramatically expanded the paleontological record of the

ever, the stratigraphic positions of the sites were estimated by

southernmost part of WIB during the Late Cretaceous (Wolle-

measuring down from the overlying Cerro Huerta beds, plac-

ben 1977; Vega and Feldmann 1991; Rodríguez-de la Rosa and

ing the ceratopsian material in the middle (approximately 230

100 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

Latest Campanian paleogeographic relationships of northern Mexico (modified after Robinson-Roberts and Kirschbaum 1995; Lehman 1997, 2001; Scotese 2001; and Eberth et al. 2004). The Parras, La Popa, and Sabinas basins and the stratigraphic relationships of the basins in southern Coahuila, Mexico, are indicated, showing locations of ceratopsid localities.

FIGURE 7.1.

Horned Dinosaurs from the Upper Cretaceous Cerro del Pueblo Formation 101

m level) of the Cerro del Pueblo Formation. Given that the

tion to being anatomically correct, a distinct advantage of this

Cerro del Pueblo Formation dates to between approximately

scheme is that it follows the convention of other named der-

72.5 and 71.4 Ma, we estimate the age of these remains at about

mal ossifications on the ceratopsid skull, the epinasal and epi-

72.0 Ma. The ceratopsid specimens were recovered from fluvial

jugal. The single complication is the presence of a marginal

coastal plain deposits within the formation. In this paper we

ossification in centrosaurines and in Triceratops that bridges

describe the salient morphologic features of these specimens

the squamosal-parietal contact. When referring to this particu-

and, where possible, place them into phylogenetic context.

lar element, we propose that the more general term ‘‘parieto-

Institutional Abbreviations. CPC: Colección Paleontológica

squamosal marginal ossification’’ be applied. In support of this usage, the dermal ossification at the parietal-squamosal con-

de Coahuila, Saltillo. Anatomical Abbreviations. ao: midline parietal accessory ossifications; EN: ectonaris; ENDO: endonaris; es1–4: epis-

tact differs from other such ossifications in that it typically does not attach to a marginal undulation or scallop.

quamosal positions; f: frontal contact; jp: jugal process (quadratojugal/jugal contact); lac: lacrimal contact; LTF: lateral temporal fenestra; no: nasal ossification; ns: narial strut; O: orbit; P: parietal; PAL: palpebral; PF: parietal fenestra; pm: premaxillary process (premaxilla/nasal contact); pob: accessory post orbital boss; qc: quadrate cotylus; SQ: squamosal.

Systematic Paleontology Ornithischia Seeley 1888 Ceratopsia Marsh 1890 Ceratopsidae Marsh 1888 Ceratopsidae, gen. et sp. indet.

Materials and Methods MATERIALS

New Material. CPC 278, a partial left orbit with supraorbital horncore from a juvenile ceratopsid (Fig. 7.2A–C). Locality. CPC 278 was recovered in the Parras Basin near the

The new ceratopsid materials described herein were recovered

town of Presa San Antonia, approximately 70 km west of Sal-

from three localities within the Cerro del Pueblo Formation

tillo, Coahuila, Mexico (Fig. 7.1). Exact locality information is

(Fig. 7.1). One site has produced isolated specimens: a partial

on file at the Museo del Desierto, Saltillo, Coahuila.

right centrosaurine squamosal (CPC 279; Fig. 7.2D) and an-

Horizon. The locality is located in the middle part (approxi-

other a partial postorbital of uncertain affinity (CPC 278; Fig.

mately 230 m level) of the Campanian Cerro del Pueblo For-

7.2A–C). The third locality consists of a monodominant

mation, the basal formation of the Difunta Group.

bonebed containing the remains of at least two ceratopsid

Comments. Only the dorsal and caudal portion of the orbit is

individuals—one adult and one juvenile. This bonebed as-

preserved, including a well developed supraorbital horncore.

semblage, including both craniofacial and postcranial ele-

The preserved orbital margin is formed by the coossified post-

ments, appears to pertain to a single chasmosaurine taxon.

orbital and a blocky palpebral, forming a well developed

Based on relative size, surficial bone texture, state of preserva-

antorbital buttress (sensu Sampson 1995). The laterally proj-

tion, and elements preserved (e.g., partial left and right su-

ecting horncore is elongate (280 mm) and dorsoventrally

praorbital horncores of adult size), we have parsimoniously

compressed (rostrocaudal diameter approximately 80 mm;

segregated the material into two specimen numbers—CPC

dorsoventral diameter 40 mm), with a dorsally inflected tip. A

276 for the adult and CPC 277 for the juvenile.

distinct process that appears to be an accessory ossification

Anatomical Terminology. The methods used in this study are

(postorbital boss in Fig. 7.2A, B) occurs caudal to the horncore

standard for anatomical description and phylogenetic analysis

on the caudodorsal rim of the orbit. The lateroventral surface

of fossil vertebrates. Morphological assessments and character

of the horncore bears elongate sulci typical of ceratopsids. As

codings were based on firsthand observations of specimens,

in other ceratopsids, the articular surface for the laterosphe-

augmented with the primary literature where necessary. Ana-

noid occurs medial to the caudal third of the orbit. A small

tomical nomenclature for marginal ossifications of the parie-

portion of the supracranial cavity (i.e., frontal sinus complex

tosquamosal frill follows the system first proposed by Hatcher

of some authors) is preserved. The walls of this cavity are

et al. (1907) and most recently advocated by Forster and Samp-

smooth, and there is minimal evidence of cornual sinuses

son (2002) and by Goodwin and Horner (2008). Specifically,

(sensu Farke 2006) that extend into the base of the horncore.

marginal ossifications on the squamosal and parietal of cera-

As preserved, the supracranial cavity measures 59 mm long

topsids are referred to as ‘‘episquamosals’’ and ‘‘epiparietals,’’

and approximately 46 mm deep. The preserved portion is lo-

respectively. In referring to these elements as a group, we pro-

cated entirely rostral to the laterosphenoid contact and di-

pose the phrase ‘‘marginal ossifications of the frill.’’ In addi-

rectly medial to the horncore base.

102 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

FIGURE 7.2.

Ceratopsidae indet. postorbital (CPC 278) and squamosal (CPC 279). Postorbital in (A) ventral, (B) right lateral, and (C) rostral views. (D) Squamosal in right lateral view.

Centrosaurinae Lambe 1915

The element exhibits the imbricated lateral border (Samp-

Centrosaurinae gen. et sp. indet.

son et al. 1997) and stepped squamosal-parietal contact characteristic of all centrosaurines (Ryan 2007) except Avacera-

New Material. CPC 279, a partial right squamosal (Fig. 7.2D).

tops (Dodson 1986). Without further material, the taxo-

Locality. CPC 279 was recovered within the Parras Basin near

nomic placement of this element cannot be refined beyond

the town of Rincon Colorado, west of Saltillo, Coahuila, Mex-

Centrosaurinae.

ico. Exact locality information on file at the Museo del DeChasmosaurinae Lambe 1915

sierto, Saltillo, Coahuila. Horizon. The locality is located in the middle facies (approximately 230 m level) of the Campanian Cerro del Pueblo Formation, the basal Formation of the Difunta Group.

Coahuilaceratops magnacuerna gen. et sp. nov. Holotype. CPC 276, disarticulated elements representing a single large adult individual (Figs. 7.3–7.7), including the ros-

Comments. The partial squamosal preserves part of the lat-

tral, left premaxilla, right maxilla, fused nasals, partial left and

eral temporal fenestra, and episquamosals es2 through es4.

right supraorbital horncores, part of the parietosquamosal

Horned Dinosaurs from the Upper Cretaceous Cerro del Pueblo Formation 103

frill, predentary, both dentaries, and unprepared postcranial

(Sampson et al. 1997). Together with the smaller, putative ju-

elements.

venile specimen (CPC 277), these materials represent approxi-

Referred Material. CPC 277, disarticulated elements repre-

mately 45% of the external craniofacial skeleton and lower

senting a single juvenile individual, including a predentary,

jaw (Fig. 7.3). To date, lower level (genus and species) tax-

dentary, and unprepared postcranial elements.

onomic resolution of ceratopsids has been based almost exclu-

Etymology. The first portion of the generic name, ‘‘Coa-

sively on craniofacial materials, in particular relating to the

huila,’’ refers to the state of discovery in northeast Mexico,

skull roof (Dodson et al. 2004). Thus, the following descrip-

whereas the latter part of the name, ‘‘ceratops,’’ is Greek

tion, derived largely from the holotype specimen, is limited to

for ‘‘horned face.’’ The specific name magnacuerna combines

key aspects of skull and lower jaw anatomy. The description

the Latin word ‘‘magna,’’ meaning ‘‘great,’’ with the Spanish

is divided into three sections—circumnarial region, circum-

‘‘cuerna,’’ meaning ‘‘horn,’’ in reference to the very large su-

orbital region, and parietosquamosal frill—each of which

praorbital horncores of this taxon.

focuses on characters useful for resolving within-group rela-

Diagnosis. Large chasmosaurine ceratopsid possessing the

tionships of chasmosaurines.

following autapomorphies: premaxilla with caudally inclined premaxillary strut bearing a counterclockwise twist and a pronounced ridge (Fig. 7.4D–F); co-ossified nasals with extremely

CIRCUMNARIAL REGION

thick dorsoventral cross section relative to the condition seen

The circumnarial region is dominated by the nasal and pre-

in other chasmosaurines (Fig. 7.5E, F). The species has a

maxilla, but is here interpreted to include the rostral and max-

unique combination of rostrally placed nasal horncore rela-

illa as well. In addition, for ease of interpretation, the apomor-

tive to the external naris, and a caudally inclined premaxillary

phic predentary and dentary are also described in this section.

strut. The taxon differs from Chasmosaurus, Pentaceratops, and

Overall, the narial region of Coahuilaceratops is relatively elon-

Agujaceratops in the direction of the premaxillary strut, differs

gate as in other chasmosaurines, but differs from Chasmosau-

from Agujaceratops and Chasmosaurus in the immense size of

rus in being deep dorsoventrally.

the postorbital horncores and the presence of an enlarged cor-

Rostral. Only the dorsal portion of the nasal process of the

nual sinus within the horns, and differs from Agujaceratops in

rostral (approximately 125 mm long) is preserved (Fig 7.4A–

the cross-sectional shape of the nasal horncore. The species

C). It exhibits the rugose external surface texture typical of

differs from Triceratops, Torosaurus, Diceratops, Anchiceratops,

ceratopsid rostrals. The element resembles other chasmosau-

and Arrhinoceratops in the enlarged parietal fenestrae and

rines in being transversely narrow, whereas in centrosaurines,

greatly enlarged anterior episquamosals, and it differs from

the nasal process of the rostral tends to be relatively broad

Triceratops, Torosaurus, and Diceratops in the inclination of the

dorsally. In contrast to virtually all other chasmosaurines,

premaxillary strut.

the dorsalmost part of the preserved nasal process is straight

Locality. Holotype and referred materials were recovered

rather than caudally curved.

from a monodominant bonebed of two individuals in the

Premaxilla. Only the caudoventral portion of the left pre-

Parras Basin near the town of Porvenir de Jalpa, approximately

maxilla is preserved (Figs. 7.3, 7.4D–G). However, this ele-

70 km west of Saltillo, Coahuila, Mexico. Exact locality infor-

ment is distinctive in several respects. The ventral margin is

mation on file at the Museo del Desierto, Saltillo, Coahuila.

cupped dorsally, bearing the rounded lateral margin typical of

Horizon. The locality occurs in the middle facies (approxi-

ceratopsids. The ascending process, which contacts the max-

mately 230 m level) of the Campanian Cerro del Pueblo For-

illa, is incompletely preserved, with the preserved rostral end

mation, the basal formation of the Difunta Group.

of its long axis directed dorsolaterally. The ventromedial surface has a prominent groove that continues to the caudo-

Description of Coahuilaceratops magnacuerna (CPC 276 and CPC 277)

medial surface of the premaxillary strut. As in other ceratopsids, the medial surface of the premaxilla is flattened for contact with the opposing premaxilla.

This section describes the holotype materials of Coahuila-

The most distinctive aspect of the premaxilla is the pre-

ceratops magnacuerna (CPC 276), as well as a single juvenile

maxillary strut, which is a thickened ridge of bone that oc-

specimen (CPC 277) referred to this taxon. All other cera-

curs along the caudal margin of the premaxillary septum in

topsid materials recovered from the Cerro del Pueblo Forma-

chasmosaurines. In dorsal view, the premaxillary strut of

tion (CPC 278 and 279) are insufficient to warrant detailed

Coahuilaceratops has a pronounced counterclockwise twist

description.

from ventral to dorsal that is accentuated by a groove on the

CPC 276 is interpreted as an adult individual of Coahuila-

rostral surface (probably representing the caudal extent of the

ceratops magnacuerna based on relative size, fusion of skull ele-

septal fossa). Overall, the premaxillary strut has a rostrocau-

ments, vertebral suture closure, and bone surface texture

dally elongate cross section, particularly in the ventralmost

104 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

FIGURE 7.3.

Skull reconstructions (A and B) of Coahuilaceratops magnacuerna n. gen. and sp. based largely on adult specimen CPC 276. Rostrum, maxilla, and squamosal are photo-reversed from the right side (stipple art by Lukas Panzarin).

portion. No other chasmosaurine has this ‘‘twisted’’ strut mor-

of the nasal to adjacent bones cannot be determined. The dor-

phology. The caudally inclined orientation of the narial strut

sal surface of the nasal has a broadly V-shaped lateral profile,

is otherwise found only in Chasmosaurus irvinensis, Anchicera-

with a gently rounded apex along most of its length leading

tops and Arrhinoceratops.

up to the orbital region. The nasals are unusual among chas-

Nasal. The holotype includes completely co-ossified right

mosaurines in their dorsoventral depth (maximum 89 mm

and left nasals bearing a nasal horncore (Fig. 7.5). The external

just caudal to the nasal horn) (Fig. 7.5E, F). This dorsoven-

surface of the entire complex is deeply incised with neuro-

trally expanded condition is present throughout most of the

vascular impressions, most of which trend rostrocaudally. The

length of the nasal, transitioning to 9 mm only in the most

nasal measures 178 mm across at the dorsal border of the

caudal portion of the nasal (Fig. 7.5E, F). A broad, projecting

naris. The element is moderately saddle-shaped, with the cau-

ridge marks the ventral limit of the internasal contact. The

dal end sloping dorsally in lateral view to contact the orbital

contact with the premaxillae is not well preserved, although it

region (Fig. 7.5E, F). Due to extensive fusion, the relationship

appears to indicate that the nasals wrapped around the pre-

Horned Dinosaurs from the Upper Cretaceous Cerro del Pueblo Formation 105

tops, Anchiceratops, Agujaceratops, Pentaceratops, and Triceratops, and contrasts with that of Chasmosaurus, in which the horncore is placed more caudally. Assuming that an epinasal ossification is present, as in other chasmosaurines, the structure is completely fused to the nasals, with no visible sutures. The rostrodorsal portion is much more rugose than the rest of the nasal horncore, perhaps indicative of a separate ossification. As measured from the top of the external naris, the horncore is 67 mm in height. Maxilla. A single incomplete right maxilla is preserved, including the body, a portion of the ascending process, and a portion of the alveolar process (Fig. 7.6A, B). A prominent maxillary cavity occurs in the body of the maxilla, although its rostralmost extent has not yet been prepared. Between the alveolar process and the ascending process, a broad and shallow fossa occurs on the lateral surface of the maxilla. Two large foramina pierce the rostral end of this fossa. The lateral surface of the maxilla is only superficially incised with neurovascular impressions, typical of ceratopsids (in contrast, for example, with the deep impressions seen on the surface of the supraorbital horncores). The alveoli in the maxilla are poorly preserved and infilled with matrix, prohibiting even an approximate count of alveoli. Overall, this element is typical of chasmosaurines other than species of Chasmosaurus, which tend to possess relatively shorter maxillae. Predentary. A single predentary fragment is preserved with the holotype materials (CPC 277; Fig. 7.6C). It possesses the typical rugose surface texture of ceratopsid predentaries, but little else can be ascertained. A second possible fragment from a juvenile individual (CPC 277; Fig. 7.6D, E) is also preserved, but it displays no diagnostic or distinctive morphology. Dentary. Two dentaries lacking teeth—one right (CPC 277; Coahuilaceratops magnacuerna rostrum and left premaxilla (CPC 276). (A) Dorsal ramus of rostrum in right lateral view; (B) dorsal ramus in rostral view; (C) dorsal ramus in caudal view. (D) Left premaxilla in dorsal view; note the caudally directed narial strut and twisted cross section; (E) lateral view (note the caudally directed narial strut and twisted cross section); (F) medial view; (G) caudal view (note the caudally directed narial strut and twisted cross section).

FIGURE 7.4.

Fig. 7.6F, G) and one left (CPC 276; Fig. 7.6H–I)—are preserved in the Cerro del Pueblo bonebed. Based on size, it is clear that these elements pertain to different individuals; the larger (left) dentary is nearly 50% longer than the smaller (right) dentary. The former measures 448 mm long, is 89 mm tall at the deepest portion of the alveolar tooth row, and is 187 mm tall at the deepest portion of the preserved coronoid process. The smaller dentary measures 292 mm long, 117 mm tall at the deepest portion of the alveolar tooth row, and is 158 mm tall as preserved at the deepest portion of the preserved coronoid

maxillae. This condition occurs in Agujaceratops and Pentaceratops, whereas it is reversed in Chasmosaurus (Forster et al. 1993).

process (approximately the distal 10 or 20 mm are broken off). The larger, presumably adult dentary (Fig. 7.6H–I) is the better preserved of the two dentaries and forms the basis for

The nasal horncore is relatively small, low and subtrian-

the description of the element. Typical of ceratopsid material

gular in shape, with an oval cross section at the base and a

recovered from bonebeds, both of the dentaries lack teeth.

rounded tip. Its long axis is directed rostrally relative to the

The larger specimen preserves 18 definitive alveolar positions;

dorsal surface of the nasals, and the base of the horncore is

given that the caudal end is missing, the total alveolar count is

situated approximately at the center of the external naris as

estimated to be between 23 and 25. No alveoli are preserved in

preserved. This conformation is similar to that of Arrhinocera-

the smaller dentary. As in other ceratopsids, the proximal end

106 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

FIGURE 7.5.

Coahuilaceratops magnacuerna fused nasals (CPC 276). (A) Left lateral view; (B) right lateral view; (C) dorsal view; (D) rostral view; (E) vertical cross section along the line indicated in A; (F) oblique (rostroventral) view. Note the extreme dorsoventral thickening in (F).

of the dentary is expanded for reception of the predentary.

CIRCUMORBITAL REGION

The ventral articular surface does not extend beyond the first two or three alveolar positions, so the ventral process of the

Postorbital. The circumorbital region of Coahuilaceratops is rep-

predentary was relatively restricted. The ventral margin of the

resented only by a pair of partial supraorbital horncores

dentary is relatively sharp at its rostral end, becoming ven-

(Fig. 7.6J–K). The horncores, represented by approximately

trally flattened caudally. Surface bone texture on the lateral

300 mm long portions of the proximal midsection—are ellipti-

surface of the dentary, where preserved, is only lightly vas-

cal in cross section and hyper-robust, exhibiting minimal evi-

cularized. As is typical of ceratopsids, the coronoid process is

dence of tapering. As preserved, maximum basal diameter of

offset from the main body of the dentary. A rostrocaudally

both horncores is 155 mm, with a minimum diameter of 120

trending, dorsally directed ridge joins it to the main body of

mm. Both sides exhibit a slight curvature along their length,

the dentary, and a large foramen occurs at its termination. A

although the direction of curvature cannot be determined.

medially directed ridge connects the medial surface of the cor-

The absolute size of these horncores rivals that of the largest

onoid process to the alveoli. The distal end of the coronoid

chasmosaurines (e.g., Triceratops, Torosaurus), and their hyper-

process is projected rostrally, with a rounded rostral margin.

robust morphology (i.e., relative lack of tapering in the distal

Horned Dinosaurs from the Upper Cretaceous Cerro del Pueblo Formation 107

PARIETOSQUAMOSAL FRILL Squamosal. The right squamosal is incomplete, lacking much of the caudal end (Fig. 7.7A, B). The rostrolateral margin is gently convex rather than concave, and the rostral edge forms the caudal and dorsal portions of the otic notch and the caudal portion of the lateral temporal fenestra. The 80 mm wide base of the jugal flange is preserved. As is typical for ceratopsids, the ventral surface possesses well developed, slot-like grooves for the quadrate and paroccipital process. The first and second (i.e., es1, es2) episquamosal ossifications are present and fully fused onto the squamosal margin. Episquamosal es1 is 110 mm long and 55 mm high; es2 is 105 mm long and 40 mm high. Both episquamosals are typical of chasmosaurines. Parietal. Only the rostral end of the parietal midline is preserved (Fig. 7.7C–E). The surface texture of the parietal is poorly preserved; nevertheless, where present, the ventral side is relatively smooth, whereas the dorsal side is covered in well developed neurovascular impressions. The rostral end of the parietal fragment is broadly arched and thickest in the middle, thinning to 12 mm at its lateral extent. A prominent hollow occurs at the rostralmost end ventrally, corresponding to the position directly above the supraoccipital. Two elongate, low bumps occur on the parietal midline over the rostral threefourths of the preserved fragment. Such midline prominences occur commonly on the parietals of centrosaurines, but only rarely in chasmosaurines. In dorsal view, the parietal midline tapers caudally. The parietal bar is ovoid in cross section and Coahuilaceratops magnacuerna jaws (CPC 276 and CPC 277 v[juvenile]). Right maxilla of CPC 276 in (A) right lateral and (B) medial views. (C) Predentary fragment of CPC 276 in medial view. Predentary of CPC 277 in (D) dorsal and (E) ventral views. Right dentary of CPC 277 in (F) right lateral and (G) medial views. Left dentary of CPC 276 in (H) left lateral and (I) medial views. Supraorbital horncores of CPC 276: ( J) left; (K) right.

FIGURE 7.6.

slightly broader dorsally than ventrally, most similar to that of Pentaceratops and Agujaceratops, and differing from the more circular to square cross sections of Chasmosaurus. It measures 45 mm in dorsoventral diameter and 36 mm in mediolateral diameter (Fig. 7.7E). Although the parietal fenestrae are incompletely preserved, it is clear that they were quite large, approaching the size seen in Pentaceratops and most specimens of Chasmosaurus.

Phylogenetic Analysis portion) more closely resembles that of large specimens of

METHODS

Pentaceratops than Agujaceratops. The external surface texture, where preserved, is typical of large chasmosaurine postorbital

A phylogenetic analysis was conducted in order to place Coa-

horncores, exhibiting elongate neurovascular sulci and a large

huilaceratops n. gen. and sp. within Chasmosaurinae (Fig. 7.8).

number of irregularly spaced foramina. One of the horncores

This matrix (Table 7.1) was adapted from that of Wu et al.

preserves the distalmost portion of the cornual sinus at its

2007 (which was modified from the analysis of Dodson et al.

base. The bone on the surface of the sinus is lightly incised

2004). A total of 76 characters were arrayed among 16 ingroup

with subparallel neurovascular impressions, and a low ridge

ceratopsid taxa and two outgroup taxa (Protoceratops and

runs from one side of the sinus to the other. The distal end of

Zuniceratops). Changes from the previously published ma-

the sinus is asymmetrical relative to the horncore, centered off

trices include the separation of Chasmosaurus into Agujacera-

from the midline rather than occurring in the middle of the

tops and Chasmosaurus. Given that the goal of the present

element.

analysis was to assess the phylogenetic affinities of Coahuila-

108 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

RESULTS The tree search identified 109 equally parsimonious trees (L = 97 steps; CI = 0.8660; RI = 0.9212). Fifty-one characters were identified as parsimony informative. All of the trees recovered a monophyletic Chasmosaurinae, with Agujaceratops, Chasmosaurus, and Pentaceratops outside of a clade including the remaining chasmosaurines. All trees recovered a clade comprising Coahuilaceratops, Anchiceratops, Arrhinoceratops, Diceratops, Eotriceratops, Torosaurus and Triceratops. This clade is united by characters 6, 18, 23, 25, 27, 29, and 31. Of these, only characters 6 (narial strut inclined caudally), 18 (nasal horn centered dorsal or rostrodorsal to naris), and 29 (supracranial sinus complex large and confluent with cornual sinuses) can be coded in Coahuilaceratops, and only characters 18 and 29 are unequivocal. This node has poor decay index support, however, with only one additional step needed to remove Coahuilaceratops from that clade. Comments. Recent cladistic analyses of chasmosaurines (e.g., Forster et al. 1993; Holmes et al. 2001; Dodson et al. 2004) have grouped Pentaceratops and Chasmosaurus in a clade that is the sister taxon to all other chasmosaurines. Similarly, although Agujaceratops (Chasmosaurus) mariscalensis has been closely allied with Chasmosaurus and indeed placed within this genus (e.g., Lehman 1989; Dodson and Currie 1990; Forster et al. 1993; Dodson et al. 2004), recent studies (Holmes et al. 2001; Sampson et al. 2004; Lucas et al. 2006) raised the possibility that this West Texas taxon may be more closely allied with another southern chasmosaurine, Pentaceratops. Thus, the polytomy recovered here for Agujaceratops, Chasmosaurus, and Pentaceratops is unsurprising, given the geographic and stratigraphic proximity of these chasmosaurines. The few previous cladistic analyses of Chasmosaurinae (e.g., Dodson et al. 2004; Wu et al. 2007) have generally placed Anchiceratops with Arrhinoceratops as a sister clade to the more Coahuilaceratops magnacuerna frill (CPC 276). Right squamosal in (A) right lateral and (B) caudal views. Parietal in (C) dorsal and (D) left lateral views; (E) crosssectional outline of the distal end of the midline parietal bar.

FIGURE 7.7.

derived Maastrichtian forms Torosaurus, Eotriceratops, Diceratops, and Triceratops. The novel relationship posited here suggests that that clade of [Coahuilaceratops + Anchiceratops + Arrhinoceratops] is more closely related to the derived Maastrichtian chasmosaurines Torosaurus, Eotriceratops, Diceratops, and Triceratops than either are to the clade of [Agujaceratops + Chasmosaurus + Pentaceratops].

ceratops rather than the inter-relationships of all chasmosaurines, Chasmosaurus (C. russelli, C. belli, and C. irvinensis), Tri-

Discussion

ceratops (T. horridus and T. prorsus), and Torosaurus (T. latus and

TAXONOMIC CONSIDERATIONS

T. utahensis) were coded at the generic level. The matrix was analyzed using a branch-and-bound search in PAUP* 4.0b10.

Centrosaurinae. The isolated centrosaurine squamosal recov-

A bootstrap analysis was also performed using the same al-

ered from the Cerro del Pueblo Formation (CPC 279) lacks

gorithm, with 1,000 replicates, and decay indices were also

any characteristics that permit taxonomic assessment beyond

calculated.

Centrosaurinae. Similarly, although the isolated supraorbital

Horned Dinosaurs from the Upper Cretaceous Cerro del Pueblo Formation 109

Stratigraphic calibrated strict consensus tree from parsimony analysis of data matrix for Ceratopsidae (after Wu et al. 2007). The values at the nodes indicate the number of supporting characters, with the decay index (as an integer) and the bootstrap index (as a decimal). Black range boxes represent known ranges, whereas white boxes represent the possible range for poorly constrained specimens. See Sampson and Loewen (this volume) for a review of stratigraphic ranges of ceratopsids.

FIGURE 7.8.

horncore (CPC 278) is autapomorphic in terms of its orienta-

ships, including the premaxilla, nasal with horncore, supra-

tion (laterally directed with a dorsally deflected tip), we refrain

orbital horncore, and parietosquamosal frill. The taxon pos-

from erecting a new taxon to accommodate this specimen on

sesses at least two autapomorphies (associated with the nasal

two grounds. First, recent finds elsewhere in the WIB indicate

and premaxillae), as well as a series of synapomorphies that

that elongate supraorbital horns, long thought to be a shared,

enable a preliminary phylogenetic assessment. Coahuilacera-

derived characteristic of Chasmosaurinae, are in fact primitive

tops can be confidently placed within Chasmosaurinae on the

for Ceratopsidae (Wolfe and Kirkland 1998) and retained in

basis of several synapomorphies, including large, rostrocau-

some centrosaurines (Kirkland and DeBlieux 2006, 2007, this

dally elongate external naris (indicated by the morphology of

volume; Ryan 2007; Sampson and Loewen this volume). Thus,

the premaxillae) and elongate squamosal, which differs from

at present, we cannot ascertain with confidence if CPC 278

the more rectangular, stepped squamosal of centrosaurines.

pertains to Centrosaurinae or Chasmosaurinae. Second, and

The basic skull architecture of Coahuilaceratops includes a long

more importantly, until additional remains of this animal are

snout, extremely robust supraorbital horncores, and an elon-

recovered, its taxonomic identification and phylogenetic af-

gate frill with large parietal fenestrae.

finities cannot be resolved.

The caudal inclination of the narial strut on the premaxilla

Chasmosaurinae. If not for the unusual and apomorphic pre-

in Coahuilaceratops appears to link this Mexican species with a

maxilla and nasal, it might be argued that the materials de-

pair of taxa from the late Campanian Horseshoe Canyon For-

scribed above for Coahuilaceratops magnacuerna are also in-

mation of Alberta, Anchiceratops and Arrhinoceratops; how-

sufficient to establish a new genus and species. Additionally,

ever, neither of the latter genera possesses the twisted cross-

although the skull of Coahuilaceratops is not well known, the

sectional shape of the narial strut present in Coahuilacera-

preserved assemblage of specimens does include several key

tops. Coahuilaceratops also differs from Anchiceratops and Ar-

elements useful in assessing taxonomy and historical relation-

rhinoceratops in the relative size, shape, and cross section of

110 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

the midline parietal bar and the size of the parietal fenestra, in

trosaurines recovered from Montana and Alberta (for review

which it resembles Pentaceratops and Agujaceratops. Finally, it

see Sampson and Loewen this volume) has led some authors

is noteworthy that Coahuilaceratops dates to an interval dur-

to suggest that centrosaurines represent a northern clade that

ing the late Campanian (approximately 72.2 Ma) for which

was absent in the south. Recent reports of centrosaurines from

no other chasmosaurines (or centrosaurines) are presently

Utah (Kirkland and DeBlieux 2006, 2007, this volume; Samp-

known (Sampson and Loewen this volume). Anchiceratops has

son and Loewen this volume) greatly extend the range of cen-

been reported from the top of the Dinosaur Park Formation

trosaurs and suggest that the previous lack of centrosaurines

near Manyberries (Langston 1959, 1975). This material (NMC

from the southern portion of the WIB represents a gap in sam-

9590 and 10645: consisting of isolated horncores; and NMC

pling rather than a true absence. An undescribed ceratopsid

9813, 9814 and 9829: consisting of incomplete distal parietals

specimen recovered from northern portion of Baja Peninsula,

and a partial squamosal) is indistinguishable from Pentacera-

Mexico, has been treated as Ceratopsidae indet. (Weishampel

tops and an unnamed taxon from the Almond Formation of

et al. 2004). It is possible that this specimen represents a cen-

southwestern Wyoming (Farke 2006). The Almond Formation

trosaurine ceratopsid, but paleogeographic reconstructions

ceratopsian has a rostrally inclined narial strut, distinguishing

place northern Baja at a latitude near that of the Cretaceous

it from the Horseshoe Canyon Formation Anchiceratops speci-

position of the Cerro del Pueblo material (Page and Engebret-

mens. Based on all available data, it is unlikely that the Dino-

son 1984; Bartow 1991; Scotese 2001). Interestingly, both the

saur Park Formation chasmosaur material from Manyberries

Alaskan pachyrhinosaur (Fiorillo and Gangloff 2003) and the

represents the same taxon as Anchiceratops from the Horse-

Cerro del Pueblo centrosaurine occur during the latest Cam-

shoe Canyon Formation; that latter materials occur in sedi-

panian, suggesting that immediately prior to the Maastrich-

ments nearly 3 million years younger, and most ceratopsids

tian, centrosaurs occupied a broad geographic range spanning

possess species ranges of much shorter duration (see discus-

from about 25\–70\ north latitude. Therefore, although the

sion in Sampson and Loewen this volume). In sum, Coahuila-

phylogenetic affinities of the Cerro del Pueblo centrosaurine

ceratops differs from all other known chasmosaurines, and we

are presently unknown, this specimen firmly extends the

posit that the present specimens (CPC 276, 277) are sufficient

range of centrosaurs from Alaska to Mexico during the Late

to erect a new taxon, C. magnacuerna.

Campanian.

Coahuilaceratops differs from all known described and un-

Chasmosaurinae. The grouping of Coahuilaceratops with

described southern chasmosaurines that we have examined

northern chasmosaurines, rather than the southern chasmo-

firsthand. As indicated in the differential diagnosis, the caudal

saurines, has significant biogeographic implications. Two in-

orientation of the narial strut separates Coahuilaceratops from

terpretations currently appear plausible. First, the Mexican

Agujaceratops, Pentaceratops, Kaiparowits new taxon B (Samp-

taxon may be derived from a northern chasmosaurine that

son and Loewen 2007, this volume), Triceratops, Diceratops

migrated south. Alternatively, the Maastrichtian northern

and Torosaurus (Lucas et al. 1998; Farke 2006). The cross-

chasmosaurine taxa are derived from an ancestral population

sectional shape of the nasal of Coahuilaceratops differs from

in southern North America. The occurrence of Coahuilacera-

that of the El Picacho chasmosaurine (Lehman 1996), too. The

tops magnacuerna in northern Mexico is one of several recently

shape of the squamosal and parietal fenestra separate Coa-

announced chasmosaurines from the southern portion of the

huilaceratops from Kaiparowits new taxon C (Sampson and

WIB (Sampson and Loewen this volume). As noted above, the

Loewen 2007, this volume), a new taxon from the Ojo Alamo

great bulk of ceratopsid taxa are known only from the north-

Formation (Sullivan and Lucas 2007, this volume) and from

ern region, in particular Alberta and Montana. During the lat-

Torosaurus. The fact that Coahuilaceratops differs from every

ter part of the twentieth century, dinosaur workers (Lehman

single well known and enigmatic specimen recovered from

1997, 2001) hypothesized that the pattern of high diversity in

the southern part of Laramidia and from every known north-

the north and low diversity in the south might represent the

ern taxon provides strong support for the erection of a new

true pattern of radiation within ceratopsids. However, recent

genus and species for the Cerro del Pueblo chasmosaur.

discoveries indicate that the diversity of southern forms may be greatly undersampled. In addition to the Maastrichtian

BIOGEOGRAPHIC IMPLICATIONS

Torosaurus latus (possibly T. utahensis) and two Campanian forms—Pentaceratops sternbergi and Agujaceratops mariscalensis

Centrosaurinae. CPC 279, an isolated squamosal recovered

—southern chasmosaurines now include two undescribed

from the Cerro del Pueblo, can be confidently placed within

taxa from the late Campanian Kaiparowits Formation of Utah

Centrosaurinae based on the stepped squamosal-parietal con-

(Sampson and Loewen this volume), and another from the

tact. The geographic range of centrosaurs established by re-

Maastrichtian Ojo Alamo Formation of New Mexico (Sulli-

ported pachyrhinosaur material from the north slope of Alaska

van and Lucas 2007, this volume). Thus, if one includes C.

(Fiorillo and Gangloff 2003) and the vast diversity of cen-

magnacuerna in this tally, the known diversity of southern

Horned Dinosaurs from the Upper Cretaceous Cerro del Pueblo Formation 111

Table 7.1. Character Codings for the Taxa Presented in Fig. 7.8 Using the Matrix of Wu et al. (2007)

Protoceratops Zuniceratops Achelousaurus Agujaceratops Anchiceratops Arrhinoceratops Avaceratops Centrosaurus Chasmosaurus Coahuilaceratops Diceratops Einiosaurus Pachyrhinosaurus Pentaceratops Styracosaurus Torosaurus Triceratops Eotriceratops

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

0 ? 0 1 1 1 ? 0 1 1 1 0 0 1 0 1 1 1

0 ? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0 ? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

? ? 0 1 1 1 0 0 1 1 1 0 0 1 0 1 1 1

0 ? 0 1 1 1 0 0 1 1 1 0 0 1 0 1 1 1

? ? ? 0 1 1 ? ? 0 1 0 ? ? 0 ? 0 0 0

? ? ? 1 0 0 ? ? 1 ? 0 ? ? 1 ? 0 0 0

? ? 0 ? 1 1 0 0 1 1 1 0 0 1 0 1 1 1

0 ? 0 0 0 0 0 0 0 0 1 0 0 0 0 1 1 1

? ? ? 0 0 ? ? ? 0 0 1 ? ? 0 ? 1 1 0

0 ? 0 1 1 1 0 0 1 ? 1 0 0 1 0 ? 1 1

? ? ? 0 0 0 ? ? 0 ? 1 ? ? 0 ? ? 1 0

0 ? 1 0 0 0 1 1 0 0 0 1 1 0 1 0 0 0

0 ? 1 0 1 1 1 1 0 ? 1 1 1 0 1 1 1 ?

0 ? 0 1 0 0 ? 0 1 ? 0 0 0 1 0 0 0 ?

0 ? 1 1 1 1 ? 1 1 ? 1 1 1 1 1 1 1 1

0 ? 2 1 1 1 ? 1 1 1 1 1 2 1 1 1 1 ?

0 ? 0 0 1 1 ? 0 0 1 1 0 0 0 0 1 1 ?

0 ? ? 0 0 0 ? 1 0 0 0 1 ? 0 1 0 0 0

0 ? 1 1 1 1 ? 1 1 ? 1 1 1 1 1 1 1 ?

0 1 1 1 1 1 ? 1 1 1 1 1 1 1 1 1 1 1

? 0 1 0 0 0 ? 0 0 0 0 0 1 0 0 0 0 0

? 0 0 0 1 1 ? 0 0 ? 1 0 0 0 0 1 1 1

? 0 1 0 0 0 ? 1 ? 0 0 1 1 0 1 0 0 0

? 0 ? 0 1 1 ? 0 0 ? 1 ? ? 1 ? 1 1 1

0 ? ? ? ? ? ? 1 0 ? ? ? ? ? 1 0 0 ?

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 Protoceratops Zuniceratops Achelousaurus Agujaceratops Anchiceratops Arrhinoceratops Avaceratops Centrosaurus Chasmosaurus Coahuilaceratops Diceratops Einiosaurus Pachyrhinosaurus Pentaceratops Styracosaurus Torosaurus Triceratops Eotriceratops

0 ? 1 1 0 0 ? 1 1 ? 0 1 1 ? 1 0 0 ?

0 ? 1 1 1 1 1 1 1 ? 1 1 1 1 1 1 1 1

0 ? 1 1 2 ? ? 1 1 2 2 1 1 ? 1 2 2 2

0 ? 1 1 1 1 ? 1 1 ? 1 1 1 1 1 1 1 1

? ? 0 1 2 ? ? 0 1 ? 2 0 0 1 0 2 2 ?

0 ? 1 0 0 0 ? 1 0 ? 0 1 1 0 1 0 1 ?

0 ? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0 ? 1 0 0 0 0 1 0 0 0 1 1 0 1 0 0 ?

0 ? 1 0 0 0 1 1 0 ? 0 1 1 0 1 0 0 ?

0 ? 1 0 0 0 0 1 0 ? 0 1 1 0 1 0 0 ?

0 ? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0 ? 1 1 1 1 ? 1 1 ? 1 1 1 1 1 1 1 ?

0 ? 0 1 0 0 ? 0 1 ? 0 0 0 1 0 0 0 ?

0 ? 0 0 0 0 ? 0 0 ? 1 0 0 0 0 1 0 ?

0 ? 1 0 1 1 ? 1 0 0 1 1 1 0 1 1 1 ?

0 ? 1 1 1 ? ? 1 1 1 1 1 1 1 1 1 1 1

? ? 1 0 0 ? ? 1 0 ? 0 1 1 0 1 0 1 1

? ? 1 0 0 ? ? 1 0 ? ? 1 1 0 1 0 0 0

? ? 0 1 1 ? ? 0 1 1 1 0 0 1 0 1 1 2

? ? 1 ? 0 0 ? 1 0 ? 0 1 1 0 1 0 0 ?

? ? 0 0 0 0 ? 1 0 ? 0 0 0 0 1 0 0 ?

? ? 0 ? 0 0 ? 1 1 ? 0 0 0 1 1 0 0 ?

? ? 0 0 0 0 ? 1 0 ? 0 0 1 0 1 0 0 ?

? ? 1 ? 0 0 ? 1 0 ? 0 1 1 0 1 0 0 ?

? ? 1 ? 0 0 ? 0 0 ? 0 1 1 0 1 0 0 ?

0 ? 1 1 1 1 1 1 1 ? 1 1 1 1 1 1 1 1

53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 Protoceratops Zuniceratops Achelousaurus Agujaceratops Anchiceratops Arrhinoceratops Avaceratops Centrosaurus Chasmosaurus Coahuilaceratops Diceratops Einiosaurus Pachyrhinosaurus Pentaceratops Styracosaurus Torosaurus Triceratops Eotriceratops

0 ? 1 ? 1 ? ? 1 1 ? ? 1 1 1 1 1 1 ?

0 ? 1 1 1 1 1 1 1 ? 1 1 1 1 1 1 1 ?

0 ? 1 1 1 1 1 1 1 ? 1 1 1 1 1 1 1 ?

0 ? 1 ? 1 1 1 1 1 ? 1 1 1 1 1 1 1 ?

0 ? 1 0 0 ? ? 1 0 ? ? 1 1 0 1 0 0 ?

0 0 1 ? 1 ? 1 1 1 1 ? 1 1 1 1 1 1 ?

0 1 1 1 1 ? 1 1 1 1 ? 1 1 1 1 1 1 ?

0 0 1 1 1 1 1 1 1 ? 1 1 1 1 1 1 1 1

0 0 1 ? 1 ? 1 1 1 ? 1 1 1 1 1 1 1 1

0 0 1 1 1 ? 1 1 1 ? 1 1 1 1 1 1 1 1

0 ? 1 ? 1 ? 1 1 1 ? ? 1 1 1 1 ? 1 1

0 ? 1 1 ? ? ? 1 1 ? ? 1 1 1 1 ? 1 ?

0 ? 0 0 ? ? ? 0 1 ? ? 0 0 1 0 ? 1 ?

0 ? 1 ? 1 ? ? 1 1 ? ? 1 1 ? 1 ? 1 ?

0 ? 1 1 1 ? 1 1 1 ? ? 1 1 1 1 ? 1 ?

0 ? 1 1 1 ? ? 1 1 ? ? 1 1 1 1 ? 1 ?

0 ? 1 1 1 ? ? 1 1 ? ? 1 1 1 1 ? 1 ?

0 ? 0 1 1 ? ? 0 1 ? ? 0 0 1 0 ? 1 ?

0 ? 1 2 2 ? ? 1 2 ? ? 1 1 2 1 ? 2 ?

0 ? 1 1 1 ? 1 1 1 ? ? 1 1 1 1 ? 1 ?

0 ? 1 ? 1 ? 1 1 1 ? ? 1 1 1 1 ? 1 ?

N ? 0 0 0 0 ? 0 0 0 1 0 0 0 0 1 1 1

N ? ? 0 0 0 ? 0 0 0 0 0 0 0 0 0 1 1

? ? N 1 ? ? N N 0 ? 1 N N 1 N ? 1 1

112 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

chasmosaurines has more than doubled since 2000. These dis-

Rosario Gomez-Nuñez (Coordinadora de Paleontología de la

coveries, combined with the growing realization that cera-

Secretaría de Educación y Cultura) and her family. Funding

topsid species had small geographic ranges and possibly rela-

for this research was provided by the National Geographic

tively brief ([1 million year) species durations (Sampson and

Society, the University of Utah, and the Government of the

Loewen this volume), suggest that many more ceratopsid taxa

State of Coahuila (through the Secretaría de Educacion y

await discovery.

Cultura). References Cited

Conclusions Remains of ceratopsid dinosaurs recently recovered from the Cerro del Pueblo Formation of Coahuila, Mexico, provide an important glimpse of the southernmost portion of the ceratopsid radiation. An isolated squamosal (CPC 279) represents the first demonstrable occurrence of Centrosaurinae in Mexico, documenting the wide geographic range of this clade. Chasmosaurine material (CPC 276 and CPC 277) from Cerro del Pueblo represent a new genus and species, Coahuilaceratops magnacuerna, which possesses autapomorphies in the shape of the narial strut and the cross-sectional shape of the nasal. Additionally Coahuilaceratops possesses a unique suite of synapomorphies (i.e., caudally directed narial strut; small rostrally positioned nasal horncore; extremely robust supraorbital horncores; large parietal fenestrae; and a narrow midline parietal bar with oval cross section) that is unique among ceratopsids. Coahuilaceratops is tentatively allied (based on a single synapomorphy) with two other late Campanian chasmosaurines, Anchiceratops and Arrhinoceratops from the Horseshoe Canyon Formation of Alberta. Acknowledgments

For discussions about ceratopsids generally and for providing thoughtful reviews of this manuscript, we thank Jim Kirkland and Michael Ryan. For assistance with fieldwork, we sincerely thank the 2002 and 2005 crews from the Utah Museum of Natural History, Royal Tyrell Museum of Palaeontology, Coordinación de Paleontología de la SEC, and Museo del Desierto. For field assistance and specimen preparation, we are grateful to José López Espinosa (‘‘Pato’’), José Ignacio Vallejo González (‘‘Nacho’’), Joe Gentry, and Jerry Golden; we also thank Pato’s family at Presa San Antonio for their hospitality and friendship. We thank Francisco Martínez Perez, director of INAHCoahuila for support of this project, Daniel Posada Martínez and Fernando Cabral Valdés (Museo del Desierto), for providing specimens, and Jesus Garza-Arocha for his support and hospitality. For access to collections and specimens, we thank Jim Gardner (Royal Tyrrell Museum of Palaeontology), Jack Horner (Museum of the Rockies), Carl Mehling (American Museum of Natural History), Kevin Seymour (Royal Ontario Museum), Kieran Shepherd (Canadian Museum of Nature), and Tom Williamson (New Mexico Museum of Natural History). Finally, for her invaluable support and logistical assistance through all phases of this project, we sincerely thank

Bartow, J. A. 1991. The Cenozoic evolution of the San Joaquin Valley, California. U.S. Geological Survey Professional Paper P 1501: 40. Brinkman, D. B., D. A. Eberth, S. D. Sampson, M. C. AguillónMartínez, C. R. Delgado de Jesús, and R. A. Rodríguez-de la Rosa. 2002. Paleontology and stratigraphy of the dinosaurbearing Cerro del Pueblo Formation, Southeastern Coahuila, Mexico. Journal of Vertebrate Paleontology 22(3, Suppl.): 38A– 39A. Brown, B. 1914. Anchiceratops, a new genus of horned dinosaurs from the Edmonton Cretaceous of Alberta. With a discussion of the origin of the ceratopsian crest and the brain casts of Anchiceratops and Trachodon. Bulletin of the American Museum of Natural History 33: 539–548. Dodson, P. 1986. Avaceratops lammersi: A new ceratopsid from the Judith River Formation of Montana. Proceedings of the Academy of Natural Sciences 138: 305–317. Dodson, P., and P. J. Currie. 1990. Neoceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 593–618. Berkeley: University of California Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A., C. R. Delgado-de Jesus, J. F. Lerbekmo, D. B. Brinkman, R. A. Rodriguez-de la Rosa, and S. D. Sampson. 2004. Cerro del Pueblo Fm (Difunta Group, Upper Cretaceous), Parras Basin, southern Coahuila, Mexico: Reference sections, age, and correlation. Revista Mexicana de Ciencias Geologicas 21(3): 335–352. Eberth, D. A., S. D. Sampson, R. A. Rodríguez-de la Rosa, M. C. Aguillón-Martínez, and D. B. Brinkman. 2003. Las Aguilas: An unusually rich Campanian-age vertebrate locale in southern Coahuila, Mexico. Journal of Vertebrate Paleontology 23(3, Suppl.): 47A. Farke, A. A. 2006. Cranial osteology and phylogenetic relationships of the chasmosaurine ceratopsid Torosaurus latus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 235–257. Bloomington: Indiana University Press. Fiorillo, A. R., and R. A. Gangloff. 2003. Preliminary notes on the taphonomic and paleoecologic setting of a Pachyrhinosaurus bonebed in northern Alaska. Journal of Vertebrate Paleontology 23(3, Suppl.): 50A. Forster, C. A., and Sampson, S. D. 2002. Phylogeny of the horned dinosaurs (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 22(3, Suppl.): 54A.

Horned Dinosaurs from the Upper Cretaceous Cerro del Pueblo Formation 113

Forster, C. A., P. C. Sereno, T. W. Evans, and T. A. Rowe. 1993. A complete skull of Chasmosaurus mariscalensis (Dinosauria: Ceratopsidae) from the Aguja Formation (late Campanian) of West Texas. Journal of Vertebrate Paleontology 13: 161–170. Gates, T. A., S. D. Sampson, C. R. Delgado-de Jesus, L. E. Zanno, D. A. Eberth, R. Hernández Rivera, M. C. Aguillón-Martínez, and J. I. Kirkland. 2007. Velafrons coahuilensis, a new lambeosaurine hadrosaur (Dinosauria: Ornithopoda) from the Late Campanian Cerro del Pueblo Formation, Coahuila, Mexico. Journal of Vertebrate Paleontology 27: 917–930. Goodwin, M. B., and J. R. Horner. 2008. Ontogeny of cranial epiossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Hatcher J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49. Heckert, A. B., S. G. Lucas, and S. E. Krzyzanowski. 2003. Vertebrate fauna of the Late Campanian ( Judithian) Fort Crittenden Formation, and the age of Cretaceous vertebrate faunas of southeastern Arizona (U.S.A.). Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 227: 343–364. Hernández, R., M. C. Aguillón-Martínez, C. R. Delgado, and N. R. Gómez. 1995. The Mexican Dinosaur National Monument. Journal of Vertebrate Paleontology 15(3, Suppl.): 34A. Holmes, R. B., C. A. Forster, M. J. Ryan, and K. M. Shepherd. 2001. A new species of Chasmosaurus (Dinosauria: Ceratopsia) from the Dinosaur Park Formation of Southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Janensch, V. W. 1926. Dinosaurier-Reste aus Mexiko. Centralblatt für Mineralogie, Geologie und Paläontologie 1926B: 192–197. Kirkland, J. I., and M. C. Aguillón-Martínez. 2002. Schizorhiza: A unique sawfish paradigm from the Difunta Group, Coahuila, Mexico. Revista Mexicana de Ciencias Geológicas 19: 16–24. Kirkland, J. I., and D. D. DeBlieux. 2006. A new genus of ornate long-horned centrosaurine ceratopsian from the middle Campanian (Cretaceous) Wahweap Formation, Grand StaircaseEscalante National Monument, Southern Utah. Journal of Vertebrate Paleontology 26(3, Suppl.): 85A. ———. 2007. New centrosaurine ceratopsians from the Wahweap Formation, Grand Staircase-Escalante National Monument, southern Utah. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 90–95. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2010. New basal centrosaurine ceratopsian skulls from the Wahweap Formation (Middle Campanian), Grand Staircase– Escalante National Monument, southern Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 117–140. Bloomington: Indiana University Press. Kirkland, J. I., R. Hernández Rivera, M. C. Aguillón-Martínez, C. R. Delgado-de Jesús, R. Gómez-Núñez, and I. Vallejo. 2000. The Late Cretaceous Difunta Group of the Parras Basin, Coahuila, Mexico, and its vertebrate fauna. In Society of Vertebrate Paleontology Annual Meeting, 2000, Field Trip Guide Book, pp. 133–172. Mexico: Universidad Autónoma del Estado de Hidalgo, Avances en Investigación 3. Lambe, L. 1914. On Gryposaurus notabilis, a new genus and spe-

cies of trachodont dinosaur from the Belly River Formation of Alberta, with a description of the skull of Chasmosaurus belli. Ottawa Naturalist 27: 145–155. ———. 1915. On Eoceratops canadensis, gen. nov., with remarks on other genera of Cretaceous horned dinosaurs. Canada Department of Mines Museum Bulletin 12, Geological Series 24:1–49. Langston, W., Jr. 1959. Anchiceratops from the Oldman Formation of Alberta. National Museum of Canada Natural History Papers 3:1–11. ———. 1975. The ceratopsian dinosaurs and associated lower vertebrates from the St. Mary River Formation (Maestrichtian) at Scabby Butte, southern Alberta. Canadian Journal of Earth Sciences 12(9): 1576–1608. Lehman, T. M. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. ———. 1993. New data on the ceratopsian dinosaur Pentaceratops sternbergii Osborn from New Mexico. Journal of Paleontology 67: 279–288. ———. 1996. A horned dinosaur from the El Picacho Formation of west Texas, and review of ceratopsian dinosaurs from the American Southwest. Journal of Paleontology 17: 494–508. ———. 1997. Late Campanian dinosaur biogeography in the western interior of North America. In D. L. Wolberg, E. Stump, and G. D. Rosenburg, eds., Dinofest International: Proceedings of a Symposium held at Arizona State University, pp. 223–240. Philadelphia: Academy of Natural Sciences. ———. 1998. A gigantic skull and skeleton of the horned dinosaur Pentaceratops sternbergi from New Mexico. Journal of Paleontology 72: 894–906. ———. 2001. Late Cretaceous dinosaur provinciality. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 310–328. Bloomington: Indiana University Press. Lucas, S. G., G. H. Mack, and J. W. Estep. 1998. The Ceratopsian Dinosaur Torosaurus from the Upper Cretaceous McRae Formation, Sierra County, New Mexico. New Mexico Geological Society Guidebook, 49th Field Conference. Lucas, S. G., R. M., Sullivan, and A. P. Hunt. 2006. Re-evaluation of Pentaceratops and Chasmosaurus (Ornithischia: Ceratopsidae) in the Upper Cretaceous of the Western Interior. In S. G. Lucas and R. M. Sullivan, eds., Late Cretaceous Vertebrates from the Western Interior, pp. 367–370. New Mexico Museum of Natural History and Science Bulletin 35. Lund, E. K., M. A. Loewen, S. D. Sampson, M A. Getty, A. Aguillon Martinez, R. A. Rodriguez de la Rosa, and D. A. Eberth. 2007. Ceratopsian remains from the Late Cretaceous Cerro del Pueblo Formation, Coahuila, Mexico. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 108–113. Drumheller: Royal Tyrrell Museum of Palaeontology. Marsh, O. C. 1888. A new family of horned Dinosauria from the Cretaceous. American Journal of Science 36: 477–478. ———. 1890. Additional characters of the Ceratopsidae with notice of new Cretaceous dinosaurs. American Journal of Science 39: 418–426.

114 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

———. 1891. Notice of new vertebrate fossils. American Journal of Science 42: 265–269. McBride, E. F., A. E. Weidie, J. A. Wolleben, and R. C. Laudon. 1974. Stratigraphy and structure of the Parras and La Popa basins, northeastern Mexico. Geological Society of America Bulletin 84: 1603–1622. Murray, G. E., D. R. Boyd, J. A. Wolleben, and J. A. Wilson. 1960. Late Cretaceous fossil locality, eastern Parras Basin, Coahuila, Mexico. Journal of Paleontology 34: 368–373. Murray, G. E., A. E. Weidie Jr., D. R. Boyd, R. H. Forde, and P. D. Lewis Jr. 1962. Formational divisions of Difunta Group, Parras Basin, Coahuila and Nuevo León, Mexico. Bulletin of the American Association of Petroleum Geologists 46: 374–383. Osborn, H. F. 1923. A new genus and species of ceratopsia from New Mexico, Pentaceratops sternbergii. American Museum Novitates 93: 1–3. Page, B. M., and D. C. Engebretson. 1984. Correlation between the geologic record and computed plate motions for central California. Tectonics 3: 133–155. Parks, W. A. 1925. Arrhinoceratops brachyops a new genus and species of Ceratopsia from the Edmonton Formation of Alberta. University of Toronto Studies (Geology Series) 19:1–15. Robinson-Roberts, L. N., and M. A. Kirschbaum. 1995. Paleogeography of the Late Cretaceous of the Western Interior of Middle North America. U.S. Geological Survey Professional Paper 1561:1–115. Rodríguez-de la Rosa, R. A. 1996. Vertebrate remains from a Late Cretaceous Locality (Campanian, Cerro Del Pueblo Formation), Coahuila, Mexico. Journal of Vertebrate Paleontology 16(3, Suppl.): 45A. Rodríguez-de la Rosa, R. A., and R. S. Cevallos-Ferriz. 1994. Upper Cretaceous zingiberalean fruits with in situ seeds from southeastern Coahuila, Mexico. International Journal of Plant Science 155: 786–805. ———. 1998. Vertebrates of the El Pelillal locality (Campanian, Cerro del Pueblo Formation), Southeastern Coahuila, Mexico. Journal of Vertebrate Paleontology 18(4): 751–764. Rodríguez-de la Rosa, R. A., R. S. Cevallos-Ferriz, and A. SilvaPineda. 1998. Paleobiological implications of Campanian coprolites. Palaeogeography, Palaeoclimatology, and Palaeoecology 142: 231–254. Rodríguez-de la Rosa, R. A., D. A. Eberth, D. B. Brinkman, S. D. Sampson, and J. Lopez Espinoza. 2003. Dinosaur tracks from the late Campanian Las Aguilas locality, southeastern Coahuila, Mexico. Journal of Vertebrate Paleontology 23(3, Suppl.): 90A. Rodríguez-de la Rosa, R. A., J. López-Espinoza, J. I. VallejoGonzalez, D. A. Eberth, and J. A. Smith. 2002. Huellas de vertebrados Cretácicos (Campaniano tardío, Formación Cerro del Pueblo) del sureste de Coahuila, México (resumen). In VIII Congreso Nacional de Paleontología, Libro de Resúmenes, pp. 75–76. Guadalajara: Sociedad Mexicana de Paleontología. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., and D. C. Evans. 2005. Ornithischian dinosaurs. In

P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 312–348. Bloomington: Indiana University Press. Ryan, M. J., R. Holmes, and A. P. Russell. A revision of the late Campanian centrosaurine ceratopsid genus Styracosaurus from the Western Interior of North America. Journal of Vertebrate Paleontology 27: 944–962. Sampson, S. D. 1995. Two new horned dinosaurs from the Upper Cretaceous Two Medicine Formation of Montana, USA, with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15: 743–760. Sampson, S. D., and M. A. Loewen. 2007. New information on the diversity, stratigraphic distribution, biogeography, and evolution of ceratopsid dinosaurs. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 125– 133. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2010. Unraveling a radiation: A review of the diversity, stratigraphic distribution, biogeography, and evolution of horned dinosaurs (Ornithischia: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 405–427. Bloomington: Indiana University Press. Sampson, S. D., M. A. Loewen, E. M. Roberts, J. A. Smith, L. E. Zanno, and T. A. Gates. 2004. Provincialism in Late Cretaceous terrestrial faunas: New evidence from the Campanian Kaiparowits Formation of Utah. Journal of Vertebrate Paleontology 24(3, Suppl.): 108A. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Scotese, C. R. 2001. Atlas of Earth History. Vol. I: Paleogeography. Arlington, Texas: PALEOMAP Project. Seeley, H. G. 1888. The classification of the Dinosauria. Report of the British Association for the Advancement of Science 1887: 698– 699. Soegaard, K., K. Giles, F. Vega, and T. Lawton 1997. Structure, stratigraphy, and paleontology of Late Cretaceous-early Tertiary Parras—La Popa foreland basin near Monterrey, Mexico. Dallas: American Association of Petroleum Geologists, Guidebook, Field Trip 10. Sternberg, C. M. 1940. Ceratopsidae from Alberta. Journal of Paleontology 14: 468–480. Sullivan, R. M., and S. G. Lucas. 2007. A new chasmosaurine (Ceratopsidae: Dinosauria) from the Upper Cretaceous Ojo Alamo Formation (Naashoibito Member), San Juan Basin, New Mexico. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, p. 140. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2010. A new chasmosaurine (Ceratopsidae, Dinosauria) from the Upper Cretaceous Ojo Alamo Formation (Naashoibito Member), San Juan Basin, New Mexico. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 169–180. Bloomington: Indiana University Press.

Horned Dinosaurs from the Upper Cretaceous Cerro del Pueblo Formation 115

Vega, F. J., and R. M. Feldmann. 1991. Fossil crabs (Crustacea, Decapoda) from the Maastrichtian Difunta Group, northeastern Mexico. Annals of the Carnegie Museum 60: 163–177. Weide, A. E., and G. E. Murray. 1967. Geology of Parras Basin and adjacent areas of northeastern Mexico. Bulletin of the American Association of Petroleum Geologists 51: 678–695. Weishampel, D. M., P. M. Barrett, R. A. Coria, J. L. Loeuff, X. Xing, Z. Xijin, A. Sahni, E. M. P. Gomani, and C. R. Noto. 2004. Dinosaur distribution. In D. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 517–606. Berkeley: University of California Press. Wolfe, D. G., and J. I. Kirkland. 1998. Zuniceratops christopheri

n. gen. & n. sp., a ceratopsian dinosaur from the Moreno Hill Formation (Cretaceous, Turonian) of west-central New Mexico. Lower and Middle Cretaceous Terrestrial Ecosystems. New Mexico Museum of Natural History and Science Bulletin 24: 307–317. Wolleben, J. A. 1977. Paleontology of the Difunta Group (Upper Cretaceous–Tertiary) in northern Mexico. Journal of Paleontology 51: 373–398. Wu, X., D. B. Brinkman, D. A. Eberth, and D. R. Braman. 2007. A new ceratopsid dinosaur (Ornithischia) from the uppermost Horseshoe Canyon Formation (Upper Maastrichtian), Alberta, Canada. Canadian Journal of Earth Sciences 44: 1243–1265.

116 loewen, sampson, lund, farke, aguillón-martínez, de leon, rodríguez-de la rosa, getty, and eberth

8 New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation (Middle Campanian), Grand Staircase– Escalante National Monument, Southern Utah JAMES I. KIRKLAND AND DONALD D. DEBLIEUX

a new basal centrosaurine ceratopsid, Diabloceratops

horn are primitive character states for ceratopsids, as

eatoni, is described from the Wahweap Formation

indicated by the ceratopsid sister taxon Zuniceratops.

(lower to middle Campanian, Upper Cretaceous) of

The basal position of Diabloceratops among centro-

Grand Staircase–Escalante National Monument, south-

saurines is supported by the absence of a contact

ern Utah. The isolated, nearly complete skull is one of

between the ascending process of the premaxilla and

the geologically oldest and is the first diagnosable cen-

the lacrimal as in other centrosaurines; rather the

trosaurine recovered south of Montana. It shares with

ascending process of the premaxilla terminates

more derived centrosaurines a stepped squamosal and a

rostral to the lacrimal as in Zuniceratops and all

nasal-premaxillary process along the caudal border of

chasmosaurines.

the naris. The species is diagnosed by numerous autapo-

A second, larger partial centrosaurine skull recovered

morphies relative to other centrosaurines: (1) the pre-

from the Wahweap Formation is not represented by

orbital skull is deeper and shorter than other known

enough critical elements to be confidently diagnosed.

ceratopsids; (2) rostral to a low, subconical nasal horn

We tentatively place it in the genus Diabloceratops based

is a smaller ‘‘epinasal’’; (3) a large accessory antorbital

on the presence of long postorbital horns, and a dorso-

fenestra is present; (4) fused frontals form a steep vault

ventrally oriented attachment scar on the jugal, which

between large postorbital horns at level of palpebrals;

indicates the possible presence of a blade-like epijugal.

(5) elongate jugals expose the caudal end of maxillae in

It is distinguished from Diabloceratops eatoni in bearing

lateral view; (6) a large, triangular, vertically oriented,

epoccipitals closely appressed to either side of the

blade-like epijugal extends laterally from the jugal

squamosal-parietal suture.

bone; (7) the erect frill is widest at the laterally directed

The presence of a well-developed accessory antorbital

squamosals, tapering to half its width at the base of a

fenestra in Diabloceratops is shared with its close out-

pair of elongate caudal parietal spines separated by a

group Zuniceratops. Among more basal neoceratopsians,

medial notch; (8) epoccipitals on the lateral margin of

only Magnirostris and Bagaceratops share this character.

parietal decrease in size caudally to the base of parietal

The presence of distinct, albeit tiny, postorbital horns

spines; and (9) the parietal fenestrae are rostrocaudally

indicates that Magnirostris may be the Asian sister taxon

elongate. The long postorbital horns and small narial

to North America’s larger ceratopsids.

117

Introduction

Butte. The Nipple Butte skull is the first ceratopsid specimen identified from the Wahweap Formation. Although the skull

The study of microvertebrate fossils collected via wet screen-

had been exposed on the surface for many years and had bro-

washing by researchers from the Museum of Northern Ari-

ken into three major sections, it was obviously an important

zona, University of Oklahoma, and Weber State University

specimen that required salvage. The skull was collected in Au-

has documented that the Wahweap Formation on the Kai-

gust 2000 by the UGS, aided by the University of Utah and

parowits Plateau of southern Utah preserves the most diverse

Utah Friends of Paleontology (Fig. 8.2). The skull was found

lower-middle Campanian terrestrial fauna in North America

one-third of a mile off an established dirt road, and, because

(Fig. 8.1). These studies have documented 4 freshwater shark

wheeled vehicles are not permitted off-road, the three large

species, 2 freshwater ray species, 7 bony fish species, 2 am-

sandstone blocks were dragged out by an eight-person team

phibian species, 6 turtle genera, 2 lizard taxa, 3 crocodil-

using the roof of a Ford Mustang as a sled (Kirkland 2001).

ian taxa, 8 dinosaur taxa, and 23 mammal species (Eaton et

Following hundreds of hours of preparation, the Nipple Butte

al. 1999). With the establishment of the Grand Staircase–

skull was identified as a centrosaurine ceratopsid dinosaur

Escalante National Monument (GSENM) in 1996, the Kai-

(Stokstad 2001; Kirkland et al. 2002). However, lacking the

parowits Plateau was incorporated into the largest national

nasal area and the rear of the frill, we concluded that given the

monument in the lower 48 states. Subsequently, the U.S. Bu-

recent discovery of other centrosaurines with large postorbital

reau of Land Management (BLM) funded the Utah Geological

horns (Ryan 2003, 2007a), the specimen is not fully diagnostic

Survey (UGS) to refine the locality data for known paleontol-

as to genus.

ogical sites and begin efforts to document additional paleontological resources (Foster et al. 2001).

In 2002, Don DeBlieux discovered a skull weathering out of a sandstone ledge near the middle of the middle mudstone

In 2001, we began the Wahweap Project, aimed at inven-

member of the Wahweap Formation (Eaton 1991) near Last

torying fossil localities in the lower sandstone and middle

Chance Creek on the north side of Reynolds Point in the east-

mudstone members of the Wahweap Formation across the

ern Wahweap outcrop belt (Fig. 8.3). The collection of bone

southern end of the Kaiparowits Basin (Plateau) to serve as a

on the surface and cleaning of the block revealed a nearly

management tool for the BLM. In addition to providing data

complete skull lying on its left side; part of the right side had

on the distribution of paleontological resources, this study

eroded away, with much of the skull still imbedded in the rock

has identified and recovered specimens that are adding to our

(Fig. 8.3). We spent eight days using a gas-powered cutoff saw

knowledge of terrestrial faunas during a time interval for

to separate the block containing the skull from the surround-

which they are poorly known. As of April 2008, there are still

ing ledge. In September 2005, arrangements were finalized by

only two dinosaurs that have been identified to species from

the BLM to airlift the block containing the skull out of the

rocks of this approximate age in North America (lower Two

backcountry (Fig. 8.3). The block was then transported by

Medicine Formation of Montana), a hadrosaurine hadrosau-

helicopter to a truck, and then driven to the UGS preparation

rid Gryposaurus latidens (Horner 1992) and a leptoceratopsid-

lab in Salt Lake City, Utah. More than 800 hours of prepara-

grade basal ceratopsian Cerasinops hodgskissi (Chinnery and

tion (by D. D.) were needed to expose the beautifully pre-

Horner 2007). As discussed below, these dinosaurs are now

served skull. The Last Chance skull is the first diagnosable,

thought to be a couple of million years older than the Wah-

centrosaurine ceratopsid dinosaur older than Late Campa-

weap Formation. A number of associated hadrosaurid skele-

nian, and the first diagnosable, centrosaurine ceratopsid dino-

tons have been identified in the field, although taxonomi-

saur recovered south of Montana.

cally critical cranial remains have yet to be identified in these

Here we describe and diagnose the Last Chance skull as a

as-yet preliminary excavations. The isolated skull roof of a

new basal centrosaurine ceratopsid dinosaur and compare it

juvenile pachycephalosaur has also been collected. Addition-

with the Nipple Butte skull, and other pertinent neoceratop-

ally, carnivorous dinosaur remains have been identified at a

sian skulls. Both skulls are curated into the vertebrate paleon-

number of sites, but nothing diagnostic has yet come to light

tology collections of the Utah Museum of Natural History at

(Kirkland and DeBlieux 2005). At GSENM, cranial remains of

the University of Utah in Salt Lake City, Utah. During the

long-horned centrosaurine ceratopsids are the most signifi-

course of this study we examined ceratopsid cranial material

cant dinosaur fossils to be identified so far from the Wahweap

housed in the institutions listed below.

Formation.

Institutional Abbreviations. AMNH: American Museum of

Joshua A. Smith, while working for the UGS, discovered a

Natural History, New York; CMN: Canadian Museum of Na-

skull of a ceratopsid dinosaur during a July 16, 1998, UGS

ture, Ottawa; DMNH: Denver Museum of Natural History and

paleontological survey of GSENM, in the lower Wahweap For-

Science, Denver; IVPP: Institute of Vertebrate Paleontology

mation at the top of the lower sandstone member on the

and Paleoanthropology, Beijing; NMMNHS: New Mexico Mu-

south end of the Kaiparowits Plateau, northwest of Nipple

seum of Natural History and Science, Albuquerque; TMM:

118 kirkland & deblieux

(A) Grand Staircase–Escalante National Monument with outcrop of Campanian Wahweap and Kaiparowits Formations. General location of Nipple Butte and Last Chance ceratopsian skulls indicated. (B) Wahweap fauna mostly from Eaton et al. (1999). N = 57 identified taxa.

FIGURE 8.1.

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 119

Collection of the Nipple Butte Skull. (A) Overview of site at top of the lower sandstone member of Wahweap Formation (Nipple Butte to southeast in background); (B) field sketch showing relationship of bone-bearing sandstone blocks; (C) dragging skull block toward road; (D) winching skull block up slope in middle mudstone member; (E) lifting skull block up for loading into truck.

FIGURE 8.2.

Texas Memorial Museum, Austin; TMP: Royal Tyrrell Museum

rh: right horn; rm: impression of right maxilla; s: squamosal;

of Palaeontology, Drumheller; UMNH: Utah Museum of Natu-

sad: sulcus connecting frontal fontanelle with supratemporal

ral History, Salt Lake City.

fenestra; so: supraoccipital; st: supratemporal fenestra; t: tri-

Anatomical Abbreviations. IX–XII: cranial nerve openings;

angular process on parietal.

II–XII: cranial nerve openings; aaf: accessory antorbital fenestra; af: antorbital fenestra; bp: basipterygoid process; bt:

Systematic Paleontology

basitubera; cc: path for carotid artery; cff: area collapsed into frontal fontanelle; e: ‘‘eustachian’’ canal of Hatcher et al.

Ornithischia Seeley 1888

(1907); ec: ectopterygoid; ej: epijugal(s); ej: jugal in Fig. 8.12

Ceratopsia Marsh 1890

(sutural surface for missing epijugal); en: epinasal; ff: frontal

Neoceratopsia Sereno 1986

fontanelle; fm: foramen magnum; fo: fenestra ovalis; it: in-

Ceratopsidae Marsh 1888

fratemporal fenestra; j: jugal; l: lacrimal; lam: caudal limit of

Centrosaurinae Lambe 1915

jaw adductor muscle on parietal; lb: laterosphenoid buttress;

Diabloceratops eatoni gen. et sp. nov.

m: left maxilla; m: right maxilla (Fig. 8.11); m: ascending process of right maxilla (Fig. 8.12); mqc: medial condyle of left

Holotype. The ‘‘Last Chance’’ skull, UMNH VP 16699, a skull

quadrate; n: nasal; nh: nasal horn; np: caudal nasal process; o:

preserving the entire left side of the skull and portions of the

occipital condyle; o: orbit in Fig. 8.12; oc: occipital condyle;

right side.

oh: orbital horn; p: parietal; pal: palpebral; path: pathology;

UMNH VP 16699 preserves the entire left side of the skull

pb: parietal bar; pf: parietal fenestra; pl: palatine; pm: pre-

including the midline (Figs. 8.4, 8.5). The skull was sagitally

maxilla; po: postorbital; pp: paroccipital process; ps: parietal

sectioned by erosion through the right side, resulting in the

spine; pt: pterygoid; q: quadrate; qj: quadratojugal; r: rostral;

loss of the right squamosal, half of the right postorbital horn,

120 kirkland & deblieux

Collection of Last Chance Skull. (A) North side of Reynolds Point with the gorge of Last Chance Creek in foreground. Star indicates location of skull. Arrowheads indicate member contacts of Wahweap Formation. ls: base of lower sandstone member; mm: base of middle mudstone member; us: base of upper sandstone member; cs: base of capping sandstone member. (B) Skull as first observed; (C) skull after surface clean-up; (D) rock-sawing the skull block; (E) airlifting the skull out of the backcountry to the road.

FIGURE 8.3.

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 121

accessory antorbital opening bounded by the premaxilla, maxilla, and nasal, such that the premaxilla is excluded from contact with lacrimal by maxilla; preorbital skull is deep and shorter than other known centrosaurines; small epinasal rostral to low narial horn; large postorbital horns extending rostrally to narial horn; elongate jugal exposes caudal end of maxilla in lateral view; vertically oriented, blade-like epijugals larger than in other described centrosaurines; fused frontals form steep vault between orbital horns rostral to palpebrals; erect frill is subequal in length to skull length, widest at laterally directed squamosals, tapering by 50% to a pair of elongate caudal parietal spines separated by a midline notch; parietal epoccipitals decrease in size caudally; parietal fenestrae are triangular and rostrocaudally elongate. Type Locality. 42Ka800V, Grand Staircase–Escalante National Monument, south side of Last Chance Canyon, north flank of Reynolds Point, west of Smoky Mountain Road, central Kaiparowits Plateau, Utah. Stratigraphic Occurrence. The skull was collected from the Wahweap Formation, near the middle of the middle mudstone member, 51.72 m above its base and 105.30 m above the contact of the Wahweap Formation on the underlying Drip Tank Member of Straight Cliffs Formation. It was found in a ledge comprising medium-grained, trough cross-bedded sandstone with abundant clay clasts and bone fragments, including turtle shell fragments.

Left lateral view of Diabloceratops eatoni holotype (UMNH VP 16699). (A) Diagramatic representation of skull; (B) skull. See also plate 8.

FIGURE 8.4.

jugal, and part of the right maxilla. A number of fragments from the right side were recovered. A rib had come to rest between the two horns lying against the left side of the face. A partial splenial was also found adjacent to the skull during excavation. The bones of the skull are very well preserved and almost completely undistorted. Sutures are well fused but can be distinguished in most cases, and the epoccipitals on the squamosal and parietal are well fused, indicating that this specimen is adult. Etymology. ‘‘Diablo,’’ Spanish for devil in reference to the pair of long sweeping spines on the back of the frill; + ‘‘ceratops,’’ horned-face, Latinized Greek. ‘‘eatoni,’’ a patronym in honor of Jeffery G. Eaton, a paleontologist at Weber State University in Ogden, Utah, in recognition for his extensive work on the Cretaceous vertebrate faunas of southern Utah, and his role in the establishment of Grand Staircase–Escalante National Monument. Diagnosis. A relatively small centrosaurine ceratopsid (skull estimated at 620 mm from rostrum to condyle) with a large

122 kirkland & deblieux

Caudal and right lateral views of Diabloceratops eatoni holotype (UMNH VP 16699). (A) Caudal view; (B) right lateral (medial) view.

FIGURE 8.5.

Description The skull is 1 m long from the beak to the back of the frill, where a pair of curved spines adds another 0.5 m to the total length. The basal length of skull is estimated at 620 mm (rostrum to condyle), and the occipital condyle is 54.4 mm in diameter. The skull shares with more derived centrosaurines a stepped squamosal and a nasal-premaxillary process along the caudal border of the naris (Fig. 8.4). The preorbital skull is deep (285 mm nasal to maxilla) and short (425 mm orbit to rostrum) as in known centrosaurines.

SNOUT Most of the rostral has been lost to erosion with much of the right side extracted as a cast of the external sandstone mold using epoxy putty. The ventral portion is preserved on the left side; the rugosity of the rostral and the suture with the premaxilla are typical of other ceratopsids (Dodson et al. 2004; Horner and Goodwin 2008). Both premaxillae are preserved. The tall premaxillae contribute to a deep muzzle in Diabloceratops. The sutured dorsal and

Detail of antorbital fossa of Diabloceratops eatoni (UMNH VP 16699) from cast specimen.

FIGURE 8.6.

rostral margins are rugose delineating the rostral extent of the soft narial tissue. The narrow, blade-like premaxillae are similar

epinasal, that may have borne a second horn. The epinasal is

to those found in other centrosaurines (Figs. 8.4, 8.5). The

more coarsely textured than the nasal horn and is located

rostral portion is deep and forms a simple smooth nasal sep-

dorsal to the rostral portion of the nasal opening, extending

tum that bears one small nutrient foramen, entering ventrally,

rostrally to the suture with the premaxilla. The rostral con-

along its rostral edge adjacent the rostral bone. There is no

tact of the nasal and premaxilla is similar to that seen in

secondary opening, fossa, or narial strut as in chasmosaurines

other centrosaurines. The caudal half of the dorsal margin

(Lehman 1993; Dodson et al. 2004). At the caudal border of the

of the accessory antorbital fenestra is formed by the nasal.

external naris, the ascending process of the premaxilla, along

The sutures of the nasal with the maxilla and the lacrimal are

with the rostroventral border of the nasal, forms a small pro-

discernable along the dorsal margin of the antorbital fossa.

cess as in other centrosaurines (Dodson et al. 2004; Ryan

Sutures along the contact of the nasal with the frontal and

2007a). The caudal portion of the ascending process of the

the palpebral are completely fused and difficult to identify

premaxilla overlaps the nasal and forks to form the rostral and

(Fig. 8.4).

ventral border of an accessory antorbital fenestra. Unlike the

The triangular maxilla is preserved on the left side, whereas

condition in other centrosaurines, the ascending process of

only an impression of the interior of the maxilla is preserved

the premaxilla does not contact the lacrimal; rather, it termi-

on the right side. The maxilla is of typical ceratopsid morphol-

nates well rostral to this element as in chasmosaurines and

ogy. Caudal to its ascending process, the maxilla contacts the

more basal neoceratopsids (Dodson et al. 2004). Ventral to the

lacrimal. The contact with the nasal and lacrimal forms the

nares, the ventral margins of the premaxillae have a narrow,

dorsal margin of the antorbital fossa. The antorbital fenestra,

ventral expansion (offset just medial to the lateral margin) as

as is typical for ceratopsids, is small and lies in a crescent

in other centrosaurines (Ryan 2007b). In centrosaurines that

shaped trough, the caudal margin of which lies between the

preserve this region, this ventral expansion matches a trough

maxillary contact with the lacrimal and the jugal. The antor-

in the caudal ramus of the predentary, where it onlaps the

bital fenestra forms the caudoventral margin of the antorbital

dentary (Ryan 2007b), forming a crushing surface (pers. obs).

fossa, which extends rostrally across the ascending process of

Although present in UMNH VP 16699, this expansion is mod-

the maxilla to the accessory antorbital fenestra (Fig. 8.6). The

est compared to the condition seen in other centrosaurines.

ascending process of the maxilla contacts the premaxilla ros-

The paired nasal bones are completely fused along the mid-

troventrally, forms the caudoventral border of the accessory

line and bear a long (ovate base), low, robust, and subconical

antorbital fenestra, and contacts the nasal along its triangular

narial horn. This horn lies dorsal to the caudalmost edge of

dorsal apex. A well-defined suture is visible between the max-

the nasal opening. Rostral to this is a smaller ossification, or

illa and the jugal and exhibits typical ceratopsid morphology.

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 123

The ventral portion of the palatine contacts the medial surface of the maxilla, and the ectopterygoid extends along the dorsal surface of the maxilla caudal to its ascending process. The pterygoids extend across the medial-dorsal margins of the ectopterygoids, twisting laterally, to contact the most caudal dorsolateral surface of maxilla obscuring the most distal tooth in lateral view. Diabloceratops eatoni is unique in having the caudal end of the maxilla visible rostral to epijugals in lateral view (Fig. 8.4). The maxilla contains 24 tooth rows preserving teeth in various stages of wear and replacement. The continuous wear surface is nearly vertical as in other ceratopsids. Although not visible, we predict that the teeth are double rooted. There are several caudodorsally directed nutrient foramina entering the maxilla above the tooth row along rostral half of the maxilla.

CIRCUMORBITAL REGION AND SKULL ROOF The orbit in Diabloceratops is formed by the jugal, lacrimal, palpebral, and postorbital. The most significant features of this region are the long, erect, postorbital horns (250 mm long from top of the orbits) a feature seen primarily in chasmo-

Skull roof of Diabloceratops eatoni. Blurred right side is reconstructed in this cast.

FIGURE 8.7.

saurines and shared only with Albertaceratops among described centrosaurines (Ryan 2007a). The entire left postorbital horn is preserved, whereas only the medial half of the right

On the skull roof caudal to the postorbital horns, the well-

horn is well preserved, although fragments of the lateral por-

developed frontal fontanelle forms an elongate oval with

tion of the horn were recovered (Figs. 8.4, 8.5). The postorbital

straight sides (Fig. 8.7), as in other centrosaurines (Dodson et

horns are inclined rostrally to a point just caudal to the low

al. 2004). Although the frontal fontanelle displays a range of

narial horn and are gently curved dorsally. The surface texture

interesting morphologies in centrosaurines, this feature has

of the postorbital horns is characterized by shallow, longi-

not yet been shown to be of taxonomic importance within the

tudinal grooves, a feature seen in Albertaceratops and many

group (Sampson et al. 1997).

of the long-horned chasmosaurines (Ryan 2007a). The bases of the horns are robust and subtriangular in cross section, being flattened laterally. The cross section of the right horn

TEMPORAL REGION

demonstrates that there is no sinus in the base of the postor-

The jugal is preserved only on the left side of the skull. The

bital horns. The bases of the postorbital horns along with the

sutures delineating the contacts of the jugal bone with sur-

skull roof bones form a tall, steep vault nearly as deep as the

rounding elements are well defined and are typical of cera-

palpebrals forming a ‘‘forehead’’ that is relatively larger than

topsids. The jugal is more rostrocaudally elongate than in

that of any other ceratopsid. Ryan (2007a) noted a similar

other Neoceratopsia. In line with the ascending process of the

feature, massive ‘‘vaulted frontals,’’ in Albertaceratops (WDCB-

left maxilla, its ventral margin is gently arched dorsally and

MC-001). Complete fusion of the frontals and prefrontals has

laterally. The most prominent feature of the jugal is the ver-

obscured all sutures on the skull roof.

tically oriented, subtriangular, blade-shaped epijugal. This

Robust, blocky, strongly grooved palpebrals form a buttress

epijugal is relatively larger than in any other ceratopsid that

along the rostrodorsal margins of the orbits in continuity with

we have examined and much larger than in any known cen-

the postorbital horns and form the lateral margins of the

trosaurine. Large epijugals are more commonly found in the

‘‘forehead.’’ Ventral to this buttress, the lacrimals are flush

chasmosaurines. The epijugal was highly vascularized as

with the lateral surface of the skull and form a smooth rim

shown by the numerous blood vessel grooves on the epijugal.

along the rostroventral margin of the orbits. The orbits are

The large, wedge-shaped quadratojugal separates the jugal

oval with the long axis of the oval oriented rostrodorsally in

from the quadrate medially and supports the caudal margin of

line with the postorbital horn.

the epijugals. The exterior surface of the jugal is characterized

124 kirkland & deblieux

FIGURE 8.8.

Caudal view of rear of skull of Diabloceratops eatoni. Blurred right side is reconstructed.

by a number of depressions or shallow dimples, about 10–

with the squamosal is obscured rostrally behind the paroccipi-

15 mm across, which may have resulted from an interaction

tal process of the braincase (Fig. 8.8). The main body of the

with the integument. The jugal forms the rostral half of the

quadrate is straight and rostrocaudally flattened with a slight

dorsal margin of the circular infratemporal fenestra, which is

depression extending down its length ventrally. There is an

54 mm wide and 64 mm tall. A triangular process of the jugal

extensive medial (pterygoid) flange that is overlain caudally

bone forms the ventral border of the infratemporal fenestra.

by the pterygoid. Ventral to a moderate constriction in the

The quadrate process of the squamosal makes up the caudal

quadrate shaft, its lateral and medial condyles are 75 mm

margin of the fenestra. The contact between the jugal and

across and form a roughened surface that was the substrate for

squamosal overlying the infratemporal fenestra is obscured by

the cartilaginous articulation for the lower jaw. The medial

a presumed pathology taking the form of a circular opening

condyle is 33 mm across rostrocaudally and the lateral con-

33 mm in diameter in the rostral portion of the squamosal.

dyle is slightly lower and more rostral, measuring 41 mm

This opening overlies and is connected to the infratemporal

across rostrocaudally (Fig. 8.8).

fenestra at the position of this suture. The contacts between

The left pterygoid is completely preserved but has not been

the jugal, quadratojugal, and quadrate are well defined and

fully exposed by preparation. The extensive quadrate wings

similar to those of other ceratopsids (Dodson et al. 2004). The

expand laterally and are broadly forked where they overlap

squamosal bone is discussed below with the frill.

the pterygoid flange. Just rostral to the articulation with the pterygoid process of the basisphenoid, a dorsomedially directed fold is formed that corresponds to the ‘‘eustachian’’

PALATAL REGION

canal (Fig. 8.9) described in Triceratops (Hatcher et al. 1907).

The left quadrate is well preserved and is much like that of

This feature looks similar to that seen in Triceratops as illus-

other centrosaurines we have examined. Its dorsal contact

trated by Hatcher et al. (1907). This is in contrast to other

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 125

FIGURE 8.9.

Right lateral view of Diabloceratops eatoni braincase (UMNH VP 16699).

BRAINCASE

centrosaurines as noted by Sampson (1997) who observed this feature is typically represented by a shallow trough. The complete left palatine is preserved, but portions of it are

As with all adult ceratopsian braincases, the sutures of nearly

still obscured by matrix. Its contacts with the caudodorsal

every element are completely fused. The right side of the

margin of the maxilla, medially, and the pterygoid, caudally,

braincase was removed by erosion prior to discovery, such

are well expressed. Medially, it forms the rostral margin of

that most of the right paroccipital process and basitubera

the ventral notch that is formed caudal to the maxilla be-

are missing and the entire right side of braincase and the cra-

tween it and the quadrate. The dorsal margin of the right pala-

nial nerve openings are well exposed. Thus, all elements of

tine is preserved, where it overlaps the rostrolateral wall of the

the braincase are well represented and the braincase is highly

braincase.

informative.

As discussed above, the ectopterygoid is a narrow element

The supraoccitpital region of the parietal surrounding the

overlapping the caudodorsal surface of the maxilla, lateral to

dorsal side of braincase bears a number of recesses or fossae

the palatines. It is marked by an elongate, rugose tuberosity

for the attachment of nuccal musculature. The supraoccipital

on its exposed dorsal surface as in other centrosaurines (Samp-

forms a long keel above the foramen magnum separating two

son 1993).

of these fossae. Near its dorsal end is a pair of small tabs of

126 kirkland & deblieux

bone that extend caudally. It is well fused with the parietal

passes through the foramen ovale caudal to the opening for

dorsally, which has a deep medial fossa caudal to this contact

the facial nerve (VII).

on the rostroventral surface of the parietal at the base of the parietal bar. The ventral end of the supraoccipital preserves its suture with the underlying exoccipitals.

FRILL

The paroccipital processes are made up of the fused exoc-

In Diabloceratops, the parietal and squamosal bones form the

cipitals and opisthotics. They extend laterally, expanding dor-

frill as in all other ceratopsids. The entire left side of the frill,

soventrally to fit in a groove formed on the medial surface of

including most of the midline parietal bar, is preserved in

the squamosal caudal to articulation with the quadrate. The

UMNH VP 16699 (Figs. 8.3, 8.4). The erect frill is subequal in

ventral margin of the paroccipital extends ventrally below

length to the basal length of the skull and is widest at the

the base of the flange of the squamosal such that the base

laterally directed squamosals, tapering to about half its width

of the paroccipital process is exposed below the quadrate pro-

at the base of a pair of caudal parietal spines separated by a

cess of squamosal in lateral view. Although possibly patho-

midline notch. This gives the frill of Diabloceratops a caudally

logic in UMNH VP 16699, as discussed below, this unusual

tapered appearance, which is unique among the Centro-

condition is symmetrically developed in Centrosaurus ‘‘nasi-

saurinae (Sampson et al. 1997).

cornus’’ (AMNH 5351).

The squamosal is short and square, and forms the rostro-

The foramen magnum is oval to subrectangular (18 mm

ventral margin of the frill. The narrow squamosal bears two

wide and 34 mm tall). The exoccipitals completely enclose the

large epoccipitals laterally and possibly two weak epoccipitals

foramen magnum and form a thin shelf that extends caudally

rostrally (Kirkland and DeBlieux 2007). The caudal contact

from the base of the supraoccipital to extend over the foramen

between the squamosal and parietal bears the ‘‘stepped’’ con-

magnum. A fossa in the base of the exoccipitals, at the base of

dition that defines the Centrosaurinae (Dodson et al. 2004).

the paroccipital processes on either side of the foramen mag-

The squamosal/parietal suture at the lateral margin of frill is

num, has three openings for cranial nerves IX–XII.

separated rostrocaudally by about 10 mm. A gap in the series

The spherical occipital condyle (54.8 mm in diameter) is

of epoccipitals along the lateral frill margin, spanning this

fully fused on a short neck that is ventrally inclined. The

suture between the squamosal and parietal, may have borne

basioccipital makes up the lower one-third of the condyle and

an epoccipital spanning this suture that had been lost. The

is completely fused with the basisphenoid rostrally. At its con-

morphology of the squamosal in the region caudal to the

tact with the basisphenoid, the basioccipital supports a pair of

quadrate, the alleged jugal notch, is distinctive in UMNH VP

massive basal tuberosities of which only the left is completely

16699. In ceratopsids, the squamosal expands laterally such

preserved. The basal tuberosities are highly rugose and flare

that there is a notch between the back of the skull (quadrate)

ventrally and laterally from the base of the braincase. Rostral

and the frill (Dodson et al. 2004). In the type of Diablocera-

to the basal tuberosities the basisphenoid supports a pair of ro-

tops eatoni, Zuniceratops, and more basal neoceratopsids (You

bust basipterygoid processes that angle rostroventrally away

and Dodson 2004), the jugal notch is not developed such that

from the braincase to buttress the pterygoids. A trough be-

the caudal surface of the quadrate and the ventral surface of

tween the basal tuberosities and the basipterygoid processes

the squamosal make a right or obtuse angle. However, it was

marks the route of the carotid artery (Fig. 8.9).

pointed out to us by Andrew Farke (pers. com. 2007) that the

The lateral wall of the braincase is exposed on the incom-

ventral margin of this squamosal maybe pathologic. Close ex-

plete right side of the skull. The laterosphenoid and prootic

amination of this surface does not reveal obvious signs of

are completely fused with each other and the surrounding

pathology showing that, if present, the pathology was well

bones of the braincase. A robust laterosphenoid buttress ex-

healed. It is possible, though, that the straight ventral margin

tends rostrodorsally along the wall of the braincase to join the

and two weakly developed epoccipitals on the ventral margin

postorbital below the base of the horn and separates the me-

may actually be artifacts of a healed pathology. As the right

dial surface of the supratemporal fenestra from a rostral fossa.

squamosal is not preserved it is impossible to test the symme-

The dorsal side of the rostral fossa is partially crushed into the

try of this feature in UMNH VP 16699.

frontal sinus below the horn.

Another interesting feature is a circular opening dorsal to

The openings for cranial nerves IV, II, III, and VI extend in

the infratemporal fenestra at the suture between the squamo-

a line ventrally just rostral to the laterosphenoid buttress

sal and the jugal on the rostral part of the left squamosal. This

(Fig. 8.9.). Extending caudally to the base of the paroccipital

feature is likely pathological, possibly the result of intraspe-

process and ventral to the laterosphenoid buttress, the area

cific conflict (Sampson 1997). However, a recent study by

of the prootic is a shallow fossa with the openings for the

Tanke and Farke (2007) has shown that healed traumatic in-

trigeminal nerves (V2 –V3 and V1) caudal to the opening for VI

juries are quite rare in ceratopsids, and lesions and fenestrae

and rostral to the opening for VII. The auditory nerve (VIII)

not attributable to trauma are found on many centrosaurine

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 127

squamosals. Lacking a right squamosal, we cannot be certain

Diabloceratops sp.

that this opening is pathological, but we think it is unlikely that this feature would have been present on the other side.

Material. The ‘‘Nipple Butte’’ skull (UMNH VP 16704), dis-

The broad parietal bar is broadly convex dorsally and pre-

covered by Joshua A. Smith in 1998 near the top of the lower

serves barely discernable undulations on its dorsal surface, but

sandstone member of the Wahweap Formation (Eaton 1991)

it has no distinct ornamentation. It is ventrally concave in

on the south side of the Wahweap outcrop belt. Collected in

cross section, unlike other centrosaurines. There is a small

2000, the skull was eroded into three large blocks (Fig. 8.2) and

(10 mm) triangular process that projects laterally into the pari-

numerous fragments; it preserves part of the right side of the

etal fenestra near its widest point, 50 mm from the rostral mar-

frill, occipital condyle, caudal braincase, portions of the right

gin of the parietal fenestra (Fig. 8.4). This feature may to be

pterygoid and maxilla, right jugal, a portion of the right or-

related to the connective tissues spanning the parietal fenestra.

bital and antorbital region (lacrimal and palpebral), palate,

Rostrally, the parietal bar overlaps the caudal margin of

left maxilla, and pterygoid, a partial postorbital horncore.

the frontal fontanelle, where it divides into three rounded

Locality. 42Ka586V, Grand Staircase–Escalante National

prongs as in other centrosaurines (Fig. 8.7). The lateral mar-

Monument, on the top of a low ridge on the north side of

gins of the frontal fontanelle are formed by smooth, depressed

Nipple Butte, southwestern Kaiparowits Plateau, Utah.

rostral projections of the parietal forming sulci connecting

Stratigraphic Occurrence. The skull was found in the Wah-

the fontanelle with the supratemporal openings on either

weap Formation in the uppermost sandstone of the lower

side. Penkalski and Dodson (1999) report that this character is

sandstone member, about 50 m above contact with under-

variable even within species, but it appears to be widespread in

lying Drip Tank Member of Straight Cliffs Formation. The

centrosaurines (pers. obs.; Farke 2006).

skull eroded out of a ledge comprising medium-grained,

Laterally, the parietal is thin and bears five triangular epoc-

slabby, cross-bedded sandstone ledge with abundant clay

cipitals (Fig. 8.4), which are slightly flexed dorsally, with the

clasts and bone fragments, including turtle shell fragments

largest near its suture with the squamosal. The epoccipitals are

(one partial turtle) and a variety of isolated teeth (found at the

progressively smaller caudally such that the epoccipital at the

base of the ledge a few meters to the north of the skull).

base of the caudal parietal spine is less than one square centimeter in size. Numerous vascular grooves are appressed into

Description

the lateral surface of the parietal; they are especially prominent on the ventral surface where they form a branching net-

The following description of UMNH VP 16704 is organized by

work. Two robust spines dominate the caudal parietal margin

the groups of elements occurring in each of the three major

and are nearly as long as the entire frill. These spines extend

sandstone blocks that yielded the specimen.

caudally and then gradually sweep laterally beyond the widest portion of frill. Using the classification of marginal processes

BLOCK 1

devised by Sampson et al. (1997), this large spine would correspond to Process 3 and be homologous to the large parietal

The largest block, Block 1, preserves a portion of the right frill

spines seen at this location in Styracosaurus, Einiosaurus, and

including the distal portion of the postorbital, a portion of the

Achelousaurus, among others. There is no indication of any

jugal, the paroccipital process of the exoccipital, most of the

processes medial to the large spines (parietal loci 1 and 2) in

squamosal, and the rostral portion of the medial and lateral

Diabloceratops.

parietal. Also preserved are the braincase, the left pterygoid,

The parietal fenestra is elongate (315 mm long and 170 mm wide rostrally) and, due to the narrowing of the frill caudally,

and portions of both the right and left maxillae and possibly the left palatine.

triangular, in contrast with the oval parietal fenestrae seen in

The right frill fragment in UMNH VP 16704 preserves a large

most centrosaurines. A raised rugose area extends laterally

distally flaring squamosal with a stepped squamosal-parietal

from the parietal bar just rostral to the parietal fenestra and

suture; preserved elements are larger than the corresponding

would have defined the caudal limit of the jaw adductor mus-

elements in UMNH VP 16699 (Fig. 8.10). Two low epoccipitals

cles on the parietal, excluding them completely from the area

are present along the lateral margin of the squamosal. A por-

of the parietal fenestra. The surface texture of the preserved

tion of the lateral parietal displays two long and low epoccipi-

left parietal spine is mottled and marked by a series of longitu-

tals (Fig. 8.10). There is an even larger rostrodorsal displace-

dinal vascular grooves. There is a well-defined collar caudal to

ment of the parietal/squamosal suture on the lateral margin

the base of the caudal spine where the surface texture changes

of the frill, than is observed in UMNH VP 16699, but unlike

from smooth to mottled. Several distinct sections can be iden-

the type specimen of Diabloceratops eatoni, the epoccipitals on

tified in the caudal spine that suggests it grew as a series of

either side of this suture are closely spaced and there would be

stacked conical ossifications.

no space for an epoccipital spanning the suture. The dorsal

128 kirkland & deblieux

sponding region in UMNH VP 16699. The pterygoid is particularly well preserved with the medial side preserving a welldeveloped fold corresponding to the ‘‘eustachian’’ canal described in Triceratops by Hatcher et al. (1907; see also Penkalski and Dodson 1999). The dorsal portion of the braincase, including the foramen magnum, is damaged and poorly preserved.

BLOCK 2 Much of the bone in Block 2 has been lost to erosion, leaving mostly bone impressions. Part of the left orbital region is preserved including what may be the base of a postorbital horn. There is also an internal mold that may be from the nasal region caudal to the external nares.

BLOCK 3 Block 3 preserves much of the right jugal, a portion of the right maxilla (ascending ramus and rostromedial portion), and the rostromedial portion of the left maxilla (Fig. 8.12). The external surface of the jugal has a rugose texture and does not display the dimpled features seen on UMNH VP 16699. The rostromedial portion of the maxillae extends to the midline forming a slightly arching partial palate. The ascending process of the right maxilla and right jugal form a broad arch as in the type of Diabloceratops. The jugal has a large articular facet that compares favorably with the morphology of the articulation of the large epijugal in the type of Diabloceratops, an indication that it may have possessed a large, vertically oriented epijugal. Initially, when only partially prepared, the presence of a large epijugal was not recognized (Kirkland and DeBlieux 2007). The ventral and rostral portion of the right orbit is preserved and is similar to that seen in UMNH VP 16699, though with a dorsorostral margin (palpebral) that is triangular in cross section and less rectangular than in UMNH VP 16699. The antorbital fenestra is preserved, and the trough housing it is shorter than that in the type of Diabloceratops FIGURE 8.10. Preserved frill of Nipple Butte skull (UMNH VP 16704) from Block 1. (A) Dorsal view; (B) ventral view.

but generally similar in morphology. The presence or absence of an accessory antorbital fenestra cannot be determined because of the loss of this area to erosion. It appears that UMNH VP 16704 may have had the vaulted supraorbital ‘‘forehead’’

surface of the squamosal and parietal have a rugose surface

seen in UMNH VP 16699.

texture characterized by many pits and grooves; larger vascular channels are also present (Fig. 8.10). The rostral portion of the parietal fenestra is 234 mm wide with a fairly straight mar-

Age

gin. The preserved rostral margin of the parietal fenestra is

The age of the contact of the Wahweap Formation with the

slightly more oval than the parietal fenestra in UMNH VP

underlying Drip Tank Member of the Straight Cliffs Formation

16699.

is Early Campanian, based on the occurrence of the latest San-

The basicranium preserves the occipital condyle (57.8 mm

tonian ammonite Desmoscaphites in the upper marine tongue

in diameter), both basituberae, and basipterygoid processes,

of the John Henry Member that underlies the Drip Tank (Peter-

the ventral margin of the maxilla, and much of the left ptery-

son 1969; Eaton 1991). A maximum detrital zircon age of 84.0

goid bone (Fig. 8.11). This region is larger than the corre-

Ma from the basal Wahweap Formation (Link et al. 2007) prob-

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 129

FIGURE 8.11.

Additional elements of Nipple Butte skull (UMNH VP 16704) from Block 1. (A) Medial view of braincase and preserved partial left palate; (B) ventral view of braincase and preserved partial left palate; (C) detail of medial view of palate (Fig. 8.11A); (D) lateral view of postorbital horn fragment; (E) distal view of postorbital horn fragment; (F) lateral view of maxilla and pterygoid in Fig. 8.11A, B.

ably represents reworked Upper Santonian (Ogg et al. 2004) strata from the upper John Henry Member of the Straight Cliffs Formation. Jinnah et al. (2007) published an 40Ar/ 39Ar age of 80.1 † 0.3 from an ash near the base of the middle mudstone member of the Wahweap Formation, 54 m above the base of the formation. This is nearly the same stratigraphic level of the Nipple Butte skull and indicates an early Middle Campanian age (Ogg et al. 2004). Given Jinnah’s calculation of a general sediment accumulation rate of 8.2 cm/Ka, the type locality of Diabloceratops (Last Chance skull) would date to 79.57 Ma, well into the Middle Campanian. Thus, although younger than the Late Santonian to Early Campanian faunas from the Menefee Elements of Nipple Butte skull (UMNH VP 16704) from Block 3 and partial right maxilla from Block 1. (A) Preserved right side elements from Block 3; (B) caudolateral portion of right maxilla and ectopterygoid from Block 1. FIGURE 8.12.

Formation (Molenaar 1983; Dyman et al. 1994), Milk River Formation (Braman 2002; Eaton 2006) and lower Two Medicine Formation (Rogers et al. 1993; Chinnery and Horner 2007) these skulls are older than any of the documented centrosaurine faunas from southern Alberta and Montana (Eberth et al. 1992; Brinkman et al. 1998; Sampson and Loewen 2007, this volume).

130 kirkland & deblieux

It is unlikely that the sedimentation rate was continuous

Diabloceratops eatoni exhibits several autapomorphies that

from the base of the Wahweap Formation up into the base of

are unique to currently described centrosaurines. Although a

the overlying Kaiparowits Formation, as proposed by Jinnah et

short, deep preorbital skull is characteristic of other centro-

al. (2007), because the capping sandstone member of the Wah-

saurines (e.g., Dodson et al. 2004), in Diabloceratops this short-

weap Formation represents a reorganization of sediment dis-

ening of the skull is even more apparent. Although an epi-

persal in the region (Pollack 1999; Pollock et al.1999) and per-

nasal fused to the nasal horn is diagnostic of chasmosaurine

haps condensation (Eaton and Nations 1991). Pollack (1999)

ceratopsids, the presence of a distinct epinasal fused to the

suggests a possible unconformity above and below the cap-

nasal rostral to the nasal horn is unique to Diabloceratops. As

ping sandstone. Additionally, there are major faunal differ-

discussed below, an accessory antorbital fenestra is present in

ences between the Wahweap and overlying Kaiparowits For-

some chasmosaurines and more basal neoceratopsians. How-

mation (Cifelli and Madsen 1986; Cifelli 1990a, b, c; Eaton et

ever, this character is known in no other centrosaurines. The

al. 1999; Eaton 2002; Eaton and Kirkland 2003) and, with the

skull roof is distinctive in rising in a steep vault at the base of

recognition that the Aquilan Land Mammal Age is Santonian

the postorbital horns. Ryan (2007a) reports this character in

and not Campanian and the Aquilan fauna is taxonomically

Albertaceratops, but it does not appear to be so strongly devel-

more similar to faunas from the underlying John Henry Mem-

oped in that taxon. The jugals are elongate and inclined ros-

ber of the Straight Cliffs Formation (Eaton 2006); perhaps

trally such that the caudal margin of the maxilla is completely

a distinct Middle Campanian ‘‘Wahweapian’’ Land Mammal

exposed in lateral view. This character is developed rarely in

Age should be recognized. The latest Middle Campanian Fore-

Triceratops, as can be observed in DMNH 48617 (Hatcher et al.

most Formation microvertebrate fauna appears to be dis-

1907; Carpenter 2007). Diabloceratops has the largest epijugals

tinctly different from that of the overlying Upper Campanian

of any known centrosaurine, and their blade-like shape and

Dinosaur Park Formation in southern Alberta (Peng et al.

vertical orientation are distinctly different from the large con-

2001), adding some support to this concept. Given a possible

ical epijugals developed in some chasmosaurines, such as Pen-

condensed interval, or unconformities, near the contact be-

taceratops (Lehman 1993). The frill of Diabloceratops is com-

tween the Wahweap and Kaiparowits formations (Eaton and

pletely different from all other centrosaurines in that it is erect

Nations 1991; Pollack 1999), the sediment accumulation

and widest at its laterally directed squamosals narrowing to

rate in the lower part of the Wahweap Formation may have

half its width caudally. Two elongate caudal spines are nearly

been significantly higher than that proposed by Jinnah et al.

as long as the entire frill. The parietal fenestrae are rostro-

(2007). Therefore, the time difference between the Nipple

caudally elongate (Figs. 8.4, 8.5, 8.13). Additionally, the epoc-

Butte and Last Chance Skulls was probably significantly less

cipitals on the lateral side of parietal decrease in size caudally,

that 0.5 million years.

whereas the epoccipitals increase in size caudally or maintain a relatively constant size in all other centrosaurines (Fig. 8.13). Chasmosaurines typically decrease the size of the epoccipitals along the lateral margin of the frill caudally, but in chasmo-

Comparisons

saurines this occurs along the margin of their elongate squamosals (Dodson et al. 2004).

CENTROSAURINES

We conducted a preliminary phylogenetic analysis of DiDiabloceratops can be included within the Centrosaurinae on

abloceratops in 2006, utilizing the data set developed by Ryan

the basis of possessing a caudal narial process and a stepped

(2003) for his revision of the Centrosaurinae. This analysis

suture between the squamosal and parietal (Dodson et al.

indicated that Diabloceratops was the sister taxon of Alberta-

2004; Ryan 2007a). Additionally, the morphology of its fron-

ceratops and the other more derived centrosaurines, although

tal fontanelle is shared by other centrosaurines (Dodson et al.

its inclusion in the data set led to a significant loss of phy-

2004).

logenetic resolution among the more derived centrosaurines

The long postorbital horns and small narial horn in D.

(Kirkland and DeBlieux 2006). Further preparation of the lat-

eatoni are primitive character states within the Ceratopsidae

eral margin of the snout supports this result, but indicates

as indicated by the ceratopsid sister taxon Zuniceratops (Wolfe

that, in light of these new observations, further comparisons

and Kirkland 1998; Wolfe 2000; Wolfe et al. 2007; Ryan

with advanced basal neoceratopsians from Asia are needed to

2007a). Diabloceratops and the recently described Albertacera-

fully understand the phylogenetics of the Ceratopsidae. Merg-

tops (Ryan 2007a) are the first centrosaurines discovered that

ing the data sets for the more basal neoceratopsians (You and

have large postorbital horns and small narial horns. These

Dong 2003; You and Dodson 2004; Makovicky and Norrell

discoveries confirm predictions made after the discovery of

2006; Chinnery and Horner 2007) and the Ceratopsidae (Dod-

Zuniceratops that long horned centrosaurines should be found

son et al. 2004; Ryan 2007a) should be done, but is beyond the

in the fossil record.

scope of this descriptive paper.

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 131

FIGURE 8.13.

Parietals of various genera of centrosaurines with epoccipital nomenclature developed by Sampson et al. (1997) indicated by numbers. (A) Diabloceratops; (B) Albertaceratops; (C) Pachyrhinosaurus; (D) Achelousaurus; (E) Einiosaurus; (F) Centrosaurus; (G) Styracosaurus. Grey areas indicate actual bone preserved. Scale bars are 10 cm. Modified after Sampson et al. (1997) and Ryan (2003).

FRILL ORNAMENTATION

Albertaceratops and Diabloceratops, the nomenclature used for describing the ornamentation of the frill should be reconsid-

Ornamentation of the adult centrosaurine frill is thought to

ered in light of phylogenetic patterns as well as ontogenetic

be sexually selected for display and species recognition (Samp-

patterns.

son et al. 1997) and has proven critical to understanding the phylogenetic relationships. Sampson et al. (1997) developed a nomenclature for designating proposed homologies in the ep-

ACCESSORY ANTORBITAL FENESTRA

occipitals of centrosaurines using a numbering system from

Another feature of considerable importance is the presence of

the midline out (Fig. 8.13). Epoccipital 3 is a large spine proj-

an accessory antorbital fenestra in the side of the skull behind

ecting from the back of the frill in all genera except Cen-

the nasal opening at the front of the antorbital fossa. This

trosaurus, and Dodson et al. (2004) proposed that this is an

fenestra is not present in the more advanced centrosaurines,

ambiguous autapomorphy for centrosaurines. Given its dis-

but is present in Zuniceratops, where it is also bounded by

tinctiveness, it is often referred to as the parietal spine. In this,

the premaxilla, nasal, and maxilla (Wolfe et al. 2007). Two

we agree, but on examining the spine in Diabloceratops, we

more basal species of protoceratopsid-grade neoceratopsians

come to the conclusion that these large parietal spines may

Bagaceratops (= Breviceratops, Lamaceratops, and Platyceratops)

accrete by the fusion of multiple epoccipital ossifications, en-

(Maryanska and Osmolska 1975; Witmer 1997; Sereno 2000;

larging the spine through ontogeny. In this, it would be fun-

Alifanov 2003) and Magnirostris (You and Dong 2003) have a

damentally different from the single step fusion that occurs at

similar accessory antorbital fenestra, in which the nasal is ex-

other epoccipital loci. Additionally, when considered with Al-

cluded from the margin of the opening (Fig. 8.14).

bertaceratops (Ryan 2007a), it appears that the parietal spine

The occurrence of Protoceratops hellenikorhinus with Magni-

was the initial parietal ornamentation developed in centro-

rostris dodsoni at Bayan Mandahu, Inner Mongolia, China,

saurines, with all more derived centrosaurines developing a

suggests that these strata are a somewhat different age than

medially directed spine at the lateral margins of the parietal

the Barun Goyot Formation of Mongolia where Bagaceratops

notch (process 2). We have observed that Styracosaurus and

occurs (Maryanska and Osmolska 1975; Lambert et al. 2001;

Centrosaurus both develop a fold rostral to the parietal notch

You and Dong 2003). Eberth (1993) has correlated the redbeds

that variably develops with fusion of an epoccipital at each

at Bayan Mandahu with the Djadokhta Formation of south-

the fold into a pair of rostrally oriented spines (process 1).

ern Mongolia, further suggesting that Magnirostris and Baga-

Thus, both processes 1 and 2 are derived characters of more

ceratops did not occur contemporaneously. Together with the

restricted distribution (Ryan 2007a). Given the discovery of

presence of postorbital horns and elongate rostral bone, the

132 kirkland & deblieux

FIGURE 8.14.

Holotype of Magnirostris dodsoni in the Institute of Vertebrate Paleontology and Paleoanthropology (IVPP V12513) missing both squamosals and parietals with reconstructed left jugal, left quadrate, and partial left postorbital. (A) Right side; (B) rostral view; (C) left side; (D) dorsal view; (E) caudal view; (F) ventral view; (G) left oblique view of nasal and orbital area. Left orbital horn partially preserved and depressed along fracture.

synonymy of Magnirostris with Bagaceratops suggested by

belli, or Chasmosaurus irvinensis, which have reduced brow

Makovicky and Norell (2006) is difficult to support.

horns. Thus, the accessory foramen is lost within the well-

Lehman (1993) recognized a tiny accessory antorbital fora-

established Chasmosaurus line of the northern Western Inte-

men at the junction of the premaxilla, nasal, and maxilla in

rior (Holmes et al. 2001). The repeated loss of the accessory

Pentaceratops (AMNH 1624) and in the skull of Agujaceratops

antorbital fenestra in more derived ceratopsids may have re-

(TMM 43098-1) described by Forster et al. (1993), although we

sulted from selection for a more solid facial area due to the

have observed that preparation is such that it is visible only on

stresses imparted by intraspecific grappling combat (Sampson

the right side of skull. It is impossible to determine the pres-

1997).

ence of this feature in the poorly preserved type material of

However, whereas the accessory antorbital fenestra is bor-

Agujaceratops mariscalensis, although its presence is suggested

dered by the premaxilla and maxilla in Bagaceratops and Mag-

in the line art of the maxilla (Lehman 1989). We noted this

nirostris, the nasal contributes to the dorsal border of the ac-

feature in fossil material representing a new chasmosaurine

cessory fenestra in Zuniceratops and in all ceratopsids, where it

from the Kaiparowits Formation in Utah (Kaiparowits new

is recognized. Magnirostris also possesses small, but distinct,

taxon B, Sampson and Loewen 2007), indicating its presence

rugose, postorbital horns not present in any specimens of

in all the southern chasmosaurines of the Late Campanian.

Bagaceratops (Fig. 8.14). Thus, Diabloceratops eatoni, together

We also observed that a tiny accessory antorbital foramen

with Zuniceratops, provide substantial evidence that among all

is present in individuals of Chasmosaurus (AMNH 5401, TMP

the known Asian protoceratopsians, Magnirostris is the sister

81.19.175) with large brow horns, and is not present in any

taxon to the large horned ceratopsids of North America (Fig.

known specimens of Chasmosaurus russelli, Chasmosaurus

8.15). This repudiates the monophyly of the protoceratopsid

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 133

FIGURE 8.15. Hypothesized ceratopsian family tree plotted against a linear time scale in millions of years ago (MYA) and Cretaceous rocks in the Grand Staircase–Escalante National Monument. Note: frill is not preserved on figured skull of Magnirostris.

neoceratopsians of Asia (Sereno 2000; You and Dodson 2004;

the premaxilla across the dorsal side of the maxilla. The large

Makovicky and Norrell 2006; Chinnery and Horner 2007).

epijugal in Diabloceratops is a character shared with many

Additionally, these hypotheses lead us to predict that, when

chasmosaurines. Although differing in position, the presence

more complete specimens of the central Asian advanced neo-

of an epinasal is shared only with chasmosaurines. These addi-

ceratopsian Turanoceratops (Nessov et al. 1989; Sereno 2000)

tional characters would also qualify as shared primitive char-

are described, they will be found to have a well-developed

acter states with the chasmosaurines if our phylogenetic hy-

accessory antorbital fenestra.

potheses are correct (Fig. 8.15).

CHASMOSAURINES Diabloceratops shares a number of features with the chasmo-

COMPARISON OF NIPPLE BUTTE AND LAST CHANCE SKULLS

saurines. The large postorbital horns and small narial horn are

Ontogeny. The Nipple Butte skull appears to be more mature

well established as primitive shared character states (Wolfe

than the Last Chance skull in that the occipital condyle is a bit

and Kirkland 1998; Dodson et al. 2004; Ryan 2007a). How-

larger, and the epoccipitals are fused into the margin of the

ever, there are a number of other characters that are not recog-

frill. The caudally recurved tip of the postorbital horns in the

nized in any other centrosaurines. In particular, the caudal

holotype of Diabloceratops may indicate that this is a young

process of the premaxilla does not contact the lacrimal in

adult individual (Horner and Goodwin 2006, 2008; Goodwin

Diabloceratops or in any chasmosaurines. The closure of the

and Horner 2007, this volume). The adult age of the holotype

accessory antorbital opening in more derived centrosaurines

of Diabloceratops is suggested by the complete fusion of the

may have resulted from the extension of the caudal process of

nasals, a fused epinasal, and the exclusion of the supraoccipi-

134 kirkland & deblieux

tal from the foramen magnum (Horner and Goodwin 2006). Also, the ectopterygoid is more fully fused to the maxilla in the Nipple Butte Skull forming little more than a roughened bulge for muscle attachment, whereas the ectopterygoid in the Last Chance skull exhibits a distinct sutural contact with the maxilla. However, in contradiction to the hypothesis that the Nipple Butte skull is more mature, the epijugal is not attached to the Nipple Butte skull. Additionally, the squamosal/ parietal suture is open in both skulls. The surface morphology of ceratopsian periosteal bone of the parietals and squamosals has been used as a means of determining the relative ontogenetic age of individuals (Sampson et al. 1997; Tumarkin-Deratzin 2003; Brown et al. 2007). Both specimens described here have frills with areas of smooth, mottled, and rugose surface textures that are associated with adult specimens. No long-grained, striated bone indicative of juveniles is seen on either of these two specimens. Based on bone surface texture, along with other lines of evidence mentioned above, we propose that UMNH VP 16699 was fully adult but not an extremely old individual based on the lack of large areas of extremely rugose bone. The frill of UMNH VP 16704 generally displays more rugose bone, especially on the dorsal surface of the lateral part of the parietal, than does the frill of UMNH VP 16699, a fact that fits well with other indications that this may have been an older individual. Taxonomy. Finally, we must address whether the Nipple Butte specimen represents Diabloceratops eatoni. The stepped squamosal suture and the presence of a robust postorbital horn

FIGURE 8.16. Comparison of Santonian to Middle Campanian centrosaurine squamosals from the southwestern United States. (A) Last Chance right squamosal (UMNH VP 16699); (B) Nipple Butte right squamosal (UMNH VP 16702; arrowheads indicate positions of epoccipitals); (C) Menefee squamosal (NMMNHS P-25052). Solid arrows indicate position of epoccipitals. Open arrows indicate position of suspect apparent epoccipitals. Drawn in same orientation.

indicate that the Nipple Butte skull represents a basal, longhorned centrosaurine (Kirkland et al. 2002; Ryan 2003, 2007a; Kirkland and DeBlieux 2007). Although there appear to be dif-

and the lack of a preserved right squamosal for comparison; it

ferences in the relative maturity of the two skulls, they both are

is possible that the square rostral margin of the lateral squam-

about the same size and are the smallest known centrosau-

osal in UMNH VP 16699 may also be pathological, as noted in

rines. The jugals are very similar in being elongate and inclined

the description. The exposure of the base of the paroccipital

caudally with a dorsoventrally deep epijugal overlapping the

process laterally below the squamosal may reflect this pathol-

suture between the jugal and quadratojugal. The rostral mar-

ogy, except this condition is present in some specimens of

gin of the parietal fenestra of both skulls is relatively straight

Centrosaurus apertus; symmetrically in AMNH 5351 (type of

and the fenestra appears to taper caudally in both skulls.

Centrosaurus nasicornis) and asymmetrically in NMC 348 (type

Nonetheless, there are important differences in the rostral

of Centrosaurus flexus). Additional definitive Diabloceratops

portion of the frill. In all other centrosaurines (including the

material will be needed to determine whether the squared-off

Nipple Butte skull) and in all chasmosaurines, the squamosal

squamosal is pathological, so we hesitate to give much weight

expands laterally (Fig. 8.16) so that there is a notch housing

to this character at this time.

the external ear between the back of the skull and the frill. In

Differences in the positioning of the epoccipitals across the

Diabloceratops eatoni, Zuniceratops, and more basal neocera-

squamosal-parietal suture are significant. Whereas differences

topsians, the notch is not developed and the caudal surface of

in maturity would explain the variation in the development

the quadrate and the ventral surface of the squamosal make a

of the epoccipitals in these two skulls, it would not account for

right or obtuse angle, in contrast with an acute angle seen

the differences in their relative position. There is a large gap

in all other ceratopsids. Taken together, these observations

between the most rostral epoccipital on the preserved lateral

would indicate that Diabloceratops eatoni is basal to the other

margin of the parietal and the most caudal epoccipital on the

known centrosaurines (Fig. 8.15). However, it is critical to

right squamosal in the Last Chance skull. A missing epoccipi-

note that, based on the presence of a presumed pathological

tal may well have straddled this suture as in many centro-

opening on the squamosal above the infratemporal fenestra

saurines (e.g., Styracosaurus; Ryan 2007b) and in Triceratops

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 135

FIGURE 8.17.

Reconstructed skull of Diabloceratops eatoni. (A) Dorsal view; (B) oblique lateral view; (C) caudal view; (D) rostral view.

(Horner and Goodwin 2008). In the reconstruction of this

Conclusions

skull (Fig. 8.17), Rob Gaston included an epoccipital in this position on the reconstructed right side of the frill. In the

Two species of basal centrosaurine ceratopsians are recognized

Nipple Butte skull, epoccipitals on the parietal and squamosal

from the lower Middle Campanian Wahweap Formation of

occur close to this suture, such that there would be no room

southern Utah: Diabloceratops eatoni and cf. Diabloceratops. Of

for any intermediate epoccipital straddling this suture (Figs.

the many dinosaur taxa documented from microvertebrate

8.4, 8.10). The squamosal in the Nipple Butte skull appears to

sites in the Middle Campanian Wahweap Formation of south-

be similar in morphology to the older, and much larger cen-

ern Utah, Diabloceratops is the first dinosaur that has been de-

trosaurine (NMMNHS P-25052; Fig. 8.16) from the Menefee

scribed to species (Fig. 8.18). At present, Diabloceratops is

Formation in New Mexico (Williamson 1997). Because the or-

the most southern and oldest diagnosable centrosaurine, al-

namentation of the frills in centrosaurines is critical taxonom-

though there are less complete, Late Campanian centrosau-

ically (Sampson 1993, 1996; Dodson et al. 2004; Ryan 2003,

rines known from Coahuila, Mexico (Murray et al. 1960; Kirk-

2007a, b), it is unlikely that the Nipple Butte and Last Chance

land et al. 2000; Loewen et al. this volume) and southern

skulls represent the same species. There are some shared fea-

Arizona (Heckert et al. 2003) and older. There is also an older

tures that may indicate that it may belong in the same genus

and more southern centrosaurine fossil from the Menefee For-

so we have tentatively placed it in cf. Diabloceratops. However,

mation of northwestern New Mexico (Williamson 1997). To-

until more specimens of Diabloceratops eatoni are collected, it

gether with Zuniceratops, Diabloceratops provides important

will be impossible to determine which of the many characters

supporting data regarding character state distributions at the

currently used to diagnose the species are variable within the

base of the Ceratopsidae and ceratopsid relationships with the

species, can diagnose the genus, or are shared by as yet un-

known more basal neoceratopsians of Asia.

known sister genera.

136 kirkland & deblieux

FIGURE 8.18.

Life reconstruction of the head of Diabloceratops eatoni.

We thank Jim Gardner of the Royal Tyrrell Museum of Palae-

Acknowledgments

ontology, Mike Getty of the Utah Museum of Natural History, Carl Mehling of the American Museum of Natural History, Tim All excavations were conducted under BLM permit numbers

Rowe of the Texas Memorial Museum, Kevin Seymour of the

UT-S-00-001, UT-EX-03-007, and UT-EX-05-026. Funding

Royal Ontario Museum, Kieran Sheperd of the Canadian Mu-

was provided by the BLM through assistance agreement

seum of Nature, and Tom Williamson of the New Mexico Mu-

JSA015002 and the UGS. Alan Titus is thanked for his efforts

seum of Natural History for the opportunity to study their

to facilitate this research at GSENM. Marietta Eaton and Dave

ceratopsian specimens. We thank Hai-Lu You and Xu Xing for

Hunsaker at the GSENM are thanked for their assistance. Help

arranging for JIK to study the holotype specimen of Magni-

in the field was provided by Bob and Linda Baldazzi, Jane De-

rostris dodsoni IVPP V 12513. Discussions with Peter Dodson,

Blieux, Walt Elkington, Bucky Gates, Joe Gentry, Mike Getty,

Andrew Farke, Pete Makovicky, Mark Lowen, Lukas Panzarin,

Martha Hayden, Don and Sheila Hughes, Mark Loewen, Tom

and Mike Ryan are appreciated. We thank Francois Gohier for

Mellenthin, Andrew Milner, Sandy Mosconi, George Muller,

many of the ceratopsian skull photographs used in Fig. 8.15.

Phil Policelli, Joshua A. Smith, Steve and Sally Stevenson, Alan

Brad Wolverton’s skillful rendering of Diabloceratops in Figs.

Titus, Daryl and Terry Wade, Dave Wilcots, Bill and Arlene

8.4A and 8.18 is appreciated. Scott Hartman suggested the

Yensen, David Zivcovic, Zion Helitack, and the UGS. We thank

generic name. Reviews by Jennifer Cavin, Peter Dodson, Dave

Rob Gaston for his skillful molding, casting, and reconstruc-

Eberth, Martha Hayden, Mike Lowe, and Pete Makovicky are

tion of UMNH VP 16699.

appreciated.

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 137

References Cited Alifanov, V. R. 2003. Two new dinosaurs of the infraorder Neoceratopsia (Ornithischia) from the Upper Cretaceous of the Nemegt depression, Mongolian People’s Republic. Paleontological Journal 37: 524–534. Braman, D. R. 2002. Terrestrial palynomorphs of the upper Santonian-?lowest Campanian Milk River Formation, southern Alberta, Canada. Palynology 25: 57–107. Brinkman, D. B., M. J. Ryan, and D. A. Eberth. 1998. The paleogeographic and stratigraphic distribution of ceratopsids (Ornithischia) in the Upper Judith River Group of Western Canada. Palaios 13: 160–169. Brown, C. M., A. P. Russell, and M. J. Ryan. 2007. Size-associated surficial bone texture changes of the centrosaurine frill: Patterns and implications. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 8–10. Drumheller: Royal Tyrrell Museum of Palaeontology. Carpenter, K. 2007. ‘‘Bison’’ alticornis and O. C. Marsh’s early views on ceratopsians. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 349–364. Bloomington: Indiana University Press. Chinnery, B. J., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Cifelli, R. L. 1990a. Cretaceous mammals of southern Utah, I. Marsupials from the Kaiparowits Formation ( Judithian). Journal of Vertebrate Paleontology 10: 295–319. ———. 1990b. Cretaceous mammals of southern Utah, II. Marsupials and marsupial-like Mammals from the Wahweap Formation (early Campanian). Journal of Vertebrate Paleontology 10: 320–331. ———. 1990c. Cretaceous mammals of southern Utah. IV. Eutherian mammals from the Wahweap (Aquilan) and Kaiparowits ( Judithian) Formations. Journal of Vertebrate Paleontology 10: 346–360. Cifelli, R. L., and S. K. Madsen. 1986. An Upper Cretaceous Symmetrodont (Mammalia) from southern Utah. Journal of Vertebrate Paleontology 6: 258–263. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Dyman, T. S., E. A. Merewether, C. M. Molenaar, W. A. Cobbam, J. D. Obradovich, R. J. Weimer, and W. A. Bryant. 1994. Stratigraphic transects for Cretaceous rocks, Rocky Mountains and Great Plains Regions. In M. V. Caputo, J. A. Peterson, and K. J. Franczyk, eds., Mesozoic Systems of the Rocky Mountain Region, USA, pp. 365–391. Denver: Rocky Mountain Section SEPM (Society for Sedimentary Geology). Eaton, J. G. 1991. Biostratigraphic framework for the Upper Cretaceous rocks of the Kaiparowits Plateau, southern Utah. In J. D. Nations and J. G. Eaton, eds., Stratigraphy, Depositional Environments, and Sedimentary Tectonics of the Southwestern Margin Cretaceous Western Interior Seaway, pp. 47–63. Geological Society of America Special Paper 260.

138 kirkland & deblieux

———. 2002. Multituberculate mammals from the Wahweap (Campanian, Aquilan) and Kaiparowits (Campanian, Judithian) formations, within and near the Grand Staircase–Escalante National Monument, southern Utah. Utah Geological Survey Miscellaneous Publication no. 02-4: 1–66. ———. 2006. Santonian (Late Cretaceous) mammals from the John Henry Member of the Straight Cliffs Formation, Grand Staircase–Escalante National Monument, Utah. Journal of Vertebrate Paleontology 26: 446–460. Eaton, J. G., R. L. Cifelli, J. H. Hutchinson, J. I. Kirkland, and J. M. Parrish. 1999. Cretaceous vertebrate faunas of the Kaiparowits Basin (Cenomanian-Campanian), southern Utah. In D. Gillette, ed., Vertebrate Paleontology in Utah, pp. 345–353. Utah Geological Survey, Miscellaneous Publication 99-1. Eaton, J. G., and J. I. Kirkland. 2003. Nonmarine Cretaceous vertebrates of the Western Interior Basin. In P. J. Harries, ed., High-Resolution Approaches in Stratigraphic Paleontology; Topics in Geobiology, Vol. 21, pp. 263–313. Boston: Kluwer Academic Publishers. Eaton, J. G., and J. D. Nations. 1991. Introduction; Tectonic setting along the margin of the Cretaceous Western Interior Seaway southwest Utah and northern Arizona. In J. D. Nations and J. G. Eaton, eds., Stratigraphy, Depositional Environments, and Sedimentary Tectonics of the Southwestern Margin Cretaceous Western Interior Seaway, pp. 1–8. Geological Society of America Special Paper 260. Eberth, D. A. 1993. Depositional environments and facies transitions of dinosaur-bearing Upper Cretaceous redbeds at Bayan Mandahu (Inner Mongolia, People’s Republic of China). Canadian Journal of Earth Sciences 30: 2196–2213. Eberth, D. A., R. G. Thomas, and A. Deino. 1992. Preliminary K-Ar dates from bentonites in the Judith River and Bearpaw formations (Upper Cretaceous) of Dinosaur Provincial Park, southern Alberta, Canada. In N. J. Mateer and P.-J. Chen, eds., Aspects of Nonmarine Cretaceous Geology, pp. 296–304. Beijing: China Ocean Press. Farke, A. A. 2006. Evolution and anatomical origin of the frontal sinus complex in ceratopsian dinosaurs. Journal of Vertebrate Paleontology 26(3, Suppl.): 59A. Forster, C. A., P. C. Sereno, T. W. Evans, and T. A. Rowe. 1993. A complete skull of Chasmosaurus mariscalensis (Dinosauria: Ceratopsia) from the Aguja Formation (late Campanian) of west Texas. Journal of Vertebrate Paleontology 13: 161–170. Foster, J. R., A. L. Titus, G. Winterfeld, and M. Hayden. 2001. Paleontological Survey of the Grand Staircase–Escalante National Monument, Garfield and Kane Counties, Utah. Utah Geological Survey Special Publication SS-99: 1–98. Goodwin, M. B., and J. R. Horner. 2007. Historical collecting bias and the fossil record of Triceratops. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 61– 66. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2010. Historical collecting bias and the fossil record of Triceratops in Montana. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 551–563. Bloomington: Indiana University Press.

Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49. Heckert, A. B., S. G. Lucas, and S. E. Krzyzanowski. 2003. Vertebrate fauna of the Late Cretaceous ( Judithian) Fort Crittenden Formation, and the age of Cretaceous vertebrate faunas of southeastern Arizona (U.S.A.). Neues Jahrbuch für Geologie and Paläontologie Abhandlungen 227: 343–364. Holmes, R. B., C. A. Forster, M. J. Ryan, and K. M. Shepherd. 2001. A new species of Chasmosaurus (Dinosauria: Ceratopsia) from the Dinosaur Park Formation of southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Horner, J. R. 1992. Cranial morphology of Prosaurolophus (Ornithischia: Hadrosauridae) with descriptions of two new hadrosaurid species and an evaluation of hadrosaurid phylogenetic relationships. Museum of the Rockies Occasional Papers 2. Horner, J. R., and M. B. Goodwin. 2006. Major cranial changes during Triceratops ontogeny. Proceedings of the Royal Society, Biological Sciences 273: 2757–2761. ———. 2008. Ontogeny of cranial epi-ossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Jinnah, Z., A. Deino, T. Gates, and E. Roberts. 2007. The first 40Ar/ 39Ar age date from the Wahweap Formation (Late Cretaceous of Utah): implications for faunal correlation. Journal of Vertebrate Paleontology 27(3, Suppl.): 96A. Kirkland, J. I. 2001. The quest for new dinosaurs at Grand Staircase–Escalante National Monument; Utah Geological Survey. Survey Notes 37: 1–4. Kirkland, J. I., and D. D. DeBlieux. 2005. Dinosaur remains from the lower to middle Campanian Wahweap Formation at Grand Staircase–Escalante National Monument, southern Utah. Journal of Vertebrate Paleontology 25(3, Suppl.): 78A. ———. 2006. A new genus of ornate long-horned centrosaurine ceratopsian from the Middle Campanian Wahweap Formation, Grand Staircase–Escalante National Monument, southern Utah. Journal of Vertebrate Paleontology 26(3, Suppl.): 85A. ———. 2007. New Centrosaurine ceratopsians from the Wahweap Formation, Grand Staircase–Escalante National Monument, southern Utah. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 90–95. Drumheller: Royal Tyrrell Museum of Palaeontology, Kirkland, J. I., D. D. DeBlieux, J. Smith, and S. Sampson. 2002. New ceratopsid cranial material from the Lower Campanian Wahweap Formation, Grand Staircase–Escalante National Monument, Utah. Journal of Vertebrate Paleontology 22(3, Suppl.): 74A. Kirkland, J. I., R. Hernandez-Rivera, M. C. Aguillon-Martinez, C. R. Delgado de Jesus, R. Gomez-Nunez, and I. Vallejo. 2000. The Late Cretaceous Difunta Group of the Parras Basin, Coahuila, Mexico and its vertebrate fauna. Universidad Autonoma del Estado de Hidalgo, Avances en Invesigacion 3: 133–172. Lambe, L. M. 1915. On Eoceratops canadaensis, gen. nov., with remarks on other genera of Cretaceous horned dinosaurs. Geological Survey of Canada Museum Bulletin 12: 1–49. Lambert, O., P. Godefroit, H. Li, C.-Y. Shang, and Z. M. Dong. 2001. A new species of Protoceratops (Dinosauria, Neoceratopsia) from the Late Cretaceous of Inner Mongolia (P. R. China).

Bulletin de l’Institut Royal des Sciences Naturalles Belgique, Science de la Terre 71: 5–28. Lehman, T. M. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. ———. 1993. New data on the ceratopsian dinosaur Pentaceratops sternbergii Osborn from New Mexico. Journal of Paleontology 67: 279–288. Link, P. K., E. Roberts, C. M. Fanning, and J. S. Larsen. 2007. Detrital zircon age populations from Upper Cretaceous and Paleogene Wahweap, Kaiparowits, Canaan Peak, Pine Hollow, and Claron Formations, Kaiparowits Plateau, southern Utah. Geological Society of America Abstracts with Programs 39: 7. Loewen, M. A., S. D. Sampson, E. K. Lund, A. A. Farke, M. C. Aguillón-Martínez, C. A. de Leon, R. A. Rodríguez-de la Rosa, M. A. Getty, and D. A. Eberth. 2010. Horned dinosaurs (Ornithischia: Ceratopsidae) from the Upper Cretaceous (Campanian) Cerro del Pueblo Formation, Coahuila, Mexico. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 99–116. Bloomington: Indiana University Press. Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Marsh, O. C. 1888. A new family of horned dinosaurs from the Cretaceous. American Journal of Science, Series 3, 36: 477–478. ———. 1890. Description of new Dinosaurian reptiles. American Journal of Science, Series 3, 39: 82–86. Maryanska, ´ T., and H. Osmólska. 1975. Protoceratopsidae (Dinosauria) of Asia. Acta Palaeontologica Polonica 33: 133–181. Molenaar, C. M. 1983. Major depositional cycles regional correlations of Upper Cretaceous rocks, southern Colorado Plateau and adjacent areas. In M. W. Reynolds and E. D. Dolly, eds., Mesozoic Paleogeography of the West-central United States, pp. 201–224. Denver: Rocky Mountain Section Society of Economic Paleontology and Mineralogy. Murry, G. E., D. R. Boyd, J. A. Wolleben, and J. A. Wilson. 1960. Late Cretaceous fossil locality, eastern Parras Basin, Coahuila, Mexico. Journal of Paleontology 34: 368–373. Nessov, L. A., L. F. Kaznyshkina, and G. O., Cherepanov. 1989. Mesozoic ceratopsian dinosaurs and crocodiles of central Asia. In T. N. Bagdanova and L. I. Khozatskii, eds., Theoretical and Applied Aspects of Modern Paleontology, pp. 383–405. Leningrad: Nauka Publishers. Ogg, J. G., F. P. Agterberg, and F. M. Gradstein. 2004. The Cretaceous Period. In F. M. Gradstein, J. G. Ogg, and A. G. Smith, eds., A Geological Time Scale 2004, pp. 344–383. Cambridge: Cambridge University Press. Peng. J., A. P. Russell, and D. B. Brinkman. 2001. Vertebrate microsite assemblages (exclusive of mammals) from the Foremost and Oldman Formations of the Judith River Group (Campanian) of southeastern Alberta: An illustrated guide. Provincial Museum of Alberta, Natural History Occasional Paper 25. Penkalski, P., and P. Dodson. 1999. The morphology and system-

New Basal Centrosaurine Ceratopsian Skulls from the Wahweap Formation 139

atics of Avaceratops, a primitive horned dinosaur from the Judith River Formation (Late Campanian) of Montana, with a description of a second skull. Journal of Vertebrate Paleontology 19: 692–711. Peterson, F. 1969. Four new members of the Upper Cretaceous Straight Cliffs Formation in southeastern Kaiparowits region, Kane County, Utah. U. S. Geological Survey Bulletin 1274-J: 1–28. Pollock, S. L. 1999. Provenance, geometry, lithofacies, and age of the Upper Cretaceous Wahweap Formation, Cordilleran foreland basin, southern Utah. M.Sc. thesis, New Mexico State University, Las Cruces. Pollock, S. L., T. F. Lawton, and R. A. J. Robinson. 1999. Provenance and geometry of Upper Cretaceous Wahweap Formation, southern Utah—Record of Foreland partitioning. Geological Society of America Abstracts with Programs 31: A-426. Rogers, R. R., C. C. Swisher, and J. R. Horner. 1993. 40Ar/ 39Ar age and correlation of the nonmarine Two Medicine Formation (Upper Cretaceous) northwestern Montana, USA. Canadian Journal of Earth Sciences 30: 1066–1075. Ryan, M. J. 2003. Taxonomy, systematics and evolution of centrosaurine ceratopsids of the Campanian Western Interior Basin of North America. Ph.D. diss., University of Calgary, Calgary. ———. 2007a. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. ———. 2007b. A revision of the Late Campanian centrosaurine ceratopsid genus Styracosaurus from the western interior of North America. Journal of Vertebrate Paleontology 27: 944–962. Sampson, S. D. 1993. Cranial ornamentation in ceratopsid dinosaurs: Systematic, behavioral, and evolutionary implications. Ph.D. diss., University of Toronto, Toronto. ———. 1996. Two new horned dinosaurs from the Two Medicine Formation of Montana, U.S.A., with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsia). Journal of Vertebrate Paleontology 15: 743–760. ———. 1997. Dinosaur combat and courtship. In J. O. Farlow and M. K. Surman, eds., The Complete Dinosaur, pp. 383–393. Bloomington: Indiana University Press. Sampson, S. D., and M. A. Loewen. 2007. New information on the diversity, stratigraphic distribution, biogeography, and evolution of ceratopsid dinosaurs. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 125– 133. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2010. Unraveling a radiation: A review of the diversity, stratigraphic distribution, biogeography, and evolution of horned dinosaurs (Ornithischia: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 405–427. Bloomington: Indiana University Press. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337.

140 kirkland & deblieux

Seeley, H.G. 1888. The classification of the Dinosauria. Report of the British Association for the Advancement of Science 1887: 698– 699. Sereno, P. C. 1986. Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). National Geographic Research 2: 234–256. ———. 2000. The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia. In M. Benton, M. Sishkin, D. Unwin, and E. Kuronchkin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. 480–516. Cambridge: Cambridge University Press. Stokstad, E. 2001. Utah’s fossil trove beckons, and tests, researchers: At Grand Staircase–Escalante National Monument, patience and muscle power pays off in paleontological riches. Science 294: 41–43. Tanke, D. H., and A. A. Farke. 2007. Bone resorption, bone lesions, and extracranial fenestrae in ceratopsid dinosaurs: A preliminary assessment. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 319–347. Bloomington: Indiana University Press. Tumarkin-Deratzian, A. R. 2003. Bone surface textures as ontogenetic indicators in extant and fossil archosaurs: Macroscopic and histological evaluations. Ph.D. diss., University of Pennsylvania, Philadelphia. Williamson, T. E. 1997. Late Cretaceous (Early Campanian) vertebrate fauna from Allison Butte Member, Menefee Formation, San Juan Basin, New Mexico. New Mexico Museum of Natural History and Science Bulletin 11: 51–59. Witmer, L. M. 1997. The evolution of the antorbital cavity of archosaurs: A study of soft tissue reconstruction in the fossil record with an analysis of the function of pneumaticity. Society of Vertebrate Paleontology Memoir 3: 1–73. Wolfe, D. G. 2000. New information on the skull of Zuniceratops christopheri, a neoceratopsian dinosaur from the Moreno Hill Formation, New Mexico. In S. G. Lucas and A. B. Heckert, eds., Dinosaurs of New Mexico, pp. 93–94. New Mexico Museum of Natural History and Science Bulletin 17. Wolfe, D. G., and J. I. Kirkland. 1998. Zuniceratops christopheri n. gen. & n. sp. A ceratopsian dinosaur from the Moreno Hill Formation (Cretaceous, Turonian) of west-central New Mexico. In S. G. Lucas, J. I. Kirkland, and J. W. Estep, eds., Lower to Middle Cretaceous Non-marine Cretaceous Faunas, pp. 303–318. New Mexico Museum of Natural History and Science Bulletin 14. Wolfe, D. G., J. I. Kirkland., D. Smith, K. Poole, B. ChinneryAlgeier, and A. McDonald. 2007. Zuniceratops christopheri: An update on the North American ceratopsid sister taxon, Zuni Basin, west-central New Mexico. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 159–167. Drumheller: Royal Tyrrell Museum of Palaeontology. You, H., and P. Dodson. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press. You, H., and Z. Dong. 2003. A new protoceratopsid (Dinosauria: Neoceratopsia) from the Late Cretaceous of Inner Mongolia, China. Acta Geologica Sinica (English Edition) 77: 299–303.

9 A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation (Late Campanian) of Southern Alberta, Canada M I C H A E L J . R YA N , D AV I D A . E B E R T H , D O N A L D B . B R I N K M A N , P H I L I P J . C U R R I E , A N D D A R R E N H . TA N K E

in 2001, an almost complete, but disarticulated, adult-

sp. is known from Alaska (Fiorillo et al. this volume). In regions

sized centrosaurine ceratopsid (TMP 2002.76.1) with a

with good fossil records (e.g., the Dinosaur Park Formation of

pachystotic nasal boss and deeply excavated bosses over

Alberta and the Two Medicine Formation of Montana), centro-

the orbits on each postorbital was collected from the

saurs appear to have undergone rapid stratigraphic replace-

transgressive Dinosaur Park Formation (DPF) at Dinosaur

ment on the order of approximately every 0.5 Ma (Ryan and

Provincial Park (DPP) near Iddesleigh, Alberta. Surficial

Evans 2005; Sampson and Loewen this volume). The last cen-

bone texture indicates that the specimen was an old, ma-

trosaurines to appear in the fossil record are the clade of

ture individual. Pathological deformations are present in

‘‘pachyrhinosaurs’’ (Pachyrhinosaurus + Achelousaurus) distin-

the anterior parietal bar and in many of the phalanges.

guished by having large, pachystotic, boss-like nasal ornamen-

Unfortunately, the typically diagnostic posterior parietal

tation and highly modified postorbital ornamentation, that,

was not recovered, hindering a definitive identification

while usually described as being a ‘‘boss,’’ is more typically a

of the specimen. A restricted phylogenetic analysis of

wide, thin-floored, irregularly excavated depression.

Centrosaurinae produced an unresolved trichotomy of

Pachyrhinosaurus canadensis was first described by C. M.

(TMP 2002.76.1 + Achelousaurus + Pachyrhinosaurus).

Sternberg (1950) based on material collected from the St.

TMP 2002.76.1 occurs near the top of the DPF in a

Mary’s River Formation (Late Cretaceous) of southern Alberta

stacked channel succession that cuts out the lowest coal

from badlands along the north side of the Little Bow River and

of the 15 m thick Lethbridge Coal Zone. Dates from ben-

from the limited exposures at Scabby Butte. Additional mate-

tonites that bracket this interval suggest an age of 75.1

rial was collected from the Scabby Butte bonebed in 1957 by

Ma, making TMP 2002.76.1 the first occurrence of a

Wann Langston, Jr. (1975), as well as from the Horseshoe Can-

pachyrhinosaur ceratopsid in the fossil record.

yon Formation along the Red Deer River in the Drumheller Valley. Starting in 1986 the Royal Tyrrell Museum of Palaeon-

Introduction

tology (then the Tyrrell Museum of Palaeontology) has collected Pachyrhinosaurus n. sp. (Currie et al. 2008) material

Centrosaurine ceratopsids are known primarily from the

from at least one monodominant bonebed in the Wapiti For-

northern Upper Cretaceous biogeographic zone, although

mation near Grande Prairie, Alberta. A second bonebed in the

they are now known to have been present as far south as Mex-

same region may produce a different species of Pachyrhino-

ico (Sampson and Loewen this volume), and Pachyrhinosaurus

saurus (Fanti and Currie 2007). Pachyrhinosaurus material has

141

Locality map. TMP 2002.76.1 occurs at the southeastern margin of Dinosaur Provincal Park, Alberta. Inset shows the location (star) relative to Alberta.

FIGURE 9.1.

also been collected from northern Alaska (Nelms and Clemens

maxilla; PNR: posterior raised ridge of nasal boss; POA: an-

1989; Fiorillo 2004; Fiorillo et al. this volume).

terior margin of postorbital ornamentation; POB: posterior

To date, the geologically oldest pachyrhinosaur has been

margin of postorbital ornamentation; POP: postorbital pro-

Achelousaurus from the Two Medicine Formation of Mon-

cess; PP: pathological pit?; PS: parietal-squamosal contact;

tana (approximately 74.0 Ma; Sampson 1995; Sampson and

QG: quadrate groove; QJC: quadratojugal; R1: anterior ridge

Loewen this volume). It differs from Pachyrhinosaurus is having

of postorbital ornamentation; R: rostral; RA: retroarticular

a nasal boss that does not contact the postorbitals, and having

process; RM: raised anterior margin of postorbital; S: squa-

a less excavated form of the postorbital boss. Achelousaurus

mosal; S/A: contact of articular and surangular; SA: surangu-

replaces Styracosaurus ovatus in that formation, and occurs ap-

lar; SCC: supracranial cavity; SP: fused contact of squamosal

proximately 1.0 Ma later than the last occurrence of Styraco-

and parietal; VA: ventral angle of premaxilla; VD: ventral de-

saurus albertensis (Ryan and Evans 2005; Sampson and Loewen

pression; VNR: raised ventral margin of nasal boss; VR: ventral

this volume) from the Dinosaur Park Formation of Alberta.

ridge below raised anterior margin of parietal.

We describe here a new pachyrhinosaur from the Dinosaur Park Formation of Dinosaur Park (Fig. 9.1), Alberta, which is now the geologically oldest pachyrhinosaur known, and discuss its implications for centrosaurine paleoecology.

Locality and Geological Setting In 1996, Philip Currie and colleagues from the Royal Tyrrell

Institutional Abbreviations. AMNH: American Museum of

Museum were shown a partially eroded but associated skele-

Natural History, New York; MOR: Museum of the Rockies,

ton (TMP 2002.76.1) of a ceratopsid at the eastern margin

Bozeman; TMP: Royal Tyrrell Museum of Palaeontology,

(‘‘Iddesleigh’’) of Dinosaur Provincial Park on the south side of

Drumheller.

the Red Deer River (Fig. 9.1) by local rancher Gene Johnson. In

Anatomical Abbreviations. A: angular; AF: antorbital fenestra;

2000, excavation of the quarry began and continued in the

AG: groove for angular; AR: ascending ramus; ART: articular;

summers of 2001–2003, 2005, and 2006. Almost the entire

D: dentary; EJ: epijugal; G: glenoid; J: jugal; JC: surface for

skeleton, with the important exception of most of the parietal,

jugal; L: lacrimal; M: maxilla; N: nasal; NB: nasal boss; O: orbit;

was collected, and the prepared and mounted material went

P: parietal; PC: pterygoid contact; PD: predentary; PM: pre-

on display at the Royal Tyrrell Museum in the summer of 2007.

142 ryan, eberth, brinkman, currie, & tanke

TMP 2002.76.1 was found 21 m below the Dinosaur Park– Bearpaw formational contact within, and near the base of, a 9 m thick, multistoried channel fill (Fig. 9.2) that cuts out the lowest coal of the 15 m thick Lethbridge Coal Zone. The specimen had been originally deposited on a point bar in a meandering channel that experienced seasonal-to-subseasonal variations in flow regime and depth, probably as a result of tidal influences and/or episodic rainfall. Although its stratigraphic position is generally regarded as occurring just below the Lethbridge Coal Zone—the base of which is marked by a prominent coal zone that occurs approximately 15 m below the top of the Dinosaur Park Formation (Eberth 2005)— continuous exposure in this area clearly shows that the base of the channel fill cuts down through the Lethbridge Coal Zone (Figs. 9.2, 9.3A). Thus, the channel fill and its fossils (including multiple specimens of Myledaphus teeth found with TMP 2002.76.1) are actually part of the Lethbridge Coal Zone. Across southern Alberta, the Lethbridge Coal Zone is regarded as a chronostratigraphic datum, with an age ranging between 75.3 and 74.8 Ma (Eberth 2005). Dates from bentonites that bracket the interval containing that specimen thus suggest an age of 75.1 Ma. TMP 2002.76.1 is associated with a coarsening upward heterolithic siliciclastic succession in the lowest one meter of the channel fill (Figs. 9.2, 9.3). The sediments surrounding the specimen comprise thinly interbedded grey siltstones and very fine grained sandstones, all of which are plant fragment rich and locally cemented with reddish-brown iron carbonate (Fig. 9.3B). The bedding is gently inclined ([10\) to the northwest, and can be traced eastward and updip into more steeply inclined and offlapping heterolithic sets that are approximately 30 m in horizontal extent with a vertical relief not exceeding 5 m. An extensive erosional surface overlies the host bed. Above this surface, decimeter-thick beds of lightcolored, fine-grained sandstone contain bedded conglomerates comprising rounded-to-angular pebble-to-cobble size intraclasts of ironstone, and mudstone (Fig. 9.3B). These conglomerate-rich sandstone beds are also gently inclined, but dip toward the northeast. Fragmentary fossil bone (turtle shell in particular) is common in all of these deposits. Following Thomas et al. (1987), Wood (1989), and Eberth (2005), meter-scale channel fills that comprise inclined heterolithic strata in the Dinosaur Park Formation are best interpreted as point bar deposits in medium-sized meandering rivers. The presence of sharp bedding surfaces across which there are changes in grain size and sedimentary structures inMeasured section through the upper 1⁄3 of the Dinosaur Park Formation at Iddesleigh. TMP 2002.76.1 occurs near the base of a 9 m thick channel fill that shows approximately 7 m of vertical relief. Specimen and channel fill are part of the Lethbridge Coal Zone. Note truncation of coals to the right (east).

FIGURE 9.2.

dicate that the channels experienced variations in flow rate, probably as a result of seasonal rainfall and regional storm activity (Béland and Russell 1978; Koster et al. 1987; Eberth and Getty 2005; Eberth et al. this volume). Based on the scale of the inclined heterolithic sets and their vertical relief (cf. Thomas et al. 1987; Wood 1989; Eberth 1996), we estimate

A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation 143

Geology of TMP 2002.76.1. (A) Photograph taken above the quarry looking southward. Note the deep incision of the channel fill (9 m thick). (B) Geology at the excavation site. The lower, darker beds (h) hosted the specimen. The channel fill truncates the coal succession to the left (east). Note the sharp contact of conglomeratic sandstone (ss) above the darker beds. Jacob’s staff is 1.5 m tall.

FIGURE 9.3.

that the original bankful-scale channel that hosted TMP

digits suggest that the specimen was a mature animal at the

2002.76.1 was 50 m wide and 5 m deep. The association of the

time of death.

specimen with a coarsening upward succession of inclined

Because the postcrania of ceratopsids tend to contain little

bedded mudstone and very fine sandstone suggests that the

in the way of diagnostic features, the description here will

skeleton either was buried in a relatively lower energy portion

focus solely on the cranium. A description of the postcrania

of a point bar (down stream portion), was buried during a time

will be presented elsewhere in a separate paper. Unless other-

of relatively low flow rate, or both (Wood 1985; Wood et al.

wise noted the description of the skull will be from the left

1988). The overlying conglomeratic sandstones are charac-

side.

teristic of bank collapse lag, reworked and buried on the lower portions of point bars during high energy flow (Eberth 1990:

SNOUT

fig. 12). The shift in inclined bed dip direction upward into the conglomerates, suggest that the onset of high energy flow was

The preserved snout includes the rostral, the complete left and

associated with a readjustment in the direction of flow.

part of the right premaxillae, the left maxilla, and the fused nasals.

Description

Rostral. The deeply rugose rostral (Figs. 9.4, 9.5) has the typical short, triangular centrosaurine shape, although its mar-

The disarticulated but associated skeleton was collected from

gins are partially obscured by its advanced fusion with the

2

an area approximately 4 — 3 m . In general, all elements of the

underlying premaxillae. A relatively short ventral process ex-

skeleton were coated in a several-millimeter-thick ironstone

tends posteriorly to just in front of the ventral angle of the

that was removed with air abrasive units. The skull (Figs. 9.4,

premaxilla. The anteroventral tip of the rostral has been bro-

9.5) was discovered lying on its left side and much of the lat-

ken away and may represent a pathology (Tanke and Roth-

eral right face (including the circumorbital elements and por-

schild this volume). The dorsal margin of the rostral cannot be

tions of the premaxilla, maxilla, nasal, and squamosal) and

determined but it is assumed to be below the expanded rostral

frill are missing and were not recovered. The skull has under-

shelf of the fused nasals.

gone moderate mediolateral crushing and has several deep

Premaxilla. The left premaxilla (Fig. 9.4) is complete al-

fractures running through the otherwise well preserved left

though sutural contact with the nasal, both dorsal to the naris

side. The size of the elements, the advanced fusion of the cra-

and along the posterior margin of the ascending posterior bar,

nial sutures, and the extreme pathologies on some of the

cannot be discerned due to fusion. This differs from some

144 ryan, eberth, brinkman, currie, & tanke

FIGURE 9.4.

TMP 2002.76.1. Skull in (A) left lateral view; (B) line diagram; and (C) dorsolateral view of left postorbital ornamentation. Lens parallax in (A) accounts for the slight difference between the photograph (A) and the line diagram (B). The left quadratojugal was rearticulated for the line diagram (B). Mandible was drawn from the rearticulated elements. Scale bar is 20 cm.

specimens of Achelousaurus (Sampson 1995), where the con-

tion that contacts the jugal. Between the two processes is an

tact between the nasal and the posterior bar of the premaxilla

elongate notch leading into the moderately sized antorbital

remains unfused even in the most massive skulls. The inflated

foramen (Fig. 9.4). As in all ceratopsids, the lateral surface be-

ventral margin of this bar forms a broad contact with the ante-

tween the two processes is slightly concave. In other cera-

rior portion of the maxilla, and its caudoventral portion forms

topsids the maxilla contacts the ectopterygoid, lacrimal, pal-

the dorsal margin of a large, elongate maxillary fossa. Ante-

atine, pterygoid and vomer, and meets the opposite maxilla

riorly the bar forms a small portion of the base of the ante-

where they slot into the premaxilla, but most of these contacts

riorly directed narial process. The thickened ventral angle

are either not preserved or not visible on this specimen.

projects slightly below the level of the ventral margin of the

Nasal. The skull of TMP 2002.76.1 (Figs. 9.4, 9.5) is domi-

maxilla. Anteriorly, the premaxillae are capped by the fused

nated by the large nasal boss. In mature centrosaurines the

midline rostral. The crescentic premaxillary nasal septum fills

paired nasals contact and fuse along the dorsalmedial margin,

the anterior two-thirds of the narial fossa and forms the ante-

developing into either a horn or boss-like ornamentation. The

rior margin of the nares. As in other ceratopsids, the surface of

unfused, ventral plate-like portions form much of the lateral

the septum is smooth, but does have several small pits ran-

face. The anteroventral margin of the lateral nasal forms the

domly distributed over the surface.

posterior margin of the naris and produces, at its anteroven-

Maxilla. The maxilla (Fig. 9.4) has the typical broad triangu-

tral corner, the diagnostic triangular centrosaurine narial pro-

lar profile with a medial ascending process that contacts the

cess (42 mm in length). The upper half of this process is con-

premaxilla anterodorsally, and an ascending caudodorsal por-

vex while the lower surface is flat and slightly imbricated

A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation 145

FIGURE 9.5.

TMP 2002.76.1. Skull in (A) dorsal; (B) right lateral; and (C) anterior views. Dashed line indicates the posterior margin of the preserved right side of face. Scale bar is 20 cm.

medially. The caudoventral margin of the nasal is fused to the

n. sp. (TMP 87.55.156; Currie et al. 2008: fig. 14B). The lateral

medial ascending process of the premaxilla so the extent of

surfaces are gently concave, and slightly inclined medially

their contact cannot be determined, but, typically in centro-

with rounded dorsal margins that wrap onto a flattened dorsal

saurines, the premaxilla excludes the nasal from contacting

surface. The latter bears a series of longitudinal grooves that

the maxilla. The anteroventral surfaces of the fused nasals are

start at its posterior margin and merge with disorganized

assumed to form the roof of the nares as in other centrosau-

rugosities on the anterior two-thirds of the dorsal surface. The

rines, and probably contributes a flange of bone to the dorsal

posterior margin of the boss forms a raised thickened ridge

portion of each premaxillary septum, but no suture can be

(Figs. 9.4, 9.5A, B) that sits directly above the antorbital fenes-

discerned.

tra. The anterior base of this ridge is ringed by at least six deep

Anterolateral projections of the nasals clasp the anterodor-

pits or fossae. In profile the boss has a slight midline hump

sal margins of the premaxillae as they do in all centrosaurines.

(similar to that seen on Albertaceratops; Ryan 2007) that occurs

Many centrosaurs have a small, boss-like swelling at this point

at a point perpendicular to the posteriormost embayment of

of contact (e.g., C. ‘‘nasicornus,’’ AMNH 5351). In Pachyrhino-

the nasal fossa. On either side of this high point the dorsal

saurus n. sp. this forms part of a rostral comb (sensu Currie

surface is slightly saddle-shaped. The anterior portion of the

et al. 2008) that can take the form of one or more horizon-

boss overhangs the rostrolateral extensions of the underlying

tal ridges extending from the anterior margin of the snout

nasal as it does in some Pachyrhinosaurus n. sp. skulls (e.g.,

between the nasal ornamentation and the rostral. On TMP

TMP 1986.55.206), and to a lesser degree in Achelousaurus.

2002.76.1 the swelling across the nasal-premaxillae contact

This anterior projecting portion has a concave anterior face so

forms a laterally expanded (approximately 75 mm at mid-

that the amount of overhang is 75 mm on either side, but only

point) and deep (approximately 60 mm) shelf-like structure

40 mm on the midline. TMP 2002.76.1 lacks the ‘‘spout’’-like

that tapers anteriorly to a rounded apex. Multiple sub-

anterodorsal projection seen on many specimens of the nasal

horizontal ridges that wrap across the surface of the pre-

bosses of Pachyrhinosaurus n. sp., as well as the prominent

maxillae contribute to this structure.

raised sagittal ribbon of bone that marks the point of fusion of

The boss-like nasal ornamentation resembles the general

the two nasals.

form of some of the well developed bosses seen on Pachyrhino-

The nasal boss has a distinct ventrolateral margin that forms

saurus n. sp. nasals from Grande Prairie (e.g., TMP 87.55.156

as a slightly raised ridge (VNR; Fig. 9.4). This ridge traces a shal-

and TMP 89.55.188). The bosses on this new taxon are highly

low concave arc from the anterior margin of the boss posteri-

polymorphic, ranging from inflated bosses (e.g., TMP

orly, where it merges with the thickened posterior ridge (PNR;

86.55.206 and TMP 87.55.285) to broadly dished out, thick-

Fig. 9.4). As in some specimens of Pachyrhinosaurus n. sp., TMP

walled cups. Viewed dorsally, the boss is approximately rec-

2002.76.1 has at least three shallow, anterodorsally oriented

tangular with rounded corners and closely resembles that of P.

flutes (‘‘palisades’’) on the lateral margins of the boss that origi-

146 ryan, eberth, brinkman, currie, & tanke

nate from the midpoint of the ventrolateral margin and end at

gin of the postorbital ornamentation. The peaked anterior

a point just dorsal to the base of the rostral overhung.

margin of the groove diminishes in height laterally as the

It is unclear if the prefrontal participates in the formation of

grooves shallow, but continues as a distinct, narrow ribbon of

the nasal boss. Currie et al. (2008) indicate that no specimens

bone that describes an inverted arc above the orbit, ending at

of Pachyrhinosaurus n. sp. show any indication of prefrontal or

approximately the midpoint of the lateral wall of the orna-

frontal involvement in the boss, a condition that mirrors that

mentation. The position of the anterior ridge suggests that, as

of Achelousaurus (Sampson 1995), and probably P. canadensis

on Achelousaurus, it is formed, in part, by the palpebrals and

(Currie et al. 2008). As in Achelousaurus and Pachyrhinosaurus n.

extends across at least a portion of the frontal and prefrontals.

sp., the nasal boss of TMP 2002.76.1 is separated from the post-

At the midline of the skull the posterior margin of the nasal

orbital ornamentation by a transverse saddle-shaped groove

ornamentation and the anterior margin of the groove are sep-

that is probably composed primarily of the fused prefrontals.

arated by 125 mm.

In Achelousaurus and Pachyrhinosaurus n. sp. this groove ex-

The margins of the palpebral cannot be discerned but, as in

tends posteriorly between the postorbital bosses to form a ‘‘T’’-

all centrosaurines, it probably forms the anterodorsal margin

shaped groove, and, although the midline between the post-

of the orbit. Although the palpebrals can be inferred to be large

orbitals is not completely preserved on TMP 2002.76.1, this

and block-shaped, the distinctive antorbital buttress (sensu

specimen appears to have had the same feature.

Sampson 1995) of Achelousaurus, Einiosaurus and Pachyrhinosaurus canadensis, is lacking in TMP 2002.76.1, as it is in

CIRCUMORBITAL REGION

Pachyrhinosaurus n. sp. Supracranial Cavity. As in other centrosaurs the postorbitals

As in most adult-sized centrosaurines, the circumorbital re-

are bordered medially, and fuse with, the frontals. The frontals

gion (Fig. 9.4) is composed of the fused unit of adjacent ele-

expand dorsoventrally ontogenetically to become vaulted,

ments (frontal, jugal, lacrimal, palpebral, postorbital and pre-

forming the walls of the supracranial cavity (Fig. 9.5B), except

frontal) forming the supraorbital (sensu Dodson et al. 2005).

for the posterior margin that is formed by the anteroventral

On putatively old centrosaurines, the postorbital can also fuse

face of the midline parietal bar. On all centrosaurs except Cen-

to the squamosal, as it has on TMP 2002.76.1.

trosaurus and Styracosaurus, the supracranial cavity at least par-

Lacrimal. Portions of the left lacrimal, and probably the pre-

tially excavates the base of the postorbital ornamentation. On

frontals can be seen internally through the broken margins of

TMP 2002.76.1, as on many skulls of Pachyrhinosaurus n. sp.,

the right lateral surface (Fig. 9.5B), but they do not seem to

the base of the preserved left postorbital boss is highly exca-

differ from other centrosaurines (in the case of the former), or

vated, and perforated by the three large foramina previously

cannot be clearly differentiated (for the latter). The antero-

described. The anteriormost opening has a smaller, secondary

ventral margin of the lacrimal forms the posterodorsal margin

cavity underlying it that is confluent with the supracranial

of the antorbital fenestra in Pachyrhinosaurus (Currie et al.

region. Anteriorly, the supracranial cavity appears to be ex-

2008) and is assumed to also do so in this specimen.

tended by a large sphere-shaped cavity under the putative left

Postorbital. The postorbital ornamentation (Fig. 9.4C) takes

prefrontal.

the form of a rugose ovoid depression (usually described as a

The dorsomedial margin of the skull adjacent to the left

boss in specimens of Achelousaurus and Pachyrhinosaurus), ap-

postorbital boss appears to preserve the left margin of the ‘‘U’’-

proximately 25 mm deep, that sits on the raised, dorsolater-

shaped frontal fontanelle (Fig. 9.5A). Since many mature cera-

ally facing surface of the postorbital. The raised margins of the

topsids variably have the frontal fontanelle at least partially

depression have a distinct sharp lip anteriorly and postero-

closed off, the seemingly large size of this opening may simply

laterally (POA and POB, respectively; Fig. 9.4). In the center of

be due to taphonomic deformation.

the depression are three (two large and one small) smooth walled fenestrae that perforate the thin floor of the depres-

TEMPORAL REGION

sion. These probably represent the advanced stages of resorption pit development, typical of all centrosaurs (and at least

The temporal region typically includes the epijugals, jugals,

some chasmosaurs; Sampson et al. 1997; Tanke and Roth-

quadrates, quadratojugals and squamosals, although the lat-

schild this volume). The largest of these are 20 mm — 31 mm

ter will be discussed with the frill. The left jugal is complete

(anterior) and 26 mm — 40 mm (posterior) wide, respectively.

and articulated with the skull. Parts of both quadrates and

The postorbital depression is bordered anteriorly by a shal-

quadratojugals are preserved with those from the left being

low, saddle-shaped groove with an anterior margin formed by

more complete.

a sharp-peaked ridge of bone (R1, Fig. 9.4). The saddle is ante-

Jugal. The jugal (Fig. 9.4) has the typical triangular shape of

roposteriorly longest and deepest medially, narrowing and di-

all ceratopsids, forming the ventral and posterior margins of

minishing in depth as it wraps around the anterolateral mar-

the orbit. It presumably has the standard ceratopsid contacts

A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation 147

FIGURE 9.6.

TMP 2002.76.1. Left quadrate in (A) posterior and (B) lateral views. Left quadratojugal in (C) lateral and (D) medial views. Scale bar is 10 cm.

with its adjacent elements but these are obscured by breakage

FRILL

or fusion. Viewed in profile the anterior margin is slightly convex while the ventral margin is expanded into a low pyra-

The centrosaurine frill is dominated by the parietal and the

midal thickening that bears a small epijugal fused to its sur-

relatively short, paired squamosals that make up its margins.

face. Posteriorly, the jugal forms the anterior and dorsal mar-

On TMP 2002.76.1, only the anteriormost portion of the mid-

gins of the infratemporal foramen, but is excluded from the

line parietal bar (Fig. 9.7A, B) and the lateralmost portion

posterior and ventral margin of the infratemporal fenestra by

of the left parietal ramus (fused to the squamosal; Fig. 9.4A, B)

the squamosal.

were recovered. The complete left squamosal is present in con-

Epijugal. The small, crescent-shaped epijugal (Fig. 9.4A, B) is

tact with the postorbital. A water-worn fragment of the right

53 mm in length and 27 mm deep at its midpoint. There is a

squamosal was also recovered that preserves a portion of

small ovoid depression on the posterior portion of its dorsal

the sutures for the quadrate and exoccipital. Other than

surface.

the ‘‘stepped-up’’ medial margin (mostly broken away on this

Quadrate. The preserved left quadrate (Fig. 9.6A, B) of TMP

specimen) that separates all centrosaurines from Avaceratops,

2002.76.1 is typical of most isolated ceratopsid quadrates, in

the squamosal does not appear to have any diagnostic utility.

that most of the pterygoid flange and proximal head for con-

Squamosal. The posterior portion of the squamosal (Fig.

tact with the squamosal is broken away. It has a preserved

9.4A, B) is slightly concave, as it is in other centrosaurs, but

length of 260 mm, and its proximal head is 90 mm wide. The

crushing of the skull may have reduced the curvature. The

lateral and medial condyles, measured anteroposteriorly, are

posterior margin has four loci for epoccipitals on its scalloped

43 mm and 33 mm thick, respectively. The triangular ven-

margin, but, unlike other centrosaurs of its putative age and

trolateral surface of the element is thickened and rugose for

size, these dermal ossifications had not fused to the margin.

contact with the quadratojugal. As in other centrosaurines,

The three lateralmost scallops abruptly taper within 20 mm

the quadrates are not diagnostic.

of the margin forming slightly dorsally concave, sharp-edged

Quadratojugal. The left quadratojugal (Fig. 9.6C, D) is 171

surfaces with low, elongate, rounded profiles. Each scalloped

mm high and has a preserved anteroposterior width of 82

loci is distinctly imbricated with their medial margins hav-

mm. In natural articulation the thin squamose anterior flange

ing a greater dorsal inflection than their lateral margins. The

rests tightly against the medial surface of the jugal, and only

fourth locus adjacent to the parietal-squamosal contact is an

the thickened posterior and ventral margins are visible be-

elongate (approximately 64 mm) raised oval scar to which an

yond these margins of the jugal (Fig. 9.4B). The long ascend-

epoccipital could have attached.

ing process inserts between the anteroventral process of the

The caudomedial corner of the squamosal is expanded rela-

squamosal and the quadrate. The deep suture for the quadrate

tive to the body of the element and bears several inscribed post-

is only partially preserved on both quadrates, as a small por-

eriorly directed grooves suggesting that in life an epoccipital

tion of the quadrate canal bounded by the quadratojugal.

straddled the contact between the squamosal and the parietal.

148 ryan, eberth, brinkman, currie, & tanke

TMP 2002.76.1. Anterior parietal bar in (A) dorsal and (B) ventral views. (C) Epoccipital. Scale bar for parietal bar is 10 cm. Scale bar for epoccipital is 2 cm.

FIGURE 9.7.

The squamosal forms the ventral margin of the infratem-

subconical epoccipital (Fig. 9.7C) was recovered from the quarry. In cross section the element is ovoid with rounded

poral fenestra. Parietal. The preserved portion of the anterior parietal bar

margins. One surface has two distinct ovoid pits separated

(Fig. 9.7) has a highly irregular shape that could represent a

by an irregular midline septum. It resembles some small

pathological condition or be due to plastic deformation of the

epoccipital-like excrescences associated with the midline spike

element after death, or both. The posterior portion of the wall

on Pachyrhinosaurus n. sp., but its placement on TMP 2002.76.1

of the frontal sinus is preserved but is highly distorted. This

is uncertain.

region is capped by the divergent and anterolaterally projecting processes that contact the posteromedial margins of

MANDIBLE

the postorbitals. These robust bars have a dorsoventrally depressed ovoid cross section. The right process is broken prox-

Most of the mandible, excluding the splenials, is preserved.

imally, but the left counterpart appears to be complete. Near

The left dentary is the best preserved with the posterior com-

the anteroventral margin of the parietal bar most centro-

plex of elements still in place. The right dentary has been

saurines variably develop a deep, wide fossa that is separated

highly eroded but the well preserved articulated angular/

from the distal portion of the ventral bar by a transverse lip of

prearticular/surangular unit and isolated articular are present

bone. In TMP 2002.76.1, this fossa is exceptionally deep, pos-

at a distance from the dentary. Predentary. The predentary (Fig. 9.8) has a maximum pre-

sibly as a result of its advanced age. Although the element appears to be abnormal, and infer-

served length and width of 255 mm and 175 mm, respectively.

ences about it should be made with caution, it should be

At the time of death the anterior peduncle of the right dentary

noted that posteriorly the preserved dorsal surface of the bar

was tightly fused to the predentary, but has since been broken

has a dorsally inflected, ‘‘U’’-shaped, rounded margin. Al-

through at this point. As in other centrosaurines the dorso-

though this margin could have been shaped by breakage and

lateral cutting surfaces are highly inclined and concave along

erosion, the ventral margin the bar has a ridge of bone that

their length.

follows the arch of the inflected dorsal surface and is separated

Dentary. The dentary (Fig. 9.9A) has morphology typical for

from it by a shallow sulcus. With the exception of Pachyrhino-

a large, mature centrosaur. All the teeth are missing with the

saurus n. sp., in all other centrosaurines the dorsal surface of

exception of four partially preserved replacement teeth in the

the anterior parietal bar is flat. On Pachyrhinosaurus n. sp. the

anteriormost alveoli of the right dentary, and several isolated

dorsal parietal bar can be slightly arched dorsally at the ante-

teeth recovered from the quarry (Fig. 9.9B and C). The tooth

rior margin of the fused midline spike(s). It is possible that the

row had at least 29 alveoli, which is not unusual for an old,

preserved posterior midline bar of TMP 2002.76.1 bar may

large centrosaur, as tooth rows are known to increase onto-

indicate the presence of similar epoccipitals.

genetically. The tooth row is well inset from the ramus and is

Epoccipital. Only one small (36 mm in length), compressed,

sharply inclined relative to the straight ventral margin of the

A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation 149

The glenoid on the articular takes up the anterior half of the dorsal surface and is separated from the distinct retroarticular process by its stepped-up margin (Fig. 9.9E). The glenoid surface (87 — 60 mm) has a modest sigmoid shape being convex medially and slightly concave laterally.

Phylogenetic Analysis The presence of the nasal and postorbital bosses on TMP. 2002.76.1 indicates that this centrosaur is probably more closely related to the ‘‘pachyrhinosaur’’ (Pachyrhinosaurus and Achelousaurus) clade diagnosed by Currie et al. (2008), rather than any of the other non-boss-bearing centrosaurs. Despite the lack of the ornamentation-bearing portions of the parietal, a limited phylogenetic analysis was conducted using 16 parsimony informative characters (Appendix 9.1) derived from Sampson (1995), Ryan (2007), and Currie et al. (2008), using all centrosaurine ceratopsid genera other than Avaceratops (see Ryan and Russell 2005 for a discussion of the problems associated with Avaceratops in phylogenetic analyses of Centrosaurinae). The data matrix (Table 9.1) was analyzed using the beta version of PAUP 4.02b10 (Swofford 2002). The Branch and Bound search algorithm produced three best trees, the consensus of which is presented in Fig. 9.10 (tree length = 39; Consistency Index = 0.87; Homoplasy Index = 0.13; Retention Index = 0.82; and Rescaled Consistency Index = 0.72). TMP 2002.76.1 could not be differentiated from Achelousaurus and Pachyrhinosaurus, forming a trichotomy with these two genera. TMP 2002.76.1. Predentary in (A) dorsal, (B) lateral, and (C) ventral views. Scale bar is 10 cm.

FIGURE 9.8.

Discussion Morphologically TMP 2002.76.1 shares a number of characters with Achelousaurus from the Two Medicine Formation of

ramus. The posterolateral surface below the broadest point of

Montana and both of the described Pachyrhinosaurus species

the dentary at the base of the coronoid process is broken away

from Alberta. Although TMP 2002.76.1 shares more similari-

and the broken margin well rounded. The anterior end of the

ties with the Pachyrhinosaurus n. sp. from Grande Prairie, Al-

left dentary, which is still in articulation with the predentary,

berta, than the much larger, more massive P. canadensis, it is

has broken off the central ramus of the bone. The total pre-

possible that many of these differences (e.g., the development

served length of the left dentary is 427 mm, with a depth

of the ‘‘transverse tunnel’’ formed by the overgrowth of the

through the coronoid of 225 mm.

postorbital exostosis on P. canadensis) may be size and/or age

Surangular/articular. Both fused left (Fig. 9.9A) and right (Fig.

related. The nasal boss of TMP 2002.76.1 is very similar in

9.9D, E) surangular/articular complexes are preserved, as are

size and shape to the known bosses of Achelousaurus (e.g.,

both angulars. Each element is similar to their described coun-

MOR 485 [holotype]) and those of some Pachyrhinosaurus n.

terparts for other centrosaurines, including multiple speci-

sp. (e.g., TMP 87.55.156), but, notably, the posterior margin

mens of Pachyrhinosaurus n. sp. that were available for com-

of the nasal boss of TMP 2002.76.1 is anterior to the ante-

parison with TMP 2002.76.1. The ascending process of the

rior orbital margin as it is in Achelousaurus, but not in Pachy-

surangular has at least four small foramina on its external sur-

rhinosaurus. The nasal bosses of Pachyrhinosaurus n. sp. from

face adjacent to the articulating surfaces for the dentary and

Grande Prairie Alberta are extremely variable in their overall

angular. As on other centrosaurs, the external surface has a

morphology, ranging from a flattened ‘‘mushroom’’-shape

modestly developed finger-like process that projects between

(e.g., TMP 86.55.206) to an elongate, dished out boss with a

the contacting margins of the dentary and angular.

anterodorsally oriented ‘‘sprout’’ rising from its anterior mar-

150 ryan, eberth, brinkman, currie, & tanke

FIGURE 9.9.

TMP 2002.76.1. (A) Left dentary in medial view; (B) and (C) isolated teeth; right surangular-articular complex in (D) lateral and (E) medial views. (F) Right angular in external view. Scale bars in A, D–F are 10 cm; bars in B and C are 2 cm.

to have a series of deep crenulations running across the

Table 9.1. Character Matrix Used for Phylogenetic Analysis

postorbital surface rather than having well developed pitting. However, as the postorbital ornamentation of centrosaurs un-

Taxon

Character

dergoes significant ontogenetic change in all species (Samp-

Outgroup

0 0 0 0 0

0 0 0 0 0

0 N 0 0 0

0

son et al. 1997), it is difficult to compare postorbital orna-

C. apertus

0 1 1 1 1

1 1 1 3 3

1 N 1 0 1

1

mentation between closely related taxa without having a

C. brinkmani

0 1 1 2 1

1 1 1 3 4

2 3 1 1 1

1

reasonable inference of the state of maturity for the specimens

Styracosaurus

0 1 1 1 1

1 1 1 2 2

3 1 1 0 1

1

under consideration. TMP 2002.76.1 was clearly of advanced

Achelousaurus

1 2 1 3 2

2 1 1 0 1

3 1 1 0 1

1

age when it died compared to the known specimens of Ache-

Einiosaurus

0 1 1 3 1

2 1 1 0 1

3 0 1 0 1

1

P. n. sp.

1 2 1 3 2

2 1 1 0 3

3 2 1 2 1

2

P. canadensis

2 2 1 N 2

? 1 1 0 3

3 2 1 1 1

2

Chasmosaurine

0 1 0 0 1

0 0 0 1 2

1 N 0 0 0

0

TMP 2002.76.1

1 2 1 ? 2

? 1 1 ? ?

? ? ? ? 1

1

Albertaceratops

1 3 1 ? 1

0 1 1 0 0

4 2 0 0 1

?

lousaurus so it is difficult to make definitive statements about the significance of the differences between their postorbital ornamentation. Similar to Pachyrhinosaurus n. sp. (Currie et al. 2008) but different to Achelousaurus, TMP 2002.76.1 does not have well developed postorbital buttresses. Unfortunately, without the usually diagnostic parietal ornamentation, TMP 2002.76.1

gin; however, it is unclear if these extremes represent tapho-

cannot be distinguished from Achelousaurus or Pachyrhino-

nomic deformation or possible sexual dimorphism. Relatively

saurus at this time.

few nasal bosses are known from Achelousaurus so the extent

Although it is not possible to make a definitive referral for TMP 2002.76.1, the specimen it is important because it repre-

of variation in this taxon is unknown. The dished-out postorbital bosses of TMP 2002.76.1, with

sents the oldest occurrence of a pachyrhinosaur-grade ceratop-

their extremely thin and perforated floors, closely resemble

sid in the fossil record, and currently is the stratigraphically

those of most mature Pachyrhinosaurus n. sp specimens, and

youngest ceratopsid in Alberta known prior to the transgres-

differ from the ornamentation of Achelousaurus which tends

sion of the Bearpaw Sea across Alberta in the late Campanian.

A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation 151

FIGURE 9.10.

Cladogram showing relationships of TMP 2002.76.1.

TMP 2002.76.1 is at least 0.5 Ma older than Achelousaurus

tana and Alberta also confirms a shift towards the geographi-

from the Two Medicine Formation (Sampson and Loewen this

cally broad establishment of pachyrhinosaurs in western

volume) and approximately 1.0 Ma older than the Pachyrhino-

North America at the end of the Late Campanian.

saurus n. sp. from Grande Prairie, Alberta. Given the relatively

The fact that all of the latest Campanian centrosaurines are

rapid turnover of ceratopsids in the Late Campanian in Al-

pachyrhinosaurs, and that all of these are associated with wet,

berta and Montana (Ryan and Evans 2005; Sampson and

coastal paleoenvironments may partially help to explain the

Loewen this volume) in relatively restricted geographic re-

stratigraphic succession of ceratopsids within southern Al-

gions, it is probable that TMP 2002.76.1 will eventually be

berta and northern Montana. Within the DPF in DPP, the

recognized as a new taxon.

pachyrhinosaur represented by TMP 2002.76.1 replaces Styra-

Stratigraphically, TMP 2002.76.1 sits near the base of a chan-

cosaurus stratigraphically, and is an additional component in a

nel fill that has cut down from the Lethbridge Coal Zone that

significant dinosaur faunal change in the uppermost DPF. This

occurs at the top of the Dinosaur Park Formation in southern

transition is also marked by the first occurrence of the chas-

Alberta. The Lethbridge Coal Zone preserves a wet, coastal

mosaurine Chasmosaurus irvinensis and the lambeosaurine

paleoenvironment associated with the transgressing Bearpaw

Lambeosaurus magnicristatus. This three-taxon assemblage fur-

Sea, suggesting that the pachyrhinosaur represented by TMP

ther supports the concept that there are discrete dinosaurian

2002.76.1 may have preferred this type of environment. Sig-

biostratigraphic zones in the DPF at DPP as first proposed by

nificantly, following the regression of the Bearpaw Sea in Al-

Ryan and Evans (2005) and, in that context, comprises a dis-

berta, the only centrosaurine present in the Late Campanian is

crete dinosaur biostratigraphic zone in the uppermost 20 m of

Pachyrhinosaurus. P. canadensis in the Horseshoe Canyon For-

the DPF.

mation is associated with a swampy, coastal paleoenviron-

In their original scheme, Ryan and Evans proposed the pres-

ment, and P. n. sp. found to the north in the Wapiti Formation

ence of two dinosaur biostratigraphic zones, each consisting of

(Currie et al. 2008) is time equivalent to the late Bearpaw For-

a unique centrosaurine and chasmosaurine, and a unique had-

mation. To the south in Montana, Einiosaurus (with horn-like

rosaurid. They identified the lowest as the Centrosaurus apertus-

nasal ornamentation) was replaced stratigraphically in the

Chasmosaurus russelli-Corthyosaurus zone (0–30 m), and the

Two Medicine Formation by Achelousaurus during the time

overlying as the Styracosaurus albertensis-Chasmosaurus belli-

interval represented by the marine Bearpaw Formation in Al-

Lambeosaurus lambei zone (30–50 m). The documented pres-

berta, and the stratigraphic span between TMP 2002.76.1 and

ence of TMP 2002.76.1—a pachyrhinosaur centrosaurine—in

the first occurrence of P. canadensis (Sampson and Loewen this

combination with L. magnicristatus and C. irvinensis confirms

volume). This pattern of centrosaurine replacement in Mon-

the presence of the third dinosaur biozone (‘‘pachyrhino-

152 ryan, eberth, brinkman, currie, & tanke

saur’’-Chasosaurus irvinensis-Lambeosaurus magnicristatus) as

3. Narial spine of nasal, a pronounced process

proposed by Ryan and Evans (2005) in the uppermost 20 m of

projecting anteriorly into the nasal vestibule from

the formation, an interval that is essentially equivalent to the

the posterior narial margin: (0) absent

Lethbridge Coal Zone.

(1) present Acknowledgments

4. Form of postorbital ornamentation (subadult) (after TMP 2002.76.1 was discovered by Gene Johnson, who graciously granted access across his land to recover the specimen. The DPP staff facilitated our work in the park. TMP 2002.76.1 was collected as part of the Royal Tyrrell Museum of Palaeontology’s Field Experience Program. We thank the many participants and students who assisted the museum staff over the years of collection under that program and later (2006), when we searched in vain for the rest of the parietal. We would especially like to thank Brad Belluk, Carolyn Van Mackelberg, and Paul McNeil for their hard work in the quarry, and Kevin Kruger and Geoff Harding for their outstanding work in extracting the specimen. Donna Sloan produced the line drawing for Fig. 9.4B and prepared the images in Fig. 9.9B,

Sampson 1995; Holmes et al. 2001; Ryan 2007): (0) height at least three times as long as anteroposterior basal length; apex is dorsally or anterodorsally oriented (1) Pyramidal horncore with an approximate 1:1 ratio of height to antero-posterior basal length; apex is dorsally oriented (2) Elongate pyramidal horncore (attenuated) with a greater than 1.5:1 ratio of height to antero-posterior length; apex is anterolaterally oriented (3) Horncore longer antero-posteriorly than high, with rounded apex; apex is oriented dorsally

C. Donna Sloan and Mark Mitchell assisted with photography. Insightful reviews by Don Henderson and Rob Holmes

5. Supraorbital ornamentation type (unmodified adult) (Sampson 1995):

improved the paper.

(0) absent Appendix 9.1. Characters Used in the

(1) present, horn

Phylogenetic Analysis

(2) present, boss

1. Nasal ornamentation, basal length (adult) (this study): (0) short-based, less than 5% basal skull length (1) long-based, between 10% and 20% basal skull length (2) long-based, greater than 25% basal skull length All ceratopsids with nasal horns have short basal lengths relative to those of the bosses seen on Achelousaurus and Pachyrhinosaurus. The nasal boss of Pachyrhinosaurus is relatively longer than that of Achelousaurus, because it extends onto the frontals in this taxon (Sampson 1995).

6. Supraorbital horncore shape (unmodified adult) (after Sampson 1995): (0) elongate with pointed apex and round to oval base (1) pyramidal with rounded apex, at least as tall as base is long (2) rounded apex, base longer than horn is tall 7. Length of squamosal relative to parietal (Sereno 1986): (0) equal or subequal in length (1) squamosal less than 60% total parietal length 8. Shape of medial margin of squamosal (after Dodson

2. Nasal ornamentation type (adult) (Sampson 1995;

1986; Penkowski and Dodson 1999): (0) bowed (not ‘‘stepped-up’’)

Holmes et al. 2001):

(1) posterior portion ‘‘stepped-up’’ relative to

(0) absent or poorly developed

rostral portion (sensu Dodson 1986)

(1) elongated horn (2) boss

Centrosaurines (except Avaceratops) with relatively short

(3) elongate hump

squamosals have the posterior portion of these elements (pos-

All centrosaurines have juvenile and subadult-sized nasals

terior to the ventral paraoccipital groove) ‘‘stepped-up’’ rela-

that bear a thin triangular or blade-like structure that develops

tive to the rostral margin. Chasmosaurines have relatively

into a horn or boss after the paired elements fuse during on-

long squamosals with bowed medial margins.

togeny. Albertaceratops has an elongate raised surface as the adult ornamentation. All adult chasmosaurines carry a short nasal horn, formed by a separate element (epinasal) in at least some taxa.

9. Process at locus 1 (after Sampson 1995; Ryan and Russell 2005; Ryan 2007): (0) absent

A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation 153

(1) unelaborated epoccipital on posterior margin (2) short procurving hook on dorsal margin; length of hook ? diameter of base (3) long procurving hook on dorsal margin; length of hook ⱕ twice diameter of base 10. Process at locus 2 (after Sampson 1995; Ryan and Russell 2005): (0) absent (1) unelaborated epoccipital on posterior margin (2) small medially curled hook; length of hook ? length of base (3) large medially curled hook; length of hook ? twice length of base (4) multipronged posteriorly directed process 11. Process at locus 3 (after Ryan and Russell 2005; Ryan 2007): (0) absent (1) small, unelaborated epoccipital on posterior margin (2) narrow-based hook or spike; length between 1 and 3 times basal diameter (3) narrow-based long spike; spike greater than 4 times basal diameter (4) wide-based hook; length ] width of base. 12. Orientation of spike-like epoccipital at locus 3 (after Ryan 2007): (0) posteriorly directed (1) caudolaterally directed (2) laterally or rostrolaterally directed (3) dorsolaterally directed 13. Profile of midline parietal ramus (after Sampson 1995): (0) straight (1) series of low rounded bumps 14. Extra ossifications on parietal (Ryan 2005): (0) absent (1) multiple short spikes on posterior rami (2) one on more spikes on midline ramus 15. Orientation of the triturating surface of predentary relative to the horizontal plane of the element (sensu Lehman 1990; Forster 1996): (0) nearly horizontal (1) steeply inclined laterally 16. Size of antorbital foramen: (0) large (1) small (2) reduced

154 ryan, eberth, brinkman, currie, & tanke

References Cited Béland P., and D. A. Russell. 1978. Paleoecology of Dinosaur Provincial Park (Cretaceous), Alberta, interpreted from the distribution of articulated vertebrate remains. Canadian Journal of Earth Sciences 15: 1012–1024. Currie, P. J., W. Langston, Jr., and D. H. Tanke. 2008. A new species of Pachyrhinosaurus (Dinosauria, Ceratopsidae) from the Upper Cretaceous of Alberta. In P. J. Currie, W. Langston Jr., and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. 1–108. Ottawa: NRC Research Press. Dodson, P. 1986. Avaceratops lammerci: A new ceratopsid from the Judith River Formation of Montana. Proceedings of the Academy of Natural Sciences 138: 305–317. Eberth, D. A. 1990. Stratigraphy and sedimentology of vertebrate microfossil sites in the uppermost Judith River Formation (Campanian), Dinosaur Provincial Park, Alberta, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology 78: 1–36. ———. 1996. Origin and significance of mud-filled incised valleys (Upper Cretaceous) in southern Alberta, Canada. Sedimentology 43(3): 459–477. ———. 2005. The geology. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 54–82. Bloomington: Indiana University Press. Eberth, D. A., D. B. Brinkman, and V. Barkas. 2010. A centrosaurine mega-bonebed from the Upper Cretaceous of southern Alberta: Implications for behavior and death events. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 495–508. Bloomington: Indiana University Press. Eberth, D. A., and M. A. Getty. 2005. Ceratopsian bonebeds: Occurrence, origins, and significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 501–536. Bloomington: Indiana University Press. Fanti, F., and P. J. Currie. 2007. A new Pachyrhinosaurus bonebed from the late Cretaceous Wapiti Formation. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 39–43. Drumheller: Royal Tyrrell Museum of Palaeontology. Fiorillo, A. R. 2004. The dinosaurs of Arctic Alaska. Scientific American 291(6): 84–91. Fiorillo, A. R., P. J. McCarthy, P. P. Flaig, E. Brandlen, D. W. Norton, P. Zippi, L. Jacobs, and R. A. Gangloff. 2010. Paleontology and paleoenvironmental interpretation of the Kikak-Tegoseak quarry (Prince Creek Formation: Late Cretaceous), northern Alaska: A multi-disciplinary study of a high-latitude ceratopsian dinosaur bonebed. In M. J. Ryan, B. J. ChinneryAllgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 456–477. Bloomington: Indiana University Press. Forster, C. A. 1996. Species resolution in Triceratops: Cladistic and morphometric approaches. Journal of Vertebrate Palaeontology 16: 259–270.

Holmes, R. B., C. A. Forster, M. J. Ryan, and K. M. Shepherd. 2001. A new species of Chasmosaurus (Dinosauria: Ceratopsia) from the Dinosaur Park Formation of Southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Koster, E. H., P. J. Currie, D. A. Eberth, D. B. Brinkman, P. A. Johnston, and D. R. Braman. 1987. Sedimentology and palaeontology of the Upper Cretaceous Judith River/Bearpaw Formations at Dinosaur Provincial Park, Alberta, Field Trip #10. Geological Association of Canada, Mineralogical Association of Canada, Joint Annual Meeting, Saskatoon, Saskatchewan, 130 p. Langston, W., Jr. 1975. The ceratopsian dinosaurs and associated lower vertebrates from the St. Mary River Formation (Maestrichtian) at Scabby Butte, Southern Alberta. Canadian Journal of Earth Sciences 12: 1576–1608. Lehman, T. M. 1990. The ceratopsian subfamily Chasmosaurinae: Sexual dimorphism and sytematics. In P. J. Currie and K. Carpenter, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 211–299. Cambridge: Cambridge University Press. Nelms, L. G., and W. A. Clemens. 1989. Pachyrhinosaurus from the Kogosuhruk Tongue, Prince Creek Formation (Late Cretaceous), north slope of Alaska. American Association for the Advancement of Science, 40th Arctic Science Conference, Fairbanks. Proceedings: 36. Penkowski, P., and P. Dodson. 1999. The morphology and systematics of Avaceratops, a primitive horned dinosaur from the Judith River Formation (Late Campanian) of Montana, with the description of a second skull. Journal of Vertebrate Paleontology 19: 692–711. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., and D. Evans. 2005. Review of the Ornithischia of Dinosaur Provincial Park. In P. J. Currie and E. Kopplehus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 313–348. Bloomington: Indiana University Press. Ryan, M. J., and A. P. Russell. 2005. A new centrosaurine ceratopsid from the Oldman Formation of Alberta and its implications for centrosaurine taxonomy and systematics. Canadian Journal of Earth Sciences 42: 1369–1387. Sampson, S. D. 1995. Two new horned dinosaurs from the upper Cretaceous Two Medicine Formation of Montana, with a phy-

logenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15: 743–760. Sampson, S. D., and M. A. Loewen. 2010. Unraveling a radiation: A review of the diversity, stratigraphic distribution, biogeography, and evolution of horned dinosaurs (Ornithischia: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 405–427. Bloomington: Indiana University Press. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Sereno, P. C. 1986. Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). National Geographic Research 2: 234–256. Sternberg, C. M. 1950. Pachyrhinosaurus canadensis, representing a new family of the Ceratopsia, from southern Alberta. National Museum of Canada, Bulletin 118: 109–120. Swofford, D. L. 2002. PAUP 4.0b10, beta version. Sinauer Associates, Sunderland, Massachusetts. Tanke, D. H., and B. M. Rothschild. 2010. Paleopathologies in Albertan Ceratopsids and their behavioral significance. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 355–384. Bloomington: Indiana University Press. Thomas, R. G., D. G. Smith, J. M. Wood, J. Visser, E. A. Calverley-Range, and E. H. Koster. 1987. Inclined heterolithic stratification—terminology, description, interpretation and significance. Sedimentary Geology 53: 123–179. Wood, J. M. 1985. Sedimentology of the Late Cretaceous Judith River Formation, ‘‘Cathedral’’ area, Dinosaur Provincial Park, Alberta. M.Sc. thesis. University of Calgary, Calgary. ———. 1989. Alluvial architecture of the Upper Cretaceous Judith River Formation, Dinosaur Provincial Park, Alberta, Canada. Bulletin of Canadian Petroleum Geology 37: 169–181. Wood, J. M., R. G. Thomas, and J. Visser. 1988. Fluvial processes and vertebrate taphonomy: The Upper Cretaceous Judith River Formation, south-central Dinosaur Provincial Park, Alberta, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology 66: 127–143.

A New Pachyrhinosaurus-Like Ceratopsid from the Upper Dinosaur Park Formation 155

10 New Material of ‘‘Styracosaurus’’ ovatus from the Two Medicine Formation of Montana A N D R E W T. M C D O N A L D A N D J O H N R . H O R N E R

a partial skull, MOR 492, is attributed to Styracosaurus

ceratops’’ includes the holotype partial juvenile skull, USNM

ovatus Gilmore 1930, hitherto known from only the

7951, and the partial remains of four other juvenile individ-

holotype, USNM 11869, a partial parietal. With the addi-

uals excavated from the same quarry (Gilmore 1917). ‘‘Brachy-

tion of MOR 492, the cranial ornamentation of S. ovatus

ceratops montanensis’’ was designated a nomen dubium by

now includes a tall, erect, long-based nasal horncore and

Sampson et al. (1997) due to the lack of diagnostic adult fea-

low, rounded supraorbital horncores. Newly recognized

tures, a designation followed herein. A partial subadult skull,

morphological features and a phylogenetic analysis indi-

USNM 14765, was also referred to this taxon by Gilmore

cate that S. ovatus is most closely related to Einiosaurus

(1939), but it too lacks diagnostic adult features and cannot be

procurvicornis and is not congeneric with S. albertensis. A

conclusively shown to pertain to the same species as the origi-

new generic name, Rubeosaurus, is created for ‘‘S.’’ ovatus,

nal ‘‘Brachyceratops.’’ Styracosaurus ovatus has been problem-

bringing to three the number of valid centrosaurine spe-

atic since its discovery by George F. Sternberg in 1928, and has

cies in the Two Medicine Formation and making Styraco-

variously been considered dubious (Dodson et al. 2004) or

saurus a monotypic taxon confined to the Dinosaur Park

valid (Ryan et al. 2007).

Formation of Alberta. Hypotheses regarding centrosaurine evolution are reviewed in light of this analysis.

MOR 492 (Figs. 10.1–10.4) was discovered in 1986 and comprises an associated partial adult skull, including a partial left premaxilla, nearly complete co-ossified nasals with horncore,

Introduction

a partial left postorbital, and most of the right half of the parietal lacking the median bar. Comparison of the parietal of

The Two Medicine Formation of Montana has yielded mate-

MOR 492 with other centrosaurines reveals that it is referable

rial of more than 20 species of dinosaur over the last century

to Styracosaurus ovatus due to the medially inclined P3 spikes,

(Trexler 2001; Weishampel et al. 2004), including four taxa of

a feature shared with only USNM 11869. With the addition of

centrosaurine ceratopsids: ‘‘Brachyceratops montanensis’’ (Gil-

the more complete MOR 492, this taxon may now be coded

more 1914; nomen dubium), Styracosaurus ovatus (Gilmore

with fewer missing data, allowing a more robust phylogenetic

1930), Einiosaurus procurvicornis (Sampson 1995), and Achelou-

analysis of Centrosaurinae.

saurus horneri (Sampson 1995). The latter two species are well

Institutional Abbreviations. AMNH: American Museum of

established upon two paucispecific bonebeds and three iso-

Natural History, New York; CMN: Canadian Museum of Na-

lated partial specimens, respectively. The material of ‘‘Brachy-

ture, Ottawa; MOR: Museum of the Rockies, Bozeman; ROM:

156

Royal Ontario Museum, Toronto; TMP: Royal Tyrrell Museum of Palaeontology, Drumheller; USNM: United States National Museum, Washington, DC. Anatomical Abbreviations. BNH: base of nasal horncore; BNO: base of nasal ornamentation; BP3: base of parietal process 3; BP4: base of parietal process 4; CS: cornual sinuses; LM: lateral margin; NR: narial rim; OM: oral margin; OR: orbit; P1– P7: parietal processes 1–7; RM: rugose rostral margin; SOH: supraorbital horncore; SPR: sutural surfaces for premaxillae; SRO: sutural surface for rostral; TP: tab-like process; VT: vascular traces.

Description of Rubeosaurus ovatus Specimen MOR 492 CRANIAL MATERIAL Premaxilla. Only the rostroventral portion of the left premaxilla is preserved (Fig. 10.1A); the nasal and nasomaxillary processes are missing. It is of the typical centrosaurine structure. Nasals. The preserved portion of the co-ossified left and right nasals includes most of the left lateral nasal and an elongate, fused horncore that is recurved at its apex, as well as the proximal portions of the broken left and right premaxillary processes (Fig. 10.1B, C). Although the nasal horncore is com-

Systematic Paleontology Ornithischia Seeley 1888 Ceratopsia Marsh 1890 Ceratopsidae Marsh 1888 Centrosaurinae Lambe 1915 Rubeosaurus ovatus (Gilmore 1930) gen. et comb. nov. Holotype. USNM 11869, a partial adult parietal consisting of most of the transverse bar together with three pairs of fused, spike-like processes at the P3, P4, and P5 positions.

pletely co-ossified, these processes are separated in ceratopsids by the paired nasal processes of the premaxillae that fit between them (Hatcher et al. 1907). The sutures between the processes of the premaxillae and those of the nasals are visible on the latter’s dorsolateral surfaces. The lateral surfaces of the premaxillary processes are inflated and rugose, with series of fine ridges oriented rostrodorsally to caudoventrally. The most prominent feature of the facial skeleton is the large nasal horncore (Fig. 10.2). It is missing its tip, but is 390 mm in length measured along the rostral margin. The nasal horncore

Referred Specimen. MOR 492, an adult partial skull including partial left premaxilla, co-ossified left and right nasals with horncore, partial left postorbital with horncore, and nearly complete right parietal with two spikes. Generic Etymology. Rubeus, Latin adjective for thornbush or bramble, plus saurus, Latinized Greek noun for lizard; ‘‘thornbush lizard’’ in reference to the array of spikes adorning the parietals. Generic Diagnosis. As for type and only species. Specific Diagnosis. Centrosaurine ceratopsid diagnosed by a single autapomorphy: caudomedially inclined elongate spikes at the P3 positions. Differential Diagnosis. Elongate, long-based nasal horncore as in Einiosaurus; nasal horncore erect and slightly recurved as in adult specimens of Centrosaurus and Styracosaurus; small postorbital horncore with rounded apex, as in unmodified adult specimens of Styracosaurus and Einiosaurus; straight P3 spikes as Einiosaurus; P4 process variably developed as elongate spike, as in all adult-sized Styracosaurus, or as short spike as in some specimens of Einiosaurus; P5 process variably developed as short spike as in some specimens of Styracosaurus, or as unmodified epoccipital as in all other centrosaurines. Horizon. Upper Two Medicine Formation, approximately 60 m below the contact with the Bearpaw Formation, late Campanian, between 75 and 74 Ma (Rogers et al. 1993). Locality. Landslide Butte Field Area, Glacier County, Montana. Specific locality data for MOR 492 are on file at the Museum of the Rockies; exact locality datum for USNM 11869 is unknown.

FIGURE 10.1. Rubeosaurus ovatus (MOR 492) partial left premaxilla (A) in lateral view. Premaxilary processes of the nasals in (B) right and (C) left lateral view. Scale bars are 10 cm.

New Material of ‘‘Styracosaurus’’ ovatus 157

FIGURE 10.2.

Rubeosaurus ovatus (MOR 492) co-ossified left and right nasals in (A) left lateral, (B) rostral, and (C) caudal views. Scale bar is 10 cm.

of MOR 492 is the longest of any centrosaurine, with the

150 mm in height (measured vertically from its base to a point

exception of the horncore of AMNH 5351, the holotype of

directly medial to its tip). Although it is only partially pre-

‘‘Monoclonius nasicornus’’ (= Centrosaurus apertus), which mea-

served, the shape of the horncore can be inferred using the

sures 460 mm (Brown 1917; Ryan and Russell 2005). The

broken edges at the base and along the rostral margin of the

long-based horncore covers an estimated two-thirds of the

preserved portion (Fig. 10.3A); this conformation indicates a

dorsal surface of the nasals, as in Einiosaurus (e.g., MOR

low horncore with a rounded apex, as on unmodified adult

373-8-20-6-14, MOR 456-8-9-6-1, and MOR 456-8-13-7-5) and

postorbitals of Styracosaurus (Ryan et al. 2007) and Einiosaurus

the boss-bearing Achelousaurus (e.g., MOR 485). The horncore

(Sampson 1995). The lateral surface of the horncore is flat,

of Rubeosaurus differs from that of Einiosaurus in being erect

whereas the medial surface is slightly concave. The lateral sur-

and recurved at the apex rather than procurved (Fig. 10.2A).

face of the postorbital above the orbit is invaded by numer-

The horncore appears to have been very well vascularized,

ous, variably sized foramina. The ventromedial surface of

with its left lateral surface pierced by numerous small fora-

the postorbital exhibits two shallow depressions interpreted

mina; in addition, four prominent vascular traces emerge

as cornual sinuses, similar to those of all other centrosaurines

from a point slightly more than halfway up the left lateral

(Fig. 10.3B). The dorsal surface of the postorbital medial to the

surface of the horncore. The horncore is laterally compressed

horncore is rugose (Fig. 10.3C).

with a narrow base (Fig. 10.2B, C). The left and right sides of

Parietal. MOR 492 includes a nearly complete right pari-

the horncore are fully co-ossified, indicating an adult individ-

etal bar missing only the portion medial to the base of the

ual (Sampson et al. 1997).

P3 process and the sutural surface for the right squamosal

Postorbital. The left postorbital (Fig. 10.3) is incomplete and

(Fig. 10.4A, B). The dorsal surface of the parietal is gently con-

includes fragments of the caudodorsal orbital rim and the pos-

vex, whereas the ventral surface is flat; both surfaces display

terior process that articulates with the squamosal and jugal.

pronounced vascular traces. An especially prominent trace on

The orbital fragment includes a small, incomplete horncore

the dorsal surface follows the curvature of the parietal and

158 mcdonald and horner

face of the parietal. The dorsal surface of the P4 spike is convex, whereas the ventral surface is flat. The isolated P3 process is an elongate spike with a preserved length of 327 mm (Fig. 10.4E). The P3 spike of MOR 492 is straight in dorsal and ventral views, as in USNM 11869. Although the preserved left P3 of USNM 11869 is slightly longer than the right (Ryan et al. 2007), it cannot be determined whether the left and right P3 spikes of MOR 492 were asymmetric in length or to what degree. As in the P4 process, the ventral surface of the P3 displays marked vascular traces; three such traces are present, all of which arise from a point near the base and course towards the tip (Fig. 10.4E). MOR 492 also includes an eroded isolated epoccipital (Fig. 10.4F). This element is highly rugose and tapers to a thin edge laterally. The medial margin bears a sutural surface with a tablike process. This epoccipital probably inserted into the lateral surface of the contact between the squamosal and parietal; the epoccipital at this position typically does not fuse to the lateral margin of the frill (D. H. Tanke pers. com. 2007).

Phylogenetic Analysis MATERIALS AND METHODS A phylogenetic analysis of Centrosaurinae was performed to elucidate the relationships of Rubeosaurus ovatus to other centrosaurines. This analysis was carried out using a branch-andbound search in PAUP 4.0 beta 10 (Swofford 2005). We included 27 unordered cranial characters; two outgroup taxa, Leptoceratops gracilis and Chasmosaurinae; and nine ingroup taxa: Albertaceratops nesmoi, Centrosaurus brinkmani, Centrosaurus apertus, Styracosaurus albertensis, Rubeosaurus ovatus, Rubeosaurus ovatus (MOR 492) partial left postorbital in (A) lateral, (B) medial, and (C) dorsal views. Dashed line indicates extrapolated shape of horncore. Scale bars are 10 cm. FIGURE 10.3.

Einiosaurus procurvicornis, Achelousaurus horneri, Pachyrhinosaurus canadensis, and Pachyrhinosaurus sp. nov. (Currie et al. 2008). The character statements used in this analysis differ from those of previous analyses in the use of separate neomorphic and transformational statements to describe the cranial or-

gives rise to branches that run towards the bases of the P3 and

namentation, after the method recently suggested by Sereno

P4 spikes (Fig. 10.4A).

(2007). For example, Sampson (1995, character 14) and Ryan

Parietal ornamentation consists of three unmodified epoc-

and Russell (2005, character 11) combine the presence or ab-

cipitals and a pair of enlarged spikes. The unmodified epoccip-

sence of the P1 process with the form it takes if present into

itals are simple tab-like objects and are interpreted here as the

a single character statement. Our analysis, in contrast, has

P5, P6, and P7 parietal processes (sensu Sampson 1995). The

a statement that pertains to the presence or absence of the

P4 process (Fig. 10.4C, D) is broken at its base, but does fit onto

P1, and a separate statement regarding its form (see Appen-

the parietal. It is a short, caudolaterally inclined spike with a

dix 10.1).

preserved length of 126 mm. The tip is eroded, but given the degree to which the spike tapers, less than a few millimeters of the apex are inferred to be missing. Two very deep vascular

RESULTS

traces on the spike’s dorsal surface course distally. These arise

The possible centrosaurine Avaceratops lammersi was included

from a single trace near the base of the spike, which corre-

in the first running of the matrix. This produced three most

sponds to one of the aforementioned traces on the dorsal sur-

parsimonious trees differing only in the placement of Ava-

New Material of ‘‘Styracosaurus’’ ovatus 159

FIGURE 10.4. Rubeosaurus ovatus (MOR 492) partial right parietal in (A) dorsal and (B) ventral views. Parietal process 4 in (C) dorsal and (D) ventral views. (E) Parietal process 3 in ventral view. (F) Epoccipital at squamosal-parietal contact in dorsal view. Scale bars are 10 cm.

ceratops; this taxon could appear as the most basal centro-

of Albertaceratops (96%), as in Ryan (2007). There is weaker

saurine, the sister taxon of Albertaceratops, or the sister taxon

support for the remaining clades, including (Centrosaurus +

to all centrosaurines to the exclusion of Albertaceratops. The

Styracosaurus), with a value of 58%, and (Rubeosaurus + Einio-

relationships of all other centrosaurines remained the same in

saurus), with a value of 53% (Fig. 10.5).

all three trees. Following the removal of Avaceratops, reanalysis resulted in a single most parsimonious tree of 46 steps (Fig. 10.5),

Discussion

with CI = 0.957 and RI = 0.923. Rubeosaurus and Einiosaurus

REFERRAL OF MOR 492 TO RUBEOSAURUS OVATUS

are united by a single unambiguous synapomorphy (231, straight P3 spikes), and one ambiguous synapomorphy under

MOR 492 has been discussed by various other authors (e.g.,

ACCTRAN optimization (25 , possession of a short P4 spike).

Sampson 1995; Ryan and Russell 2005), but none have as-

Synapomorphies that unite centrosaurine sub-clades are

signed it a specific affiliation. Our referral of MOR 492 to

given in Table 10.1. Bremer support values for Centrosau-

Rubeosaurus ovatus is based upon the morphology of the pari-

rinae and the clade of all centrosaurines to the exclusion of

etal and is supported by its stratigraphic position.

1

Albertaceratops were fairly high (4 and 3 steps, respectively),

The partial right parietal bar of MOR 492 shows the broken

whereas those of the other centrosaurine sub-clades were low

bases of two parietal processes. The two bases are partially

(Fig. 10.5). A bootstrap analysis with 10,000 replicates was

coalesced, as are the bases of the P3 and P4 spikes of USNM

performed to determine the support for the groupings found

11869 (Fig. 10.6A). The break in the more complete base corre-

by the phylogenetic analysis. Strong support (99%) was found

sponds with the break in the shorter of the two spikes of MOR

for the monophyly of Centrosaurinae, as in Ryan and Russell

492, allowing the reconstruction of this process, which is in-

(2005), and for the clade of all centrosaurines to the exclusion

terpreted as the right P4 (Fig. 10.6B, C). Medial to the base of

160 mcdonald and horner

right P3 position (Fig. 10.6B, C). In summary, we conclude that the shorter spike of MOR 492 pertains to the right P4 position (sensu Sampson 1995), the longer spike to the right P3 position, and that the elongate P3 spike was medially inclined, a morphology that MOR 492 shares with only USNM 11869 (Gilmore 1930; Ryan et al. 2007). MOR 492 was collected 60 m below the contact between the Two Medicine Formation and the overlying Bearpaw Formation (Horner et al. 1992). The precise stratigraphic position of the holotype of Rubeosaurus ovatus, USNM 11869, has previously been considered uncertain (e.g., Ryan et al. 2007). We examined Gilmore’s field notes and map from the 1928 expedition and compared this information with stratigraphic data from more recently excavated dinosaur localities in the Two Medicine Formation. The locality of USNM 11869 is at approximately the same level as a Hypacrosaurus bonebed excavated by the MOR from 1985 to 1986, 60 m below the Two Medicine/Bearpaw contact. Thus, USNM 11869 came from low in the section, at approximately the same level as MOR 492. Specimens assigned to the three valid Two Medicine centrosaurines demonstrate that these taxa do not overlap stratigraphically, with Rubeosaurus occurring at 60 m, Einiosaurus at Single most parsimonious tree produced by a phylogenetic analysis of 11 taxa and 27 characters conducted in PAUP 4.0 beta 10. Tree graphic prepared using TREEVIEW (Page 1996). Numbers adjacent to nodes are Bremer support values and bootstrap percentages, respectively. FIGURE 10.5.

45 m, and Achelousaurus at 20 m below the Two Medicine/ Bearpaw contact (Horner et al. 1992; Sampson 1995) (Fig. 10.7). The inferred similar stratigraphic placement of USNM 11869 and MOR 492 supports the referral of the latter to Rubeosaurus ovatus. A similar pattern of non-overlapping stratigraphic ranges also occurs among the centrosaurines of the upper Oldman Formation and the Dinosaur Park Formation,

the short spike is a saddle-shaped rim, with the floor of the

with the sequence Centrosaurus brinkmani, C. apertus, and

saddle directed caudally. The left P3 of USNM 11869 is still

Styracosaurus albertensis (Ryan and Russell 2005). We must

attached to the parietal, but a fracture at its base is expressed as

note that, even though the holotype of ‘‘Brachyceratops mon-

a medially inclined, saddle-shaped raised rim similar to that of

tanensis’’ was also found approximately 60 m below the con-

MOR 492. Therefore, the saddle-shaped rim of MOR 492 is

tact (Sampson 1995), this taxon is a nomen dubium based

interpreted as the base of a medially directed P3 spike. We

upon indeterminate juvenile material and should not be con-

assign the isolated elongate spike of MOR 492 to this position

sidered synonymous with R. ovatus.

based upon comparison with specimens of other centro-

A tantalizing specimen, MOR 449, which was also discov-

saurines, such as Styracosaurus albertensis (e.g., TMP 89.97.1,

ered 60 m below the contact, exhibits characteristics quite

TMP 99.55.5, and TMP 2005.12.58), Einiosaurus (e.g., MOR

different from those of R. ovatus (Fig. 10.8). MOR 449 consists

456-8-14-19-8), Achelousaurus (e.g., MOR 485), and the Pipe-

of the co-ossified nasals of an adult centrosaurine with the

stone Creek Pachyrhinosaurus (e.g., TMP 87.55.141, TMP

nasal ornamentation broken off immediately above its base.

87.55.1085, TMP 87.55.1144, TMP 87.55.1503, TMP 88.55.46,

The base of the nasal ornamentation is wider than the portion

and TMP 89.55.258), as well as USNM 11869. These specimens

of the nasals directly ventral to it, and the preserved anterior

demonstrate that right and left spikes of the same position are

margin of the ornamentation dorsal to the bifurcation of the

usually of similar lengths; for example, the right P3 of an indi-

premaxillary processes is rugose, quite different from the lat-

vidual is not significantly shorter than the left P3. Further-

erally compressed nasal horncore of MOR 492 that is smooth

more, even if some degree of asymmetry exists between left

along its anterior margin. Extremely wide nasal ornamenta-

and right spikes of the same position, the longer spike is never

tion is a synapomorphy of the clade (Achelousaurus + Pachy-

more than 1.5 times the length of the shorter one. It is there-

rhinosaurus) (see Table 10.1), allowing for the tentative sug-

fore unlikely that the long spike of MOR 492 pertains to the

gestion that the species represented by MOR 449 might be

left P4 position; instead, it is interpreted as belonging to the

a member of this group. The concurrence of MOR 449 and

New Material of ‘‘Styracosaurus’’ ovatus 161

FIGURE 10.6. Rubeosaurus ovatus. (A) Holotype USNM 11869 in dorsal view; (B) reconstructed parietal of MOR 492 in dorsal view with right parietal fragment mirrored; (C) reconstructed skull of MOR 492 in left lateral view. Scale bars are 20 cm (A, B) and 100 cm (C).

MOR 492 indicates that two (possibly three, if ‘‘Brachycera-

tion and curvature of a spike. In Rubeosaurus, the shaft of

tops’’ can be determined to be a distinct taxon) centrosaurine

the spike is straight and the entire spike, from its base to its

species might be present 60 m below the upper contact of the

tip, is inclined medially (Fig. 10.6A, B). The spikes of Einio-

Two Medicine Formation.

saurus are similarly straight with a slight lateral inclination (Fig. 10.9A, B). In contrast to Rubeosaurus and Einiosaurus, the

CHARACTERS THAT UNITE RUBEOSAURUS AND EINIOSAURUS

elongate P3 spikes of other centrosaurines curve along their shafts. The P3 spikes of the Wahweap new taxon (Kirkland and DeBlieux this volume), Styracosaurus, Achelousaurus, and

In our phylogenetic analysis, Rubeosaurus and Einiosaurus are

Pachyrhinosaurus all display curved lateral and medial margins

sister taxa based on one unambiguous synapomorphy (231)

(Fig. 10.9D–F).

and an additional ambiguous synapomorphy (251) under only ACCTRAN optimization (Table 10.1).

The ambiguous synapomorphy is the presence of a short spike at the P4 position. This character is polymorphic in

The unambiguous synapomorphy is the presence of straight

Rubeosaurus (long or short P4 spike; Fig. 10.6A, B) and Einio-

P3 spikes. It is crucial here to distinguish between the inclina-

saurus (short P4 spike or unmodified epoccipital; Fig. 10.9A–

162 mcdonald and horner

FIGURE 10.8. Indeterminate centrosaurine MOR 449 co-ossified left and right nasals in oblique rostrodorsal view. Scale bar is 10 cm.

Evolution of Centrosaurines from the Two Medicine Formation Horner et al. (1992) proposed that Styracosaurus, MOR 492, Stratigraphic chart showing distribution of centrosaurine taxa and specimens in the Two Medicine Formation. Numbers 10–70 are depths (in meters) below the contact between the Two Medicine and Bearpaw formations. Age datum of 74.1 Ma from Rogers et al. (1993). FIGURE 10.7.

Einiosaurus, Achelousaurus, and Pachyrhinosaurus comprise an anagenetic lineage. Sampson (1995) disagreed and instead regarded each taxon as representing a separate lineage. Furthermore, there is a specimen from the uppermost Dinosaur Park Formation of Alberta that has affinities with Achelousaurus and Pachyrhinosaurus (Ryan et al. this volume), at least as old if not older than Rubeosaurus (Rogers et al. 1993; Eberth 2005). This calls into question the Einiosaurus-Achelousaurus progression of Horner et al. (1992). However, the close relationship

C), and these taxa are coded as such in the phylogenetic analy-

between Rubeosaurus ovatus and Einiosaurus procurvicornis and

sis (see Table 10.2). Such variation is widespread among cen-

their lack of stratigraphic overlap mean that anagenesis be-

trosaurines. The P5 processes of Rubeosaurus (short spike or

tween these two species cannot be ruled out.

unmodified epoccipital; Fig. 10.6A, B) and Styracosaurus (long

It is important to note the observation of Gingerich (1985:

or short spike; Ryan et al. 2007) are also variable and are coded

35) that ‘‘cladogenesis reduces to anagenesis’’ when isolated

as polymorphic. The P2 process of Styracosaurus can be ex-

segments of an ancestral population spawn modified descen-

pressed as a small, tab-like process as in Rubeosaurus, Einio-

dents during allopatric speciation. Thus, the evolution of

saurus, and Achelousaurus, or as an enlarged, medially curled

middle-late Campanian centrosaurines might have entailed

hook as in Centrosaurus apertus and Pachyrhinosaurus (Ryan et

both anagenesis and cladogenesis as parts of the same overall

al. 2007). Finally, adult C. apertus exhibit a highly variable P3

process of speciation. In any case, it is clear that the middle-

process, ranging from a tab-like epoccipital (e.g., TMP 97.85.1,

late Campanian centrosaurine record is extremely diverse for

USNM 8897), to an inflated, rounded epoccipital larger than

so narrow a span of time (Sampson and Loewen this volume)

the P4 (e.g., AMNH 5351), to a short spike (e.g., CMN 348).

and that evolution must have been rapid (Horner et al. 1992).

New Material of ‘‘Styracosaurus’’ ovatus 163

FIGURE 10.9.

Centrosaurine parietals. Einiosaurus procurvicornis parietals in dorsal view. (A) Holotype MOR 456-8-9-6-1; (B) MOR 373-7-9-87 (note: break near the distal end of the spike inflects the tip ventrally, causing the spike to appear curved in photograph); (C) MOR 373-6-28-6-4. (D) Pachyrhinosaurus sp. nov. parietal TMP 87.55.141 in dorsal view. (E) Achelousaurus horneri holotype MOR 485 parietal in left dorsolateral view. (F) Styracosaurus albertensis parietal ROM 1436 in dorsal view (courtesy of M. J. Ryan).

Table 10.1. Synapomorphies Uniting Centrosaurine Clades in the Phylogenetic Analysis Unambiguous Clade

(both optimizations)

ACCTRAN

(Pachyrhinosaurus [ Albertaceratops)

(92, 101, 113, 191, 223)

(71)

(17 )

(201)

(Centrosaurus + Styracosaurus)

1

(C. apertus + C. brinkmani )

(9 , 18 )

(Pachyrhinosaurus [ Centrosaurus)

(61, 73)

(Rubeosaurus + Einiosaurus) (Achelousaurus + Pachyrhinosaurus) (P. canadensis + P. sp. nov.)

164 mcdonald and horner

1

1

(231)

(251)

(2 , 3 , 9 ) 2

1

(22 ) 2

3

(41, 201)

DELTRAN

(71)

FIGURE 10.10.

Life restoration of Rubeosaurus ovatus based upon USNM 11869 and MOR 492. Artwork by Lukas Panzarin.

Rapid evolution in ceratopsids fits with the interpretation of

Conclusion

their bizarre cranial ornamentations as secondary sexual characters (Sampson et al. 1997; Sampson 2001); sexual selec-

The referral of MOR 492 to ‘‘Styracosaurus’’ ovatus reveals a

tion can lead to rapid speciation (Lande 1981; Sampson 1999;

trove of new anatomical data for this species, which previously

Coyne and Orr 2004). Horner (1984) suggested for the Two

included only a partial adult parietal, USNM 11869. The phy-

Medicine centrosaurines that their rapid evolution might

logenetic analysis indicates that ‘‘S.’’ ovatus is not congeneric

be attributable to the westward transgression of the Bearpaw

with the type species of Styracosaurus, S. albertensis, but is rather

Sea; as the sea transgressed, terrestrial habitats and species

most closely related to Einiosaurus procurvicornis. Furthermore,

shifted and were displaced. If the Rubeosaurus-Einiosaurus-

features of the cranial ornamentation of MOR 492 enhance the

Achelousaurus sequence is not due to a strictly anagenetic pro-

diagnosis of ‘‘S.’’ ovatus and reveal key differences between this

gression, then the non-overlapping stratigraphic occurrences

species and S. albertensis. Therefore, ‘‘S.’’ ovatus becomes the

might reflect a series of displacements and extinctions as the

type species of a new genus, Rubeosaurus. The addition of MOR

transgression forced different species into the Two Medicine

492 allows the skull of Rubeosaurus ovatus to be reconstructed

area.

and a life restoration to be created (Fig. 10.10).

New Material of ‘‘Styracosaurus’’ ovatus 165

Acknowledgments

We are indebted to C. Ancell for discovering MOR 492, and to

Table 10.2. Character-Taxon Matrix Used in the Analysis of Centrosaurinae

the people of the MOR preparation lab for their exquisite pres-

10

20

27

ervation of the specimen. ATM thanks the staff of the USNM, L. gracilis

0? ? ?0? ?0? ?

?000000?0?

0? ?0?0?

Chasmosaurinae

1000000100

0000110?0?

10?1010

Al. nesmoi

1100120100

1111110?0?

11?1010

C. apertus

1000101111

3111111111

14?1010

the Academy of Natural Sciences (ANSP), and to J. Gardner for

C. brinkmani

1000102111

2111111112

12?1010

access to specimens at the TMP. ATM thanks R. Sullivan for

S. albertensis

1000101121

311111101 ( 01 )

130121 ( 12 )

logistical support during visits to the ANSP and USNM, and

R. ovatus

1000? ? ?121

3? ? ?110?10

1311 ( 12 ) 1 ( 01 )

K. Bader and A. Farke during the Ceratopsian Symposium

E. procurvicornis

1000113121

3111110?10

1311 ( 01 ) 10

at the TMP. ATM is grateful to D. Brinkman, P. Dodson,

Ach. horneri

1210113131

?111110?10

1301010

D. Eberth, A. Farke, W. Kimbel, J. Kirkland, G. Orti, L. Panza-

P. canadensis

12111? ?131

?111110?11

1301010

rin, R. Rogers, M. Ryan, S. Sampson, P. Sereno, D. Tanke, and

P. sp. nov.

1211113131

?111110?11

1301010

especially R. Purdy for access to specimens and M. BrettSurman for assistance with Gilmore’s notes and map. Thanks also to T. Daeschler for access to the holotype of Avaceratops at

D. Wolfe for their correspondence and discussion. We are grateful to L. Panzarin for his spectacular restoration of Rubeosaurus and for his permission to include his artwork in this paper. Thanks to P. Dodson for reading an early version of this manuscript and providing discussion and insights, and to M. Ryan and S. Sampson for their critiques of a later version that greatly improved this paper. Thanks to A. Farke for assistance with the phylogenetic analysis. ATM is very grateful to D. Loope for serving as faculty advisor at University of Nebraska. This research was supported by funds from the University of Nebraska UCARE Program and by a grant from the Museum of the Rockies Paleontology Program, both awarded to the senior author. Finally, we wish to express gratitude toward the conveners and organizers of the Ceratopsian Symposium and the editors of this volume for their dedication and help in the presentation of this and other projects. Appendix 10.1.

6. Nasal ornamentation (subadult): short-based horn (0); long-based horn (1); long-based low crest (2) (modified from Sampson 1995). 7. Supraorbital horncore shape (subadult): conical apex, height at least twice as long as rostrocaudal basal length (0); pyramidal, with approximately 1:1 ratio of height to rostrocaudal basal length (1); attenuated, pyramidal with approximately 1.5:1 ratio of height to rostrocaudal basal length (2); rounded apex, horncore longer rostrocaudally than high (3) (modified from Ryan and Russell 2005). 8. Supraorbital ornamentation: absent (0); present (1). 9. Supraorbital ornamentation type (unmodified adult): elongate horn with conical apex and round to oval base (0); pyramidal horn with rounded apex, at least as tall as base is long (1); horn with rounded

Description and polarity of characters used in phylogenetic

apex, shorter than base is long (2); pachyostotic boss

analysis

(3) (modified from Sampson 1995 and Ryan and Russell 2005).

1. Nasal ornamentation: absent (0); present (1). 2. Nasal ornamentation type (adult): elongate horn (0); long-based, low thickened ridge (1); boss (2) (modified from Ryan 2007). 3. Nasal ornamentation (adult), transverse width of base: less than width of portion of nasals directly ventral to ornamentation (0); as wide as or wider than portion of nasals directly ventral to ornamentation (1). 4. Nasal ornamentation, basal length (adult): less then 25% basal skull length (0); more than 25% basal skull length (1) (modified from Ryan and Russell 2005). 5. Narial spine of nasal protruding rostrally into narial fenestra from caudal narial margin: absent (0); present (1) (Sereno 1986).

166 mcdonald and horner

10. Height of unmodified adult supraorbital ornamentation: greater than 40% length of face (0); less than 40% length of face (1) (modified from Holmes et al. 2001). 11. Orientation of supraorbital horns: caudodorsal (0); rostrolateral (1); lateral (2); dorsal (3) (modified from Ryan 2007). 12. Length of squamosal relative to parietal: equal or subequal (0); much shorter (1) (Dodson et al. 2004). 13. Shape of medial margin of squamosal: bowed (0); caudal portion stepped up relative to rostral portion (1) (Penkalski and Dodson 1999). 14. Processes on midline parietal bar: absent (0); present (1) (modified from Sampson 1995). 15. Parietal fenestrae: absent (0); present (1) (modified from Holmes et al. 2001).

16. Epoccipitals on squamosal and parietal: absent (0); present (1) (Dodson et al. 2004). 17. P1 parietal process: absent (0); present (1) (modified from Dodson et al. 2004). 18. Process at locus 1, form: abbreviated procurving hook on dorsal margin (0); elongate procurving hook on dorsal margin (1) (modified from Sampson 1995). 19. P2 parietal process: absent (0); present (1) (modified from Dodson et al. 2004). 20. Process at locus 2, form: small tab-like process (0); enlarged, medially directed process (1); enlarged epoccipital with several associated accessory ossifications (2); (modified from Ryan and Russell 2005). 21. Process at locus 3: absent (0); present (1). 22. Process at locus 3, form: unmodified epoccipital (0); short, pachyostotic hook (1); short rostrolaterally directed spike or hook, can attain length between one and two times basal diameter (2); elongate spike, can attain length greater than twice the basal diameter (3); variable, can range from tab-like epoccipital, to inflated, rounded epoccipital larger than the P4, to short spike (4) (modified from Ryan 2007). 23. Elongate P3 spikes, form: lateral and medial margins curved (0); lateral and medial margins straight (1). 24. Process at locus 4: absent (0); present (1). 25. Process at locus 4, form: unmodified epoccipital (0); short spike (1); elongate spike (2). 26. Process at locus 5: absent (0); present (1). 27. Process at locus 5, form: unmodified epoccipital (0); short spike (1); elongate spike (2). References Cited Brown, B. 1917. A complete skeleton of the horned dinosaur Monoclonius, and description of a second skeleton showing skin impressions. Bulletin of the American Museum of Natural History 37: 281–306. Coyne, J. A., and H. A. Orr. 2004. Speciation. Sunderland: Sinauer Associates, Inc. Currie, P. J., W. Langston, Jr., and D. H. Tanke. 2008. A new species of Pachyrhinosaurus (Dinosauria, Ceratopsidae) from the Upper Cretaceous of Alberta. In P. J. Currie, W. Langston, Jr., and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. 1–108. Ottawa: NRC Research Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P.Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A. 2005. The geology. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 54–82. Bloomington: Indiana University Press.

Gilmore, C. W. 1914. A new ceratopsian dinosaur from the Upper Cretaceous of Montana, with a note on Hypacrosaurus. Smithsonian Miscellaneous Collection 63: 1–10. ———. 1917. Brachyceratops: A ceratopsian dinosaur from the Two Medicine Formation of Montana. U.S. Geological Survey Professional Papers 103: 1–45. ———. 1930. On dinosaurian reptiles from the Two Medicine of Montana. Proceedings of the U.S. National Museum 77: 1–10. ———. 1939. Ceratopsian dinosaurs from the Two Medicine Formation, Upper Cretaceous of Montana. Proceedings of the U.S. National Museum 87: 1–18. Gingerich, P. D. 1985. Species in the fossil record: Concepts, trends, and transitions. Paleobiology 11: 27–41. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49. Holmes, R. B., C. Forster, M. Ryan, and K. M. Shepherd. 2001. A new species of Chasmosaurus (Dinosauria: Ceratopsia) from the Dinosaur Park Formation of southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Horner, J. R. 1984. Three ecologically distinct vertebrate faunal communities from the Late Cretaceous Two Medicine Formation of Montana, with discussion of evolutionary pressures induced by interior seaway fluctuations. Montana Geological Society 1984 Field Conference, Northwestern Montana Guidebook, pp. 299–303. Horner, J. R., D. J. Varricchio, and M. B. Goodwin. 1992. Marine transgressions and the evolution of Cretaceous dinosaurs. Nature 358: 59–61. Kirkland, J. I., and D. D. DeBlieux. 2010. New basal centrosaurine ceratopsian skulls from the Wahweap Formation (Middle Campanian), Grand Staircase–Escalante National Monument, southern Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 117–140. Bloomington: Indiana University Press. Lambe, L. M. 1915. On Eoceratops canadensis, gen. nov., with remarks on other genera of Cretaceous horned dinosaurs. Bulletin of the National Museum of Canada, Geological Series 24: 1–49. Lande, R. 1981. Models of speciation by sexual selection on polygenic traits. Proceedings of the National Academy of Sciences of the USA 78: 3721–3725. Marsh, O. C. 1888. A new family of horned Dinosauria from the Cretaceous. American Journal of Science 36: 477–478. ———. 1890. Additional characters of the Ceratopsidae with notice of new Cretaceous dinosaurs. American Journal of Science 38: 418–426. Page, R. D. M. 1996. TREEVIEW: An application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences 12: 357–358. Penkalski, P., and P. Dodson. 1999. The morphology and systematics of Avaceratops, a primitive horned dinosaur from the Judith River Formation (late Campanian) of Montana, with the description of a second skull. Journal of Vertebrate Paleontology 19: 692–711. Rogers, R. R., C. C. Swisher III, and J. R. Horner. 1993. 40Ar/ 39Ar

New Material of ‘‘Styracosaurus’’ ovatus 167

age and correlation of the nonmarine Two Medicine Formation (Upper Cretaceous), northwestern Montana, U.S.A. Canadian Journal of Earth Sciences 30: 1066–1075. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., D. A. Eberth, D. B. Brinkman, P. J. Currie, and D. H. Tanke. 2010. A new Pachyrhinosaurus-like ceratopsid from the upper Dinosaur Park Formation (Late Campanian) of southern Alberta, Canada. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 141–155. Bloomington: Indiana University Press. Ryan, M. J., R. Holmes, and A. P. Russell. 2007. A revision of the late Campanian centrosaurine ceratopsid genus Styracosaurus from the Western Interior of North America. Journal of Vertebrate Paleontology 27: 944–962. Ryan, M. J., and A. P. Russell. 2005. A new centrosaurine ceratopsid from the Oldman Formation of Alberta and its implications for centrosaurine taxonomy and systematics. Canadian Journal of Earth Sciences 42: 1369–1387. Sampson, S. D. 1995. Two new horned dinosaurs from the Upper Cretaceous Two Medicine Formation of Montana; with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15: 743–760. ———. 1999. Sex and destiny: The role of mating signals in speciation and macroevolution. Historical Biology 13: 173–197. ———. 2001. Speculations on the socioecology of ceratopsid dinosaurs (Ornithischia: Neoceratopsia). In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 263–276. Bloomington: Indiana University Press.

168 mcdonald and horner

Sampson, S. D., and M. A. Loewen. 2010. Unraveling a radiation: A review of the diversity, stratigraphic distribution, biogeography, and evolution of horned dinosaurs (Ornithischia: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 405–427. Bloomington: Indiana University Press. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Seeley, H. G. 1888. On the classification of the fossil animals commonly named Dinosauria. Proceedings of the Royal Society of London 43: 165–171. Sereno, P. C. 1986. Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). National Geographic Research 2: 234–256. ———. 2007. Logical basis for morphological characters in phylogenetics. Cladistics 23: 565–587. Swofford, D. L. 2005. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4 beta 10. Sunderland: Sinauer Associates. Trexler, D. 2001. Two Medicine Formation, Montana: Geology and fauna. In D. H. Tanke and K.Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 298–309. Bloomington: Indiana University Press. Weishampel, D. B., P. M. Barrett, R. A. Coria, J. Le Loeuff, X. Xu, X. Zhao, A. Sahni, E. M. P. Gomani, and C. R. Noto. 2004. Dinosaur distribution. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 517–606. Berkeley: University of California Press.

11 A New Chasmosaurine (Ceratopsidae, Dinosauria) from the Upper Cretaceous Ojo Alamo Formation (Naashoibito Member), San Juan Basin, New Mexico R O B E R T M . S U L L I VA N A N D S P E N C E R G . L U C A S

Ojoceratops fowleri, a new genus and species of chas-

Gilmore (1916) subsequently reported ceratopsid remains

mosaurine ceratopsid from the lower Maastrichtian

from the same region. These included four specimens from

Naashoibito Member (Ojo Alamo Formation), San Juan

the Ojo Alamo Formation; one specimen consisting of frill

Basin, New Mexico, is diagnosed by an apomorphic

fragments, a second specimen consisting of frill fragments

squamosal. Ojoceratops fowleri is presently the only diag-

with ‘‘deep radiating vascular impressions,’’ a third consisting

nosable ceratopsid from the Ojo Alamo Formation and

of ‘‘similar fragments,’’ and the fourth consisting of isolated

antedates Eotriceratops xerinsularis, Triceratops horridus,

teeth. Lull (1933) later cited these (and two others) as un-

Diceratops hatcheri, and Torosaurus latus, and is a tem-

cataloged material in the USNM collections, but none of this

poral equivalent of ‘‘Torosaurus’’ utahensis.

uncataloged material is presently housed at the Smithsonian (M. Brett-Surman pers. com. 2006), and their disposition is

Introduction and Background

unknown. All of these specimens reportedly came from strata above the ‘‘lower conglomerate [which is at the base of the Ojo

Nearly 100 years ago, Barnum Brown (1910) briefly described

Alamo Formation],’’ with the possible exception of the third,

the first ceratopsid remains from the Upper Cretaceous strata

whose stratigraphic position is less certain. In a subsequent

of the San Juan Basin, New Mexico (Fig. 11.1). The material

paper, Gilmore (1919) noted the fragmentary remains (‘‘pos-

he noted included uncollected thin squamosal fragments

terior median or dermosupraoccipital’’) of a ceratopsid from

with ‘‘deep vascular grooves’’ and a small, smooth ‘‘supraorbi-

the Ojo Alamo Formation. Comparing it to Triceratops, Cera-

tal horncore’’ (AMNH 5798). While the material came from

tops, or Monoclonius, he interpreted it as ‘‘a fragmentary frill in

an area known to contain rocks of both the upper part of

which the openings are apparently smaller than in either of

the Kirtland Formation (‘‘upper shale’’ of traditional usage)

the genera mentioned above. In all probability it represents an

and the Ojo Alamo Formation (formerly Ojo Alamo Sand-

undescribed form’’ (Gilmore 1919: 65). Unfortunately, the

stone), the exact stratigraphic horizon remains uncertain. The

whereabouts of this specimen is also unknown. For over 50

light brown-tan preservation of the ‘‘supraorbital horncore’’

years, no additional ceratopsid material was collected from

(AMNH 5798) suggests that this specimen came from the

the Ojo Alamo Formation.

lower Kirtland or upper Fruitland Formation and not from the

In 1977, the Bureau of Land Management conducted a pale-

stratigraphically higher De-na-zin Member (upper Kirtland

ontological survey in advance of potential coal mining ac-

Formation) or the overlying Ojo Alamo Formation.

tivity in the San Juan Basin, New Mexico. A few fragmentary

169

FIGURE 11.1.

Regional map of the San Juan Basin, New Mexico, showing outcrops of the Ojo Alamo Formation and principal localities for Ojoceratops fowleri, gen. nov., sp. nov. (1) Head of Hunter Wash (SMP localities: 313b, 319b, 364b); (2) head of Willow Wash/Alamo Wash (NMMNH localities 4726 and 5841; SMP localities 410, 411, 415); (3) head of De-na-zin Wash (SMP locality 403); (4) head of Coal Creek (SMP loc. 375); and (5) north flank of Betonnie Tsosie Wash (SMP loc. 374).

ceratopsid teeth were collected that year, and mandibular and

many of them from just above the lower conglomerate. Ver-

rostral fragments were later recovered in 1983 and 1984, re-

tebrate taxa (below family level) from this unit are poorly

spectively. Lehman (1981) reviewed the history of collecting

known—most are represented by very incomplete remains.

vertebrate fossils in the Upper Cretaceous strata of the San

Only three dinosaur taxa are currently recognized from the

Juan Basin and added to the list of material from the Naashoi-

Ojo Alamo Formation (Naashoibito Member) at the species

bito Member, a unit previously considered part of the Kirtland

level: the titanosaurid Alamosaurus sanjuanensis (Gilmore

Formation but recently reassigned to the Ojo Alamo Forma-

1922), the nodosaurid Glyptodontopelta mimus (Ford 2000;

tion (see Sullivan and Lucas 2003, 2006; Sullivan et al. 2005b).

Burns in press) and the ceratopsid Ojoceratops fowleri. Other

It wasn’t until 1997 that a major effort to resample the dino-

dinosaur taxa (such as Kritosaurus and Parasaurolophus) pre-

saur faunas from the Ojo Alamo and underlying Kirtland for-

viously reported from the Naashoibito Member are actually

mations commenced, separately initiated by both the New

from the De-na-zin Member (Kirtland Formation), which is

Mexico Museum of Natural History and Science and the State

disconformably overlain by the Naashoibito Member. Hunt

Museum of Pennsylvania. Consequently, numerous cera-

and Lucas (1991) described a partial hadrosaurid skeleton

topsid and other specimens have been recovered from these

from the Naashoibito Member as being a lambeosaurine. Wil-

formations during the last 13 years.

liamson (2000) countered that it was a hadrosaurine, based on

In addition to acquiring new specimens, the stratigraphy

the narrow neck of the pubis. It is possible that the hadrosaur

and biostratigraphy of the terrestrial Upper Cretaceous rocks

represents a form similar to, if not identical with, Edmonto-

of the San Juan Basin continues to be refined. All specimens

saurus, but this has yet to be demonstrated.

herein come from the Naashoibito Member of the Ojo Alamo

The resurgence in sampling the vertebrate fauna from the

Formation. Exposures of this member stretch along the north-

Naashoibito Member has resulted not only in a better un-

ern border of the Bisti/De-na-zin Wilderness (head of Hunter

derstanding of the vertebrate biostratigraphy, but in a re-

Wash across to Willow and Alamo washes, down to De-na-zin

interpretation of the age of the unit. Fassett (2005, in press)

Wash [Barrel Springs area]) on the eastern side of the wilder-

and Fassett et al. (2000, 2002) have argued for a Paleocene

ness. Additional exposures crop out in Betonnie Tsosie Wash

age for the Naashoibito Member based largely on pollen anal-

(Lucas and Sullivan 2000). Locally, the Naashoibito Member is

ysis, an interpretation met with much criticism (Fastovsky

rather thin, ranging from 4.2 m to 18.4 m in thickness (see

and Sheehan 2005a, b; McKenna 2007; Sullivan et al. 2003,

Lucas and Sullivan 2000). Most of the vertebrate fossils come

2005b); while others have argued for a ‘‘Lancian’’ (late Maas-

from the lower part of the unit (a thickness of 10 m or less),

trichtian) age (Lehman 1981; Weil and Williamson 2000; Weil

170 sullivan & lucas

et al. 2004; and Williamson and Weil 2004). We have presented data elsewhere (Sullivan and Lucas 2003, 2006) that suggests an older ‘‘pre-Lancian’’ (early Maastrichian) age. This correlation is based on the occurrence of Alamosaurus, which is tied to a precise U-Pb date of 69.0 +/- 0.9 Ma in the Big Bend region, Texas (McDowell et al. 2004; Lehman et al. 2006), an age interpretation we maintain here. Institutional Abbreviations. AMNH: American Museum of Natural History, New York; NMMNH: New Mexico Museum of Natural History and Science, Albuquerque; SMP: State Museum of Pennsylvania, Harrisburg; USNM: National Museum of Natural History (Smithsonian), Washington, D.C. Anatomical Abbreviations. as: articulation surface; av: arcuate vessel; be: beveled edge; bif: (superior) border of interpremaxillary fossa; dlw: depression: lateral wing; EN: epinasal ossification; ESQ: episquamosal ossification; f: furrow; g: groove; gc: glenoid cavity; ipvp: inferior posterior ventral process; mb: medial bar; ns: narial strut; ocl: occusal surface; or: orbital rim; pas: parietal articulation surface; pmxf: premaxillary fossa; s: scapular spine; tb: transverse buttress; tde: truncated distal end.

Systematic Paleontology Ornithischia Seeley 1888 Ceratopsia Marsh 1890 Ceratopsidae Marsh 1888 Chasmosaurinae Lambe 1915 Ojoceratops gen. nov. Type Species. Ojoceratops fowleri gen. et sp. nov. Etymology. The prefix ‘‘Ojo’’ is derived from Ojo Alamo,

FIGURE 11.2. Ojoceratops fowleri, gen. nov., sp. nov. holotype. A nearly complete left squamosal (SMP VP-1865) in (A) lateral and (B) medial views; (D) line drawing of B. (C) Holotype of ‘‘Torosaurus’’ utahensis (USNM 15583; a right squamosal) in lateral view. ESQ 1 indicates the position of episquamosal in both species. Scale bar is 10 cm.

name of the formation from which it came; suffix ‘‘ceratops’’ is derived from the Greek cerato meaning ‘‘horn,’’ the common root name given to horned dinosaurs.

Localities of Other Material. Specific locality information for

Diagnosis. As for the type and only known species.

the specimens cited herein are on file with the New Mexico Museum of Natural History and Science (Albuquerque) and

Ojoceratops fowleri gen. et sp. nov.

the State Museum of Pennsylvania (Harrisburg).

Holotype. SMP VP-1865, complete left squamosal (Fig. 11.2A, B, D).

Horizon and Age. Naashoibito Member of Ojo Alamo Formation (sensu Sullivan et al. 2005b), early Maastrichtian of New

Etymology. The name of the type species honors Denver

Mexico.

Fowler, who discovered the holotype and who also discovered, and collected, a number of the specimens described herein.

cf. Ojoceratops fowleri. Material. NMMNH P-36200, fragmentary skull elements, in-

Diagnosis. A chasmosaurine distinguished from other chas-

cluding sections of the squamosals, parietal, upper portion of

mosaurines by its unique squamosal morphology: enlarged

the premaxilla (including the narial struts and premaxillary

base; lacks embayment of the otic notch; wide and arched,

fossa), and occipital condyle; NMMNH P-44477, partial pari-

forming an angle of 115\ at the position of epijugal 1; trun-

etal with attached epiparietal; SMP VP-1243, frill fragment

cated and squared off distally; medial border concave; lateral

and two unidentified fragments; SMP VP-1245 and SMP VP-

border convex forming an arc of 80\.

1246, numerous frill fragments; SMP VP-1248, isolated frill

Type Locality. SMP locality 410b, Willow Wash, San Juan Basin, New Mexico.

fragment; SMP VP-1575, nearly complete parietal median bar (with lateral ‘‘wings’’); SMP VP-1576, two frill fragments; SMP

A New Chasmosaurine from the Upper Cretaceous Ojo Alamo Formation 171

VP-1719, numerous skull fragments, including left jugal/

dial edge, terminating 280 mm from the distalmost point. The

orbital rim region and maxilla fragments; SMP VP-1828, nu-

quadrate articular groove is preserved and is distinguished by

merous skull elements, including nearly complete nasal and

an internal flange. The articulation for the exoccipital is not

isolated rib; SMP VP-1829, incomplete right squamosal, pos-

readily identifiable. The postfrontal border of the squamosal

terior portion of parietal median bar, rostral and unidentified

is slightly damaged and measures approximately 230 mm

elements; SMP VP-1849, frill fragments; SMP VP-1877, squa-

across. The thickness of the squamosal varies, reaching a max-

mosal fragments; SMP VP-2013, incomplete parietal median

imum thickness of 30 mm. This specimen may be from a sub-

bar; and SMP VP-2090, nearly complete predentary.

adult based on the overall thickness of the squamosal and

Remarks. The material designated cf. Ojoceratops fowleri is

relatively smooth surface texture.

based primarily on having similar frill (squamosal and parietal) morphology and surface texture.

Descriptions of Assigned Material

?Ojoceratops fowleri

The following material is assigned to cf. Ojoceratops fowleri, ex-

Material. SMP VP-1872, ?frontal or parietal fragment; SMP

cept where noted, based on comparison of the holotype with a

VP-1873, incomplete left dentary; SMP VP-1874, anterior part

partial squamosal (SMP VP-1829) and other elements in com-

of right nasal (juvenile); and SMP VP-1875, incomplete left

mon between the latter specimen and other individuals.

scapula and nearly complete right mandible. Remarks. The 7-Up Sandbar individual (SMP VP-1872, 1873, 1874, 1875) is questionably referred to O. fowleri based only on size as no part of the frill is associated with this material.

CRANIA Squamosal. SMP VP-1829 is the transverse buttress region of a right squamosal (Fig. 11.3A, B). This section readily conforms

Description of Ojoceratops fowleri

in size and shape to the corresponding part of the holotype squamosal but differs in that it is more massive. The dorsal

The holotype is erected based on a single diagnostic element, a

surface appears to be more vascularized, but essentially dis-

nearly complete left squamosal (SMP VP-1865, Fig. 11.2A, B,

plays the same external texture as the holotype.

D). The squamosal is characterized by a distinctive outline

Another specimen, NMMNH P-36200, consists, in part, of

where the lateral margin and anterior articulation surface

numerous squamosal fragments, some 320 mm thick. The

form an obtuse angle of 115\. The base of the squamosal is

fragments differ from the holotype squamosal (SMP VP-1865)

long and lacks the embayment formed by the otic notch that is typical of chasmosaurines. The distal end of the squamosal is truncated or squared off. The lateral border is convex, and the medial border is concave. Episquamosals (sensu Horner and Goodwin 2008) are incipient, low, elongate to spindleshaped and indistinguishably fused along the lateral and distal ends of the squamosal. Only a few episquamosals are discernable in the holotype: the ‘‘first,’’ what appears to be the fourth, and three additional episquamosals are fused along the distalmost part. Along the medial edge is a surface approximately 205 mm long where the squamosal articulates with the parietal. This surface may, in fact, be longer (up to 390 mm), but due to poor preservation in this region the total length cannot be confirmed with certainty. The margin of the supratemporal fossa is approximately 310 mm long. Dorsally, external surface texture is relatively smooth with only minor traces of vascularization. Eroded areas are present on the dorsal surface of the squamosal, especially along the left lateral edge. The outer or dorsal surface of the squamosal is convex laterally, with the greatest convexity above the region of the exoccipital buttress. Ventrally the squamosal is concave, with the greatest concavity above the transverse buttress. This buttress extends across the breadth of the squamosal and distally along the me-

172 sullivan & lucas

FIGURE 11.3. Cf. Ojoceratops fowleri, gen. nov., sp. nov. (SMP VP-1829). Portion of squamosal in (A) anterior and (B) posterior view (showing transverse buttress). Scale bar is 10 cm.

in having a rugose surface texture with prominent vascular grooves, so they are inferred to be from a more mature individual. Episquamosals are also rugose, elongate to spindle-shaped and are fused to the lateral margins of the squamosals. Parietal (Median Bar and Adjacent Regions). Specimens of the parietal consist of fragmentary remains of the median parietal bar and anterior portions of the parietal in the region of the supratemporal fossae. The most distinctive and complete specimen, SMP VP-1575 (Fig. 11.4A, B), consists of the posterior part of the median bar, measuring approximately 450 mm (maximum length). This element is heavily fractured on both sides. The median bar is unique in that it expands laterally to a maximum of 90 mm from the right lateral edge of the bar. This expansion forms an elongate, oval depression, 245 mm long on the right side; only a portion of the corresponding depression is preserved on the left lateral side of the median bar. The left side is damaged, but the right edge is relatively intact and along most of it what appears to be an articulation surface is preserved. This ‘‘articulation surface’’ is similar to the ‘‘pseudosutural divisions’’ described by Gilmore (1946) for the parietal of ‘‘Torosaurus’’ utahensis (see below). It is unclear what this edge articulated with, presumably another part of the parietal as it is too short to be a parietal-squamosal contact. Distally the median bar widens, forming a continuous surface. The dorsal surface is characterized by a vascular surface texture, and the ventral surface also shows this texture, confirming that this section represents the distalmost part of the median parietal bar. The ventral surface (on the posterior part of the median bar) is slightly convex along the midline. The distal and posterior portion of the right lateral edge is eroded or missing. Maximum thickness of the median bar is 65 mm. Other comparable specimens include a partial median bar, SMP VP-1829 (Fig. 11.4C, D), and an isolated partial median bar, SMP VP-2013 (Fig. 11.4E, F). The former specimen also includes a

Cf. Ojoceratops fowleri, gen. nov., sp. nov. Incomplete parietal median bars showing laterally expanded wings (platforms). SMP VP-1575 in (A) anterior and (B) posterior views; SMP VP-1829 in (C) anterior and (D) posterior views; SMP VP-2013 in (E) anterior and (F) posterior views. Scale bars are 10 cm (C–F same scale). FIGURE 11.4.

number of other identifiable elements. The SMP VP-1829 parietal includes part of the left lateral expansion (the right side of the median bar is damaged). Breakage occurs on both sides of

mum thickness is approximately 30 mm. SMP VP-1829 (Fig.

the median bar of SMP VP-2013, suggesting that it, too, bore

11.5C, D) is very similar to SMP VP-1250, but is larger and

lateral extensions of the parietal. The dorsal sculpturing is con-

preserves the articulation margin for the attachment of the

sistent with that seen in other specimens, but the ventral sides

squamosal and medial parietal bar. This specimen also pre-

of both SMP VP-1829 and SMP VP-2013 are slightly concave

serves the curved border of either the parietal fenestra or the

along the parietal median bar.

infratemporal fenestra. It appears to be the complement (op-

Additional sections of the parietal are represented by speci-

posite) of SMP VP-1250. It has a maximum thickness of 22

mens SMP VP-1250, SMP VP-1829, SMP VP-1872 and SMP

mm and has the same sculpturing seen on the squamosal frag-

VP-1877. SMP VP-1250 (Fig. 11.5A, B) is a flat, plate-like sec-

ment from the same individual described above. The SMP

tion, sub-trapezoidal in shape, with prominent articular sur-

VP-1872 parietal is triangular in shape. In general it is thinner

faces on both lateral sides, and one edge which is beveled. The

(15 mm) than the previous fragments, but thickens towards

beveled edge may represent the posterior parietal margin, al-

the lateral sides (up to 30 mm). A prominent sutural surface is

though no indication of any epiparietals exists. The surface

pierced by a large vascular canal. Presumably this surface rep-

bears traces of five vascular grooves that are sub-parallel and

resents the midline (sagittal plane) of the skull, and the bone

curve, presumably, toward the midline. Another vascular

may pertain to the frontal, if not the parietal. SMP VP-1877

groove is located parallel to the midline near the edge. Maxi-

consists of four large parietal sections found in association,

A New Chasmosaurine from the Upper Cretaceous Ojo Alamo Formation 173

where the base of the epinasal attached is eroded. It appears that this element is from an immature individual, based on the fact that it is only slightly smaller than that of SMP VP-1828 and that it is not fused, as are the nasals of the latter specimen. It belongs to the 7-Up Sandbar individual which is questionably referred to O. fowleri (see below). Premaxilla. The superior portion of the premaxilla is among the more identifiable fragmentary elements of specimen NMMNH P-36200 (Fig. 11.6D, E). It consists of the region delimited by the arcuate vessel (above), the narial strut (posteriorly) and the superior part of the premaxillary fossa and superior border of the interpremaxillary fenestra (below), as illustrated by Forster (1996: fig. 1). The dorsal ascending portion of the premaxilla is broken where it joins with the nasal, as are the ventral and anterior portions that articulate with the rostral. The outer surface of this element bears distinctive rugose sculpturing laterally and anteriorly. The premaxillary fossa thins ventrally towards the midline and is broken anteCf. Ojoceratops fowleri, gen. nov., sp. nov. Incomplete parietal sections. SMP VP-1250 in (A) anterior and (B) posterior views; SMP VP-1829 in (C) anterior and (D) posterior views. Scale bars are 10 cm. FIGURE 11.5.

riorly. An apparent sutural surface for the rostral is present on the right side, but the left side is damaged in this region.

two of which clearly articulate with each other. The other two sections do not readily articulate. Maximum thickness of these is 220 mm. Lastly, NMMNH P-44477, a partial parietal with an attached epiparietal, recently described in detail by Farke and Williamson (2006), is included here. Its attributes are discussed below. Nasal. A relatively complete nasal bearing an incipient epinasal ossification is preserved in SMP VP-1828 (Fig. 11.6A–C). In SMP VP-1828 (Fig. 11.6A) the right side of the nasal is nearly complete, with an overall estimated length of about 510 mm. The dorsal surface of the nasal (Fig. 11.6C) is nearly flat, and the tip of the epinasal ossification rises only slightly. The superior margin of the nasal opening is damaged along the right side. The posterior third of the left nasal is also broken to the midline. The roof of the nasal opening exposes the smooth ventral surface of the paired nasals. The trace of the fused midline is evident. A circular rugose pit is present anteroventrally, where the anterior ascending portion of the premaxilla joins with the nasal (Fig. 11.6B). Below the epinasal ossification is a groove that we identify as the arcuate vessel, which appears more prominent on the right side. Ventrally, a midline sutural scar is formed by the fusion of the nasals and extends for the entire length of the specimen. Another specimen is identified as an incomplete right nasal SMP VP-1874 (Fig. 11.7). The surface sculpturing is highly vascularized, and the ventral ridge that marks the border of the naris is smooth and well-defined. The dorsal edge represents the midline of the face, where the nasals meet. The surface

174 sullivan & lucas

FIGURE 11.6. Cf. Ojoceratops fowleri, gen. nov., sp. nov. Incomplete nasal complex (SMP VP-1828) in (A) right lateral, (B) ventral, and (C) dorsal views; upper part of premaxilla (NMMNH P-36200) in (D) right lateral and (E) posterior views. Scale bar is 10 cm.

Rostral. The rostral is known from only one specimen (SMP VP-1829, Fig. 11.8A, B). It is approximately 200 mm long and has a maximum width of 70 mm. The rostral is relatively symmetrical bilaterally, is very rugose on both the anterior and posterior surfaces, and is covered with irregularly pits and grooves. The ascending process for articulation with the nasal is relatively blunt, whereas the ventral end is tapered. It is not clear if this element is complete or if it is missing part of the ventral region due to its rugose appearance. The lateral profile of the element is convex anteriorly and flat to slightly convex posteriorly. Lacrimal. A nearly complete, well-preserved right lacrimal is included with SMP VP-1719 (Fig. 11.8C). The orbital rim is partially intact and measures 165 mm from the orbit to the anterior end. The external surface is heavily grooved along the anterolateral side. The orbital rim is moderately pitted and the posterior margin of the rim is pierced by nine foramina. Dentary and Coronoid. A badly weathered incomplete left FIGURE 11.7. ?Ojoceratops fowleri, gen. nov., sp. nov. Incomplete right nasal of immature individual (SMP VP-1874) in (A) right lateral view (anterior direction to the right) and (B) medial view (anterior direction to the left). Dorsal direction is up for both A and B. Scale bar is 10 cm.

dentary (SMP VP-1873) and a nearly complete right dentary and coronoid (SMP VP-1875 [in part], Fig. 11.9) are typically ceratopsid, but are not diagnostic beyond family level. They, along with the rest of the 7-Up Sandbar specimen (SMP VP-1872 and 1874, see below) are assigned to ?Ojoceratops fowleri. Predentary. The predentary is massive and deep, but incomplete (SMP VP-2090; Fig. 11.10). The posterior portion is broken and both of the superior processes are incomplete, with the right side more complete than the left. The paired inferior ventral processes are present, but the proximal part of the right

FIGURE 11.8. Cf. Ojoceratops fowleri, gen. nov., sp. nov. Rostral (SMP VP-1829) in (A) anterior and (B) posterior views; (C) incomplete right lacrimal with orbital border (SMP VP-1719). Scale bar is 10 cm.

FIGURE 11.9. ?Ojoceratops fowleri, gen. nov., sp. nov. Nearly complete right dentary with coronoid (SMP VP-1875) in (A) lingual and (B) labial views. Scale bar is 10 cm.

A New Chasmosaurine from the Upper Cretaceous Ojo Alamo Formation 175

Member) of the San Juan Basin, New Mexico (see Fig. 11.1). These exposures are located at the northern and eastern ends of the Bisti/De-na-zin Wilderness and in a small region along the northern edge of Betonnie Tsosie Wash. Some specimens were collected from the same locality, but because they were either loosely associated, or were collected in subsequent years, they were assigned different catalog numbers. We consider these to be composite individuals. We recognize three composite individuals. The first two are referred to cf. O. fowleri: the Alamo Wash individual (SMP VP-1828, SMP VP-1829, SMP VP-090) and the Betonnie Tsosie Wash individual (SMP VP-1243, SMP VP-1245, SMP VP-1246). The third, the 7-Up Sandbar individual (SMP VP-1872, SMP VP-1873, SMP VP-1874, SMP VP-1875) is questionably referred. The remaining specimens are considered to represent single individuals, each assigned to cf. O. fowleri based on simiCf. Ojoceratops fowleri, gen. nov., sp. nov. Nearly complete predentary (SMP VP-2090) in (A) left lateral, (B) occlusal, and (C) ventral views. Scale bars are 10 cm. Apparent distortion due to depth of field. FIGURE 11.10.

lar frill surface texture. Comparison of same skull elements among the different single and composite specimens allows us to refer some nonsquamosal elements to cf. Ojoceratops fowleri. For example, the holotype squamosal (SMP VP-1865) can be compared to the

one is missing. The anterior tip of the predentary is also miss-

Alamo Wash specimen (SMP VP-1828, SMP VP-1829, and SMP

ing. The maximum width of the predentary is estimated to be

VP-2090) using the squamosal. In turn, the Alamo Wash speci-

230 mm. The maximum depth is approximately 180 mm. The

men can be compared to SMP VP-1575 using the median bar.

overall length cannot be determined with accuracy due to

However, reference of the remaining specimens to cf. O. fowleri

breakage, but probably approaches 320–340 mm (measured

is less certain due to fragmentary and nondiagnostic nature of

along the occlusal surface). The external surface is slightly

the material. Other elements are designated cf. O. fowleri based

pitted, with irregularly placed foramina located toward the

primarily on external texture (sculpturing), where applicable,

anterior end. The lower posterior margin of the predentary is

and size.

characterized by a deeply grooved dentary facet. This facet

The squamosal is the most distinctive element of Ojoceratops

articulates with the anterior tip of the two dentaries. The

fowleri. It differs from that of other chasmosaurines in hav-

walls of the facet on the left side are damaged, but those of the

ing a pronounced lateral angle at the position of the first epis-

right side are intact. The occlusal surface of the predentary is

quamosal, a concave lateral border, a convex medial bor-

largely undamaged on the left side, but is incomplete on the

der, and a truncated distal end. It lacks the distinctive otic

right side.

notch found in other chasmosaurines. It also has incipient,

POSTCRANIA Scapula. One left scapula (SMP VP-1875 [in part], Fig. 11.11) is nearly complete and measures 920 mm in length, but it is not diagnostic. It, along with the rest of the 7-Up Sandbar individual (SMP VP-1872, 1873, 1874), is referred to ?Ojoceratops fowleri. Rib. A single incomplete rib, belonging to SMP VP-1828, was recovered with other skeletal elements of the Alamo Wash individual (see below).

Discussion The material assigned to Ojoceratops fowleri, cf. O. fowleri, and ?O. fowleri comes from a number of specimens collected from limited exposures of the Ojo Alamo Formation (Naashoibito

176 sullivan & lucas

FIGURE 11.11. Cf. Ojoceratops fowleri, gen. nov., sp. nov. Nearly complete left scapula (SMP VP-1874) in (A) left lateral and (B) medial views. Scale bar is 10 cm.

elongated episquamosals, similar in morphology to the iso-

mens of Triceratops, but is distinguished by the absence of the

lated ‘‘epoccipitals’’ of ‘‘Torosaurus’’ utahensis and the recently

facet for contact with the superior (posteriorly projecting) tips

named Eotriceratops xerinsularis (Wu et al. 2007).

of the premaxillae. In SMP VP-1828 the epinasal forms a small

The parietal median bar (cf. O. fowleri ) is also distinctive,

ossification, but the variation in size of this epi-ossification

with ovoid depressions on the expanded lateral wings adja-

is ontogenetic (Horner and Goodwin 2006, 2008). Forster

cent to the axis of the median bar. It retains the outline of the

(1996) described the nasal and epinasal in a few specimens of

primitive, strut-like bar seen in Chasmosaurus and Pentacera-

Triceratops and noted the presence of an arcuate vessel groove

tops, and differs from the broad, rather thin median bars of

winding across the front of the nasal horn, and she suggested

Torosaurus latus and Arrhinoceratops brachyops (Parks 1925;

that it was an autapomorphy of Triceratops. However, this ves-

Tyson 1981). If properly referred, this morphology appears to

sel groove is present in SMP VP-1828 (Fig. 11.6A) and it seems

be unique to O. fowleri, as it has not been seen in any other

doubtful that the presence and position of vascular grooves

chasmosaurine.

are of any taxonomic value.

The remaining material is less taxonomically informative

Wu et al. (2007) concluded that Eotriceratops possesses a

due to its incomplete nature and/or ontogenetic stage. How-

nasal horncore positioned anteriorly, much as in Triceratops,

ever, a few fragments exhibit morphologies that are briefly

Torosaurus and Diceratops. The position of the nasal horncore

worth noting. Distal parietal fragments have been recovered,

may be the same for Ojoceratops, based on the position of the

including SMP VP-1250 (Fig. 11.5A, B) and NMMNH P-44477,

nasal horn in SMP VP-1828.

the latter recently described by Farke and Williamson (2006).

Is Ojoceratops the same genus as Triceratops? It seems clear,

SMP VP-1250 is, in part, distinguished by lateral articular sur-

based on the morphology of the squamosal, and irrespective

faces. These unusual surfaces were first noted by Parks (1925)

of ontogeny, that Ojoceratops is a taxon distinct from all other

for Arrhinoceratops and discussed later by Gilmore (1946) for

crown chasmosaurines. Assuming SMP VP-1575 is properly

‘‘Torosaurus’’ (Arrhinoceratops?) utahensis. Gilmore (1946) de-

referred, the presence of a median bar with expanded lateral

scribed them in detail and called them ‘‘pseudosutural divi-

wings is another criterion that may be used to unambiguously

sions,’’ or edges, and concluded that they were natural separa-

separate Ojoceratops from Triceratops, Eotriceratops, Torosaurus,

tions and not sutures of the parietal. Sullivan et al. (2005a)

‘‘Torosaurus’’ utahenesis and Diceratops. Recognition of a new

confirmed the occurrence of the surfaces in a referred speci-

ceratopsid dinosaur in the Naashoibito Member of the Ojo

men of ‘‘Torosaurus’’ (= Torosaurus?) utahensis (USNM 16572).

Alamo Formation compels us to reexamine a couple of the

The lateral sutures on SMP VP-1250 and NMMNH P-44477

previous taxonomic assignments of ceratopsid material from

may be correlated to a subadult ontogenetic stage, as the lat-

this unit.

ter specimen has been identified as subadult based on the incomplete fusion of the epiparietals. The median epiparietal of NMMNH P-44477, flanked by three others on each side, suggests that the complete parietal bore a total of seven epiparietals and thus could be distinguished from that of Triceratops,

TAXONOMIC STATUS OF PREVIOUSLY DESCRIBED CERATOPSID SPECIMENS FROM THE OJO ALAMO FORMATION

which is known to only have five (Farke and Williamson

Sullivan et al. (2005a) reassessed all the San Juan Basin speci-

2006). Also, because no midline epiparietal exists in Torosau-

mens considered to be referable to ‘‘Torosaurus’’ (= Torosaurus)

rus latus (Farke and Williamson 2006) the parietal cannot be

utahensis by Farke (2002). Of the six examined, only one

assigned to this taxon.

(NMMNH P-22884), in retrospect, approaches similar mor-

The three cf. O. fowleri parietal specimens, as well as the

phology to the holotype of Ojoceratops fowleri. This specimen

original paratype and referred material of ‘‘Torosaurus’’ (= Ar-

is an incomplete left squamosal characterized by a smooth

rhinoceratops?) utahensis (Gilmore 1946), all lack distinct cren-

dorsal surface with a long base and, in outline, it conforms

ulations (i.e., wavy morphology) and are devoid of articulated

somewhat to the holotype of Ojoceratops fowleri. It differs from

epiparietals. However, 13 isolated ‘‘epoccipitals’’ (epiparietals

the O. fowleri squamosal, however, in having an anteriorly

and episquamosals) were recovered from the Ojo Alamo For-

directed, somewhat prominent, episquamosal (= ‘‘epoccipi-

mation, and all are elongate and narrow (Gilmore 1946; Sul-

tal’’) number 1. To further complicate matters, the precise

livan et al. 2005a) suggesting a subadult stage (Horner and

stratigraphic position of this specimen is ambiguous. It ap-

Goodwin 2006, 2008). Episquamosals of the holotype of

pears to have come from high in the De-na-zin Member (Kirt-

Ojoceratops are similar to the elongate/spindle-shaped epi-

land Formation) or possibly low in the Naashoibito Member

ossifications of the holotype of ‘‘Torosaurus’’ utahensis (see Sul-

(Ojo Alamo Formation; Lehman 1981), two units temporally

livan et al. 2005a) and those recently described by Wu et al.

separated by at least 4 Ma (Sullivan and Lucas 2006).

(2007) for Eotriceratops.

This incomplete squamosal (previously numbered UNM

The nasal of cf. Ojoceratops is similar in shape to some speci-

B-628) was first documented and interpreted by Lehman

A New Chasmosaurine from the Upper Cretaceous Ojo Alamo Formation 177

(1981) as having affinities to ‘‘Torosaurus’’ utahensis. Lehman

been identified in other chasmosaurines including Eotricera-

(1981: fig 9.12) provided a description and line drawing of the

tops xerinsularis (Wu et al. 2007), so the presence of a medially

specimen without showing the distinct episquamosal 1, and

positioned epiparietal might be a synapomorphy of the crown

implied the presence of an otic notch along the anterolateral

clade.

margin of the squamosal. The episquamosal of this specimen differs in shape from the isolated elongated, spindle-shaped episquamosals found with the holotype and other materials

Conclusions

from the Ojo Alamo Formation, but as we have noted above,

Ojoceratops fowleri is a new genus and species of chasmosau-

the shape and degree of fusion of these elements change dur-

rine ceratopsid based on a diagnostic (left) squamosal, pos-

ing ontogeny (Horner and Goodwin 2006, 2008). By implica-

sibly supplemented by a number of incomplete specimens.

tion, shape and degree of fusion of episquamosals have no

The holotype squamosal is characterized, in part, by a very

bearing on taxonomic identity. We note, too, that Lehman

distinct lateral margin with an elongated base or ‘‘neck’’ and

indicated that there were no episquamosals on the lateral

squared-off tip. Some of the supplemental fragments exhibit

margin of NMMNH P-22884, a fact which we now question.

features that appear to be unique, but most of the morphology

Subsequent to the initial description, Lucas et al. (1987) iden-

of the material is not diagnostic below the family level. All of

tified this specimen as ‘‘Torosaurus’’ cf. ‘‘T.’’ utahensis. Later,

the specimens come from a very restricted stratigraphic inter-

Farke (2002) noted some differences that challenged this iden-

val, the Naashoibito Member of the Ojo Alamo Formation.

tification, such as the elongated base of the squamosal, and

This stratigraphic unit is considered to be Early Maastrichtian

concluded it could not be referred to ‘‘Torosaurus.’’ Sullivan et

(late Edmontonian) in age based on the co-occurrence of

al. (2005a) agreed with Farke (2002) and conservatively identi-

Alamosaurus sanjuanensis, a taxon that has been dated at 69

fied it as an indeterminate chasmosaurine. The presence of a

Ma in the Big Bend region of Texas. Ojoceratops fowleri is also a

lateral margin and elongated base, similar to the holotype

temporal equivalent to ‘‘Torosaurus’’ utahensis. Thus Ojocera-

of Ojoceratops fowleri (SMP VP-1865), argues for referral of

tops fowleri antedates the newly described Eotriceratops xer-

NMMNH P-22884 to this taxon and further supports the no-

insularis, Diceratops hatcheri, Torosaurus latus and the well-

tion that it is from the Naashoibito Member, as originally

known Triceratops horridus.

stated by Lehman (1981), rather than the De-na-zin Member. Thus we tentatively refer NMMNH P-22884 to cf. Ojoceratops fowleri.

Acknowledgments

The second specimen, NMMNH P-21100, an incomplete

We thank field assistants Michael Burns, Denver Fowler, War-

right squamosal, was attributed to ‘‘Torosaurus cf. utahensis’’

wick Fowler, Paul Sealey, and Warren Slade for collecting many

by Lehman (1985) and ‘‘Torosaurus utahensis’’ by Lehman

of the specimens that are the focus of this paper. Thanks are

(1993) and later considered by Farke (2002) and Sullivan et al.

extended to Robert Purdy and Mike Brett-Surman (United

(2005a) to be ‘‘Chasmosaurinae indeterminate.’’ The speci-

States National Museum of Natural History) for access to the

men is sufficiently different from the type of Ojoceratops fowleri

USNM collections and for information regarding specimens

in morphology—the former has a distinct ‘‘L’’-shape. However,

cited herein. We also thank Justin Spielmann (New Mexico

because of its incomplete nature, a proper generic and spe-

Museum of Natural History and Science) for his help in com-

cies assignment cannot be made although it most closely ap-

piling collection data for some of the earlier collected material

proaches ‘‘Torosaurus’’ utahensis. Therefore at least two cera-

in the NMMNH collections. Special thanks to Xiao-chun Wu

topsids are present in the Ojo Alamo Formation: Ojoceratops

(Canadian Museum of Nature) for an advance copy of the

fowleri, described above, and an indeterminate taxon repre-

Eotriceratops paper and data matrix and many discussions re-

sented by NMMNH P-21100.

garding phylogenetic analysis. We also benefited from discus-

The recently described partial parietal from the Naashoibito

sions with John ( Jack) R. Horner (Museum of the Rockies) and

Formation (NMMNH P-44477; Farke and Williamson 2006) is

thank him for his help and interest in this study. We also thank

assigned to cf. Ojoceratops fowleri based on similarity of thick-

Pat Hester (Bureau of Land Management–Albuquerque) and

ness and external sculpturing (see Fig. 11.5A, B). Farke and

Rich Simmons (Bureau of Land Management–Farmington) for

Williamson (2006) noted that an epiparietal (= epoccipital)

the respective institutional Paleontological Resource Use per-

located on the midline of the parietal suggests affinities with

mits and their continued support for collecting fossil verte-

Triceratops. However, we have noted above that the specimen

brates on lands administered by the BLM. Lastly, we thank

is a subadult and conforms in morphology to Ojoceratops

Michael Ryan (Cleveland Museum of Natural History) and

fowleri. Moreover, the distribution of a medially positioned

Brenda Chinnery-Allgeier (University of Texas–Austin) for

epiparietal among other crown taxa is not known with cer-

their support and interest in this contribution and for their

tainty. We note here that other ‘‘Triceratops-like features’’ have

suggestions and editorial skills that improved this paper.

178 sullivan & lucas

References Cited Brown, B. 1910. The Cretaceous Ojo Alamo beds of New Mexico with a description of the new dinosaur genus Kritosaurus. Bulletin of the American Museum of Natural History 28: 267–274. Burns, M. E. (in press). Taxonomic utility of ankylosaur (Dinosauria, Ornithischia) osteoderms: Glyptodotopelta mimus Ford 2000—A test case. Journal of Vertebrate Paleontology. Farke, A. A. 2002. A review of ‘‘Torosaurus’’ (Dinosauria: Ceratopsidae) specimens from Texas and New Mexico, USA. Journal of Vertebrate Paleontology 22(3, Suppl.): 52A. Farke, A. A., and T. E. Williamson. 2006. A ceratopsid dinosaur parietal from New Mexico and its implications for ceratopsid biogeography and systematics. Journal of Vertebrate Paleontology 26: 1018–1020. Fassett, J. E. 2005. Comment: The extinction of the dinosaurs in North America. GSA Today 15(7): 11. ———. 2009. New geochronologic and stratigraphic evidence confirms the Paleocene age of the dinosaur-bearing Ojo Alamo Sandstone and Animas Formation in the San Juan Basin, New Mexico and Colorado. Palaeontologia Electronica 12.1.3A Fassett, J. E., S. G. Lucas, R. A. Zielinski, and J. R. Budahn. 2000. Compelling new evidence for Paleocene dinosaurs in the Ojo Alamo Sandstone, San Juan Basin, New Mexico and Colorado, USA. Lunar and Planetary Institute Contribution 1053: 45–46. Fassett, J. E., R. A. Zielinski, and J. R. Budahn. 2002. Dinosaurs that did not die: Evidence for Paleocene dinosaurs in the Ojo Alamo Sandstone, San Juan Basin, New Mexico. Geological Society of America, Special Paper 356: 307–336. Fastovsky, D. E., and P. M. Sheehan. 2005a. The extinction of the dinosaurs in North America. GSA Today 15(3): 4–10. ———. 2005b. Reply: The extinction of the dinosaurs in North America. GSA Today 15(7): 11. Ford, T. 2000. A review of ankylosaur osteoderms from New Mexico and a preliminary review of ankylosaur armor. New Mexico Museum of Natural History and Science Bulletin 17: 157–176. Forster, C. A. 1996. New information on the skull of Triceratops. Journal of Vertebrate Paleontology 16: 246–258. Gilmore, C. W. 1916. Contributions to the geology and paleontology of San Juan County, New Mexico. 2. Vertebrate faunas of the Ojo Alamo, Kirtland, and Fruitland formations. U.S. Geological Survey Professional Paper 98: 279–302. ———. 1919. Reptilian faunas of the Torrejon, Puerco, and underlying Upper Cretaceous formations of the San Juan Basin, New Mexico. U.S. Geological Survey Professional Paper 19: 1–68. ———. 1922. A new sauropod dinosaur from the Ojo Alamo Formation of New Mexico. Smithsonian Miscellaneous Collections 72: 1–9. ———. 1946. Reptilian fauna of the North Horn Formation of central Utah. Geological Survey Professional Paper 210: 1–53. Horner, J. R., and M. B. Goodwin. 2006. Major cranial changes during Triceratops ontogeny. Proceedings of the Royal Society, Biological Sciences 273: 2757–2761. ———. 2008. Ontogeny of cranial epi-ossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Hunt, A. P., and S. G. Lucas. 1991. An associated Maastrichtian

hadrosaur and a Turonian ammonite from the Naashoibito Member, Kirtland Formation (Late Cretaceous: Maastrichtian), northwestern New Mexico. New Mexico Journal of Science 31: 27–35. Lambe, L. M. 1915. On Eoceratops canadensis, gen nov., with remarks on other genera of Cretaceous horned dinosaurs. Geological Survey of Canada, Geological Series 24: 1–49. Lehman, T. M. 1981. The Alamo Wash local fauna: A new look at the old Ojo Alamo fauna. In S. G. Lucas, J. K. Rigby, Jr., and B. S. Kues, eds., Advances in San Juan Basin Paleontology, pp.189–221. Albuquerque: University of New Mexico Press. ———. 1985. Stratigraphy, sedimentology, and paleontology of Upper Cretaceous (Campanian-Maastrichtian) sedimentary rocks in Trans-Pecos Texas. Ph.D. diss., University of Texas, Austin. ———. 1993. New data on the ceratopsian dinosaur Pentaceratops sternbergii Osborn from New Mexico. Journal of Paleontology 67: 279–288. Lehman, T. M., F. W. McDowell, and J. M. Connelly. 2006. First isotopic (U-Pb) age for the Late Cretaceous Alamosaurus vertebrate fauna of West Texas and its significance as a link between two faunal provinces. Journal of Vertebrate Paleontology 26: 922–928. Lucas, S. G., N. J. Mateer, A. P. Hunt, and F. M. O’Neill. 1987. Dinosaurs, the age of the Fruitland and Kirtland Formations, and the Cretaceous-Tertiary boundary in the San Juan Basin, New Mexico. In J. E. Fassett and J. K. Rigby, Jr., eds., The Cretaceous-Tertiary Boundary in the Raton and San Juan Basins of New Mexico and Colorado, pp. 35–50. Geological Society of America, Special Paper 209. Lucas, S. G., and R. M. Sullivan. 2000. Stratigraphy and vertebrate biostratigraphy across the Cretaceous-Tertiary boundary, Betonnie Tsosie Wash, San Juan Basin, New Mexico. New Mexico Museum of Natural History and Science, Bulletin 17: 95–104. Lull, R. S. 1933. A revision of the Ceratopsia or horned dinosaurs. Memoirs of the Peabody Museum of Natural History 3: 1–175. Marsh, O. C. 1888. A new family of horned Dinosauria from the Cretaceous. American Journal of Science 36: 477–478. ———. 1890. Description of new dinosaurian reptiles. American Journal of Science, Series 3, 39: 418–426. McDowell, F. W., T. M., Lehman, and J. N. Connelly. 2004. A U-PB age for the Late Cretaceous Alamosaurus vertebrate fauna of West Texas. Geological Society of America Abstracts with Programs 36: 6. McKenna, M. C. 2007. Linked aspects of nonmarine CretaceousTertiary boundary events. In K. C. Beard and Z.-X. Luo, eds., Mammalian Paleontology on a Global Stage: Papers in Honor of Mary R. Dawson, pp. 49–56. Bulletin of the Carnegie Museum of Natural History 30. Parks, W. A. 1925. Arrhinoceratops brachyops, a new genus and species of Ceratopsia from the Edmonton Formation of Alberta. University of Toronto Studies 19: 5–15. Seeley, H. G. 1888. The classification of the Dinosauria. Report of the British Association for the Advancement of Science 1887: 698– 699. Sullivan, R. M., and S. G. Lucas. 2003. The Kirtlandian, a new land-vertebrate ‘‘age’’ for the Late Cretaceous of western North America. New Mexico Geological Society, Guidebook 54: 369–377.

A New Chasmosaurine from the Upper Cretaceous Ojo Alamo Formation 179

———. 2006. The Kirtlandian land-vertebrate ‘‘age’’ -faunal composition, temporal position and biostragraphic correlation in the nonmarine Upper Cretaceous of western North America. New Mexico Museum of Natural History and Science Bulletin 35: 7–29. Sullivan, R. M., A. C. Boere, and S. G. Lucas. 2005a. Redescription of the ceratopsid dinosaur Torosaurus utahensis (Gilmore 1946) and a revision of the genus. Journal of Paleontology 79: 564– 582. ———. 2005b. Dinosaurs, pollen, and the Cretaceous-Tertiary boundary in the San Juan Basin, New Mexico. New Mexico Geological Society, 56th Field Conference Guidebook: 395–407. Sullivan, R. M., S. G. Lucas, and D. R. Braman. 2003. No Paleocene dinosaurs in the San Juan Basin, New Mexico. Geological Society of America, Rocky Mountain Section, 55th Annual Meeting, 35: 15. Tyson, H. 1981. The structure and relationships of the horned dinosaur Arrhinoceratops Parks (Ornithischia: Ceratopsia). Canadian Journal of Earth Sciences 18: 1241–1247. Weil, A., and T. E. Williamson. 2000. Diverse Maastrichtian terrestrial vertebrate fauna of the Naashoibito Member, Kirtland

180 sullivan & lucas

Formation (San Juan Basin, New Mexico) confirms ‘‘Lancian’’ faunal heterogeneity in western North America. Geological Society of America, Abstracts with Programs, A-498. Weil, A., T. E. Williamson, F. Pignataro, and J. Colon. 2004. The teiid lizard Peneteius discovered in the Upper Cretaceous Naashoibito Member of the Kirtland Formation, San Juan Basin, New Mexico. Journal of Vertebrate Paleontology 24(3, Suppl.): 127A. Williamson, T. E. 2000. Review of the Hadrosauridae (Dinosauria, Ornithischia) from the San Juan Basin, New Mexico. New Mexico Museum of Natural History and Science Bulletin 17: 191–213. Williamson, T. E., and A. Weil. 2004. First occurrence of the teiid lizard Peneteius from the latest Cretaceous Naashoibito Member, Kirtland Formation—San Juan Basin, New Mexico. New Mexico Geology 26: 65. Wu, X.-C., D. B. Brinkman, D. A. Eberth, and D. R. Braman. 2007. A new ceratopsid dinosaur (Ornithischia) from the uppermost Horseshoe Canyon Formation (upper Maastrichtian), Alberta, Canada. Canadian Journal of Earth Sciences 44: 1243–1265.

12 A New Chasmosaurine Ceratopsid from the Judith River Formation, Montana M I C H A E L J . R YA N , A N T H O N Y P. R U S S E L L , A N D S C O T T H A R T M A N

a new chasmosaurine ceratopsid, Medusaceratops

the Western Interior Seaway of North America, with the ex-

lokii, is described based on material collected from a

ception of the fragmentary taxon Turanoceratops (discussed

bonebed in the Judith River Formation (Campanian)

by Chinnery-Allgeier and Kirkland this volume), which is

near Havre, Montana. Originally, all ceratopsid material

from the Turonian of Uzbekistan (Nessov et al. 1989). Recent

from the bonebed was referred to the basal centrosaurine

work (e.g., Ryan and Russell 2005; Ryan 2007; Lucas et al.

Albertaceratops Ryan 2007, the holotype of which was

2006; and contributions to this volume) has significantly in-

collected from the Oldman Formation of Alberta, Can-

creased the number of ceratopsid taxa, and, taken together

ada. Reassessment of several key features of parietals

with new descriptions of basal neoceratopsians (see You et al.

from the Montanan bonebed, including the number and

and Chinnery-Allgeier and Kirkland this volume, for reviews),

shape of the preserved epiparietals, necessitates referral

the number of described ceratopsians has more than doubled

of at least some material from this site to the new taxon.

in the past decade.

Although the bonebed does include centrosaurine ele-

Although the Upper Cretaceous rocks of Montana have

ments (including a lateral parietal ramus and a squam-

yielded a wealth of dinosaur taxa (more than two dozen: Trex-

osal) that may eventually be referable to Albertaceratops,

ler 2001; Weishampel et al. 2004), they have been surprisingly

it appears to be dominated by elements of chas-

sparing in producing ceratopsid material, especially consid-

mosaurine affinity. In addition to being the first un-

ering the abundance of horned-dinosaur taxa known from the

equivocal occurrence of a Campanian-aged

adjacent, and at least partially contemporaneous, Belly River

chasmosaurine ceratopsid in Montana, Medusaceratops

Group in Alberta. Basal neoceratopsian material from Mon-

lokii is also the oldest known Chasmosaurine ceratopsid

tana includes Cerasinops, Prenoceratops, and Montanocera-

(approximately 77.5 Ma).

tops from the Two Medicine Formation. Campanian-aged Montanan ceratopsids are all centrosaurines and include Ache-

Introduction

lousaurus (Sampson 1995), Einiosaurus (Sampson 1995) and Styracosaurus ovatus (Gilmore 1930; although this taxon is

Until the early 2000s, Ceratopsidae was limited to 13 named

transferred to a new genus by McDonald and Horner this vol-

genera in two clades (‘‘subfamilies’’), Centrosaurinae and

ume) from the upper portion of the Two Medicine Formation,

Chasmosaurinae (Ryan and Russell 2001). The specimens

and Avaceratops from the Judith River Formation. Additional

were known exclusively from Upper Cretaceous sediments of

problematic, putative centrosaurine elements have been col-

181

lected from both formations but all the designated taxa (Mono-

upper part of the Judith River Formation (Fig. 12.1). In this

clonius crassus, M. fissus, M. recurvicornus, and M. sphenocerus;

region of Montana the Judith River Formation is lithologically

see Hatcher et al. 1907 for illustrations and discussion) are

equivalent to the Oldman Formation of Canada, sharing

nomina dubia (Dodson et al. 2004).

the same source rocks in northwestern Montana (Eberth and

One other taxon of note, Ceratops montanus, described by

Hamlin 1993), and differing in name solely because of the

Marsh in 1888 based on an occipital condyle and a pair of

presence of the international border. To the north, the well-

large chasmosaurine-like postorbital horncores (USNM 2411,

exposed Oldman Formation is thick (over 100 m) and infor-

nomen dubium; Dodson et al. 2004), was collected from the top

mally divided into three stratigraphic units, 1–3. In ascend-

of the Judith River beds in the Cow Creek Valley, about 16 km

ing order these are unit 1 (isolated paleochannels), unit 2

upstream of the confluence of this stream with the Missouri

(the Comrey Sandstone Zone; amalgamated, multimeter-

River. Stratigraphically the host stratum of C. montanus is

thick paleochannels), and unit 3 (isolated paleochannels)

probably approximately equivalent to the locality of the type

(Eberth 2005). To the south, in the region of the Montanan

of Albertaceratops and the new chasmosaurine material re-

bonebed, the Judith River (= Oldman) Formation is thinner,

ported on in this paper.

with much less exposure, but geologically the bonebed lies in

The new bonebed material was first reported by Sweeney

what would be equivalent to the bottom of unit 1 of the Old-

and Boyden (1993), who suggested that the material rep-

man Formation (Eberth pers. com.). This places the bonebed

resented the southernmost occurrence of Styracosaurus alber-

at approximately the same level as the holotype of Albertacera-

tensis, based on misidentification of the supraorbital horns

tops, making is approximately the same age, 77.5 Ma, and thus

as parietal spikes. When the supraorbital horncores were later

represents the oldest known chasmosaurine ceratopsid.

correctly identified, Trexler and Sweeney (1995) noted the similarity of the cranial material to that of type material of

Systematic Paleontology

Ceratops montanus (nomen dubium), but could not refer elements from the bonebed to any valid ceratopsid taxon. The bonebed is located on private land and historically has

Ornithischia Seeley 1888 Ceratopsia Marsh 1890

been excavated by a number of commercial companies. It is

Neoceratopsia Sereno 1986

referred to as the ‘‘Mansfield Bonebed’’ in reference to the

Ceratopsidae Marsh 1888

landowner. Most recently, material from this bonebed, in-

Chasmosaurinae Lambe 1915

cluding the material presented here, has been purchased from

Genus Medusaceratops gen. nov.

Canada Fossils, Ltd., of Calgary, Alberta, and accessioned in the collections of the Wyoming Dinosaur Center (WDC). Ad-

Type Species. Medusaceratops lokii, new species.

ditional material was purchased and accessioned by the Royal

Generic Etymology. Medusa (mythological figure, Greek) +

Tyrrell Museum of Palaeontology (TMP). Canada Fossils, Ltd.,

ceratops (horned-face, Latinized Greek). Medusa was a mon-

has also assembled two composite skeletons using elements

ster in Greek mythology with ‘‘hair’’ comprised of snakes and

from the bonebed and these are in the collections of the WDC

a gaze that could turn men to stone. The allusion refers to the

(WDBC MC-001) and the Fukui Prefectural Dinosaur Museum

large, thick snake-like spikes that extend from the lateral mar-

(FPDM-V-10). Neither cast has an exact reconstruction of the

gins of the posterior portion of the parietal.

Medusaceratops parietal.

Diagnosis. Chasmosaurine ceratopsid with only three epi-

This paper describes diagnostic material from the Mansfield

parietals (P1–3) on the lateral parietal ramus: wide-based,

Bonebed in the Judith River Formation previously referred

pachyostotic, dorsoventally depressed, curved processes at

to the centrosaurine Albertaceratops (Ryan 2007), and assigns

the posterolateral and lateral margin that are large (P1) and

it to a new taxon that represents the first unequivocal oc-

small (P2), respectively, and; P3 small, unmodified, dorsoven-

currence of a Campanian-aged chasmosaurine ceratopsid in

tally depressed, triangular epiparietal adjacent to the squamo-

Montana.

sal contact.

Institutional Abbreviations. FPDM: Fukui Prefectural Dino-

The P1 parietal ornamentation closely resembles the P3 par-

saur Museum, Fukui; TMP: Royal Tyrrell Museum of Palae-

ietal process of Albertaceratops, but Medusaceratops differs in

ontology, Drumheller; WDC: Wyoming Dinosaur Center,

lacking tab-shaped, frequently imbricated P4–7 processes (as

Thermopolis.

do all chasmosaurines) of centrosaurines. Medusaceratops lokii sp. nov.

Geology and Setting

Holotype. WDC-DJR-001, a partial parietal (Fig. 12.2).

The bonebed is located in the badlands on the west side of

Paratype. WDC-DJR-002, a partial parietal (Fig. 12.2).

Kennedy Coulee adjacent to the Milk River, and occurs in the

Specific Etymology. Loki (mythological figure, Norse). Loki

182 ryan, russell, & hartman

FIGURE 12.1.

Locality map of the Mansfield Bonebed near Havre, Montana. Inset photograph indicates the host horizon for the bonebed in the lower part of ‘‘unit 1’’ Oldman Formation equivalent of the Judith River Formation, in Kennedy Coulee, adjacent to the Milk River.

was a Norse god who contrived mischief for his fellow gods,

Description

the name thus alluding to the confusion experienced in trying to assign taxonomic designations to the material collected

Although all the chasmosaurine material from the Mansfield

from the bonebed.

Bonebed is referred to Medusaceratops, only the holotype and

Diagnosis. As for the genus.

one additional parietal specimen are described here. The re-

Type Locality. The Mansfield Bonebed in the Judith River

maining material is presently being reexamined by MJR.

Formation (899 masl) near Havre, Montana, along the west

WDC-DJR-OO1 (Fig. 12.2) is a complete, adult-sized, right

side of Kennedy Coulee bordering the Milk River (Fig. 12.1;

lateral parietal ramus. At the caudolateral margin the putative

locality data on file with the Wyoming Dinosaur Center and

P1 epiparietal (sensu Horner and Goodwin 2008; following

the Royal Tyrrell Museum of Paleontology).

the numbering convention of Sampson et al. 1997) is a large,

Comments. The holotype and additional uncatalogued

dorsoventrally depressed, rugose, laterally directed and ros-

material from the Mansfield Bonebed are curated at the

trally curved hook-like process that superficially resembles

Wyoming Dinosaur Center. Some original material (TMP

the P3 process of Albertaceratops, as well as that of the new cen-

2002.69.1–10) and casts (TMP 2002.28–38), mostly repre-

trosaurine from the Wahweap (Kirkland and Deblieux 2006,

senting nondiagnostic cranial elements, are curated at the

this volume), but differs from the spike-like P3 process of

Royal Tyrrell Museum of Palaeontology. Aside from the cast of

Achelousaurus, Einiosaurus and Pachyrhinosaurus, or the short,

WDC-DJR-001 (TMP 2002.3.28), this material is presently re-

tongue-like process of Centrosaurus brinkmani. The P1 process

ferable only to either Centrosaurinae or Chasmosaurinae.

is thickened proximally along its caudal and medial margin

A New Chasmosaurine Ceratopsid from the Judith River Formation 183

FIGURE 12.2. Medusaceratops lokii parietals. (A) Dorsal and (B) ventral right parietal ramus, WDC-DJR-001; (C) ventral and (D) dorsal left parietal process P1, WDC-DJR-002. Inset is a reconstruction of the frill based on WDC-DJR-001. P1–3: parietal processes #1–3; SQC: squamosal contact. Scale bar is 10 cm.

and is distinctly raised above the underlying parietal. The P1

tinctly offset from the P2 process by a small, rounded notch

process is not clearly demarcated from the parietal ventrally,

and sits next to a large, rounded marginal scallop that is con-

but rather progressively grades into the smooth ventral sur-

fluent with the squamosal contact.

face. Distally the process becomes progressively thinner along

A 150 mm long squamosal suture forms the preserved ros-

its cranial and lateral margins, terminating at a broken or

tral margin of the ramus and wraps around approximately 50

eroded margin, which, if complete, would have rounded out

mm onto the lateral surface, indicating that, as in Triceratops

this side to form a robust hook.

and centrosaurines, an epi-ossification probably bridged the

The P1 process is separated from the adjacent P2 process by

marginal gap between the parietal and the squamosal.

a smooth, saddle-shaped surface. The preserved apex of P1

WDC-DJR-OO1 (Fig. 12.2A, B) is the only parietal to pre-

arches rostrally over approximately one-half the length of the

serve any of the caudal margin medial to the P1 process. A

P2 process. The P2 process is an elongate, lobe-shaped, rugose

110-mm section of the caudal ramus extends medially from

structure with a hook-shaped rostral margin rendering it as a

the base of the P1 process. The caudal surface of this section is

smaller, lower version of the P1 ornamentation. As seen on

rounded, and the sectional profile of the process tapers ros-

many large, adult-sized chasmosaurine parietals, the base of

trally. This short section of caudal parietal more closely resem-

this epiparietal is not clearly defined and grades into the adja-

bles the strap-like caudal parietal rami of Chasmosaurus belli

cent smooth surface of the parietal.

and C. russelli than it does of any centrosaurine.

The P3 process is a small, depressed, triangular epiparietal

The preserved smooth parietal ramus thins to millimeter

that sits on the parietal margin, similar to those seen at the

thickness within 100 mm from the epiparietals, suggesting

caudolateral margins of Chasmosaurus. The midpoint of its

that Medusaceratops had large parietal fenestrae. The ramus

base is penetrated by a small, 4 mm wide, foramen. P3 is dis-

also exhibits deep, smooth-walled, putative vascular depres-

184 ryan, russell, & hartman

sions on both surfaces that are also prominent features on

specimens of the geologically youngest chasmosaurines (Di-

large-bodied chasmosaurines.

ceratops, Eotriceratops, Torosaurus, and Triceratops) tend to have

WDC-DJR-OO2 is also a partial right parietal ramus. Al-

epiparietals with more ovate profiles, that, at least for Tricera-

though it is less complete than WDC-DJR-OO1, its P1 and P2

tops, are the result of ontogenetic modification. Chasmosaur-

processes closely resemble those of that specimen, whereas

ine parietal ornamentation is typically more modestly devel-

the partially preserved P3 appears to have a less sharply de-

oped than that of centrosaurines, with only the medialmost

fined apex than that of WDC-DJR-OO1.

(Anchiceratops and Pentaceratops) or caudolaterally positioned

The remaining catalogued parietal specimens (TMP

epiparietals (Agujaceratops, Chasmosaurus, Pentaceratops) oc-

2002.03.38; Ryan 2007: figs. 11.3 and 11.4; TMP 2002.69.5,

casionally being elaborated as larger, more robust triangular

TMP 2002.69.6, TMP 2002.69.7, and TMP 2002.3.29) are all

ornamentation.

from adult-sized specimens and preserve some portion of the

Superficially the P1 hooked process at the caudolateral margin of the parietal of Medusaceratops resembles the P3 process

putative P1 or P2 processes but are not diagnostic.

of Albertaceratops nesmoi, which lead to the initial referral of the ceratopsid material from the Mansfield Bonebed to this

Discussion

taxon. However, the parietal is otherwise distinct from Alber-

Although Ryan (2007) referred all the ceratopsid material

taceratops in both the shape of the remaining epiparietals and

from the Mansfield Bonebed to Albertaceratops, he was con-

having only a total of three epiparietals on the lateral parietal

founded by the distinct difference in the number and shape of

ramus. Since all centrosaurines have at least seven loci for epi-

epiparietals preserved on the best parietals versus those of Al-

parietals and only chasmosaurines exhibit three loci on the

bertaceratops. All incongruities with the epiparietals disappear

lateral parietal ramus, this necessitates the referral of at least

when specimens such as WDC-DJR-OO1 and WDC-DJR-OO2

the elements discussed here to a new chasmosaurine taxon,

are considered as chasmosaurine rather than centrosaurine.

Medusaceratops lokii, making it the first unequivocal chasmo-

Features of the epiparietals are important for making taxonomic determination at both the subfamily (Centrosaurinae

saurine ceratopsid recovered from the Judith River Formation of Montana.

vs. Chasmosaurinae) and lower taxonomic levels. Epiparietals

The complete morphology of the midline ramus of

formed in the overlying dermis and, through fusion and mod-

Medusaceratops is unknown. The only preserved portion of the

ification of the underlying parietal, helped generate diagnos-

caudal parietal ramus is that of the holotype, WDC-DJR-001.

tic frill ornamentation. All centrosaurines (with the exception

Notably this straight, 110 mm portion medial to the P1 hook

of Avaceratops which lacks parietal ornamentation and is ex-

more closely resembles the smooth, strap-like bar of Chasmo-

cluded from discussion for reasons discussed elsewhere, e.g.,

saurus than it does the gently arching caudal margin of most

Sampson et al. 1997; Ryan 2007) have seven loci (P1–7) on the

centrosaurines. It also differs from the reconstructed caudal

margin of the parietal adjacent to the midline. In centrosau-

margin caudal bar of Albertaceratops on which the pachyosto-

rines, when present, the first three epiparietals (and occa-

tic hooks develop adjacent to a ‘‘U’’-shaped, smooth-textured

sionally the fourth in Styracosaurus) are modified on adult-

notch on the caudal midline, identical to the notch seen

sized animals into robust hooks or spikes (parietal loci and

on most centrosaurines with large P3 processes. When re-

their corresponding epiparietals are numbered sequentially,

constructed to articulate with a typical chasmosaurine-shaped

progressing anteriorly from the midline; see Sampson et al.

squamosal (Fig. 12.2, inset) the dorsal profile of the frill most

1997). The remaining crescent-shaped epiparietals are typi-

resembles those of chasmosaurines with deep midline embay-

cally unmodified on adult-sized parietals, although they are

ments of the caudal parietal margin (e.g., Agujaceratops, Penta-

usually imbricated, can be slightly attenuated on Albertacera-

ceratops, and some specimens of Chasmosaurus belli and C.

tops and some specimens of Centrosaurus (becoming occa-

russelli ).

sionally spike-like on specimens of Styracosaurus), and are

So, what of the rest of the material from the Montanan bonebed? The taxonomic composition of the Mansfield Bone-

‘‘loosely’’ attached to the parietal. On most chasmosaurines only three epiparietals (P1–3) de-

bed specimens remains unclear. Confounding analysis is the

velop per side with an additional single midline epiparietal

fact that Montanan bonebed was not excavated using rigor-

occurring on Triceratops, and a pair occurring on the caudo-

ous scientific methodology, and any associated field notes

dorsal surface of the parietal midline bar of Agujaceratops, An-

were not preserved to assist in assessing the recovered mate-

chiceratops and Pentaceratops. The autapomorphic Chasmo-

rial. It is also unclear whether all excavated material was kept

saurus irvinensis has five per side, and Torosaurus can have

or whether some elements were selectively excluded from col-

up to six. The unmodified epiparietals of chasmosaurines are

lection. As suggested by Ryan (2007), the Mansfield Bonebed

typically distinctly triangular (e.g., those of Agujaceratops, An-

appears to contain a mixture of both centrosaurine and chas-

chiceratops, Chasmosaurus, Pentaceratops), although mature

mosaurine elements. Indeed, a number of centrosaurine ele-

A New Chasmosaurine Ceratopsid from the Judith River Formation 185

FIGURE 12.3.

Reconstruction of Medusaceratops lokii by Donna Sloan based on WDC-DJR-001 and putative chasmosaurine material from the Mansfield bonebed.

ments, including a distinctive centrosaurine lateral parietal

The remainder of the adult-sized ceratopsid material from

bar with four tab-shaped, imbricated epiparietals, were incor-

the Mansfield Bonebed currently available for study (much of

porated into the composite chasmosaurine indet. skull and

it postcranial), is distinctly larger than Campanian-aged cen-

skeleton (FDMJ-V-10) assembled by Canada Fossils, Ltd., now

trosaurine material other than Pachyrhinosaurus canadensis,

in the collections of the Fukui museum in Japan. Although

which also suggests that it is better referred to Chasmosau-

monodominant centrosaurine bonebeds are the norm in the

rinae that typically obtain large body size in the Late Creta-

Belly River Group in Canada, even these yield rare occurrences

ceous. Collected from the bonebed are several unnumbered

of chasmosaurine elements (e.g., a partial Chasmosaurus skull

and numbered partial skulls (e.g., WDCB-MC-001; Ryan 2007:

[TMP 79.11.147] found in the Centrosaurus apertus Bonebed 43

fig. 9) that bear large, robust postorbital horns, but that lack

in the Dinosaur Park Formation of Alberta). Chasmosaurine

attached squamosals and parietals, or other taxonomically in-

bones beds are rare (e.g., Agujaceratops in Big Bend National

formative features. Such specimens cannot be reliably referred

Park, Texas) and, to date, no centrosaurine material has been

to any known ceratopsid taxon.

identified from them (Lehman 2007).

186 ryan, russell, & hartman

One subadult-sized elongate postorbital horncore with an

oval base was tentatively suggested to be chasmosaurine (TMP

Evans, and Andy Farke for thoughtful reviews. The Phaeton

2002.2.30; Ryan 2007: figs. 10.4 and 10.5) and we reaffirm

Group assisted with logistics. The Dinosaur Research Institute

that referral here. It is possible that the large postorbital horn-

provided funding to help offset the cost of producing the casts

cores collected from the bonebed are from Marsh’s (1888)

that are curated at the TMP. Gilles Danis assisted with the

Ceratops montanus. Several of the partial skulls from the bone-

production of the molds. This work was supported, in part, by

bed have long, robust postorbital horncores; however, with-

an NSERC Discovery grant (9747-03) to A. P. Russell.

out associated frill material the horncores cannot be considered diagnostic and Ceratops must, therefore, remain a nomen

References Cited

dubium. The unequivocal taxonomic assignment of cera-

Chinnery-Allgeier, B. J., and J. T. Kirkland. 2010. An update on the paleobiogeography of ceratopsian dinosaurs. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 387–404. Bloomington: Indiana University Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A. 2005. The geology. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 312–348. Bloomington: Indiana University Press. Eberth, D. A., and A. P. Hamblin. 1993. Tectonic, stratigraphic, and sedimentologic significance of a regional discontinuity in the upper Judith River Formation (Belly River Wedge) of southern Alberta, Saskatchewan, and northern Montana. Canadian Journal of Earth Sciences 30: 174–200. Gilmore, C. W. 1930. On dinosaur reptiles from the Two Medicine Formation of Montana. Proceedings of the U.S. National Museum 77: 1–39. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49. Horner, J. R., and Goodwin, M. B. 2008. Ontogeny of cranial epiossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Kirkland, J. I., and D. D. DeBlieux. 2006. A new genus of ornate long-horned centrosaurine ceratopsian from the middle Campanian (Cretaceous) Wahweap Formation, Grand Staircase– Escalante National Monument, southern Utah. Journal of Vertebrate Paleontology 26(3, Suppl.): 85A. ———. 2010. New basal centrosaurine ceratopsian skulls from the Wahweap Formation (middle Campanian), Grand Staircase– Escalante National Monument, southern Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 117–140. Bloomington: Indiana University Press. Lambe, L. M. 1915. On Eoceratops canadensis, gen nov., with remarks on other genera of Cretaceous horned dinosaurs. Geological Survey of Canada, Geological Series 24: 1–49. Lehman, T. M. 1997. Late Campanian dinosaur biogeography in the Western interior of North America. In D. Wolberg, E. Stump, and G. Rosenberg, eds., Dinofest International: Proceedings of a Symposium Held at Arizona State University, pp. 223– 240. Philadelphia: Academy of Natural Sciences. ———. 2007. Growth and population age structure in the horned dinosaur Chasmosaurus. In K. Carpenter, ed., Horns and Beaks:

topsid material to from the Mansfield Bonebed will only be possible with the collection of more complete specimens in the future. In addition to being the first definitive chasmosaurine described from the Judith River Formation of Montana, Medusaceratops (Fig. 12.3) is also the oldest known member of this clade. At an estimated 77.5 Ma it is at least 1.0 Ma older than the next oldest chasmosaurine, Chasmosaurus russelli, from the Dinosaur Park Formation of Alberta (Ryan and Evans 2005; Sampson and Loewen this volume). The presence of the sister taxon of the Ceratopsidae, Zuniceratops, in the Turonian of west-central New Mexico suggests that Ceratopsidae may have originated in the southern paleobiogeographical zone (sensu Lehman 1997). However, the oldest chasmosaurines (Chasmosaurus and Medusaceratops) are from the northern biogeographcal zone (Alberta and Montana) suggesting that this subfamily may have originated here and then migrated to, and diversified in, the south. Finally, if the large postorbital brow horns recovered from the Mansfield Bonebed are eventually associated with Medusaceratops it would also suggest that massive chasmosaurine brow horns appeared earlier in the fossil record than has been previously reported (i.e., Pentaceratops, ca. 75 Ma). Acknowledgments

Thanks to the Royal Tyrrell Museum of Palaeontology for hosting the Horned Dinosaur Symposium, and especially Don Brinkman for driving its organization. The material described here was first brought to the attention of MJR by Canada Fossils, Ltd., Calgary, Canada, who allowed him to examine the material along with Stephen Godfrey and Darren Tanke, both of whom shared in the initial discussions of the material. Thanks to the staff of the TMP, especially Jim Gardner and Bruce Naylor, who arranged for some of the original and cast material from the bonebed to be added to its collections. David Eberth investigated the Mansfield Bonebed and graciously shared his unpublished data and the photograph in Fig. 12.1. Donna Sloan drew the reconstruction in Fig. 12.3. Thanks to Amer Dzindic, Pierre Pare, and the late René Vandervelde at Canada Fossils, Ltd., and Berkhard Pohl and the staff of the Wyoming Dinosaur Center for allowing access to their material. Thanks to Brenda Chinnery-Allgeier, David

A New Chasmosaurine Ceratopsid from the Judith River Formation 187

Ceratopsian and Ornithopod Dinosaurs, pp. 259–318. Bloomington: Indiana University Press. Lucas, S. G., R. M. Sullivan, and A. P. Hunt. 2006. Re-evaluation of Pentaceratops and Chasmosaurus (Ornithischia: Ceratopsidae) in the Upper Cretaceous of the Western Interior. New Mexico Museum of Natural History and Science Bulletin 35: 367–370. Marsh, O. C. 1888. A new family of horned dinosaurs from the Cretaceous. American Journal of Science, Series 3, 36: 477–478. ———. 1890. Description of new dinosaurian reptiles. American Journal of Science, Series 3 39: 418–426. McDonald, A. T., and J. R. Horner. 2010. New material of ‘‘Styracosaurus’’ ovatus from the Two Medicine Formation of Montana. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 156–168. Bloomington: Indiana University Press. Nessov, L. A., L. F. Kaznyshikina, and G. O. Cherepanov. 1989. Mesozoic ceratopsian dinosaurs and crocodiles of central Asia. In T. N. Bogdanova and L. I. Khozatsky, eds., Theoretical and Applied Aspects of Modern Paleontology, pp. 144–154. Leningrad: Proceedings of the XXXIII Session of the All-Union Paleontological Society. [In Russian.] Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., and D. C. Evans. 2005. Review of the Ornithischia of Dinosaur Provincial Park. In P. J. Currie and E. Kopplehus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 313–348. Bloomington: Indiana University Press. Ryan, M. J., and A. P. Russell. 2001. Dinosaurs of Alberta (Exclusive of Aves). In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 279–297. Bloomington: Indiana University Press. ———. 2005. A new centrosaurine ceratopsid from the Oldman Formation of Alberta and its implications for centrosaurine taxonomy and systematics. Canadian Journal of Earth Sciences 42: 1369–1387. Sampson, S. D. 1995. Two new horned dinosaurs from the upper Cretaceous Two Medicine Formation of Montana, with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15: 743–760.

188 ryan, russell, & hartman

Sampson, S. D., and M. A. Loewen. 2010. Unraveling a radiation: A review of the diversity, stratigraphic distribution, biogeography, and evolution of horned dinosaurs (Ornithischia: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 405–427. Bloomington: Indiana University Press. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral phylogenetic implications. Zoological Journal of the Linnean Society 121: 293–337. Seeley, H. G. 1888. The classification of the Dinosauria. Report of the British Association for the Advancement of Science 1887: 698– 699. Sereno, P. C. 1986. Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). National Geographic Research 2: 234–256. Sweeney, F., and W. M. Boyden. 1993. A first report of the southern most occurrence of the ceratopsian dinosaur Styracosaurus albertensis, the first found in the United States. Journal of Vertebrate Paleontology 13(3, Suppl.): 59A. Trexler, D. 2001. Two Medicine Formation, Montana: Geology and Fauna. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 298–309. Bloomington: Indiana University Press. Trexler, D., and F. G. Sweeney. 1995. Preliminary work on a recently discovered ceratopsian (Dinosauria: Ceratopsidae) bonebed from the Judith River Formation of Montana suggests the remains are of Ceratops montanus Marsh. Journal of Vertebrate Paleontology 15(3, Suppl.): 57A. Weishampel, D. B., P. M. Barrett, R. Coria, J. Le Loeuff, X. Xing, Z. Xijin, A Sahni, E. M. P. Gomani, and C. R. Noto. 2004. Dinosaur Distribution. In D. B. Weishampel, P. Dodson, and H.Osmólska, eds., The Dinosauria, 2nd ed., pp. 614–626. Berkeley: University of California Press. You, H., K. Tanoue, and P. Dodson. 2010. A new species of Archaeoceratops (Dinosauria: Neoceratopsia) from the Early Cretaceous of the Mazongshan area, northwestern China. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 59–67. Bloomington: Indiana University Press.

13 Description of a Complete and Fully Articulated Chasmosaurine Postcranium Previously Assigned to Anchiceratops (Dinosauria: Ceratopsia) JORDAN C. MALLON AND ROBERT HOLMES

a nearly complete ceratopsid postcranial skeleton with

Deer River valley, near Rumsey, Alberta (Fig. 13.1). Most of

associated skull fragments, long attributed to Anchicera-

the skull was missing, but the postcranium was complete and

tops, is described. The skeleton is identified as a chasmo-

fully articulated. The skeleton was attributed to Anchiceratops

saurine, but the paucity of diagnostic cranial material

by Sternberg, although he did not give a reason for his identi-

prevents more precise identification. Based on its prove-

fication. The specimen was subsequently prepared as a panel

nance from unit 2 of the Horseshoe Canyon Formation

mount in 1929 and displayed in the National Museum of Can-

of Alberta, the skeleton probably pertains to either An-

ada (now the Canadian Museum of Nature) with the cast skull

chiceratops or Arrhinoceratops. The skeleton exhibits sev-

of the type of A. ‘‘longirostris’’ (Sternberg 1929) in place.

eral unique features, including a humerus with an

Given the rarity of complete, fully articulated ceratopsid

enlarged deltopectoral crest and underdeveloped medial

postcrania (only one other specimen of comparable quality is

tubercle, thickened ribs, and a unique vertebral formula

known; Brown 1917), it is surprising that, except for a brief

(10 cervicals, 13 thoracics, 12 sacrals, and 39 caudals).

description by Lull (1933), this skeleton has remained unde-

The skeleton is also unusually stout and robust for its

scribed for so long. Here we describe the specimen in more

size, lending credence to the idea that it may have been

detail, paying particular attention to those defining features of

adapted to a semi-aquatic lifestyle much like the com-

the skeleton. The fragmentary skull material is also described

mon hippo (Hippopotamus amphibious). This interpreta-

for the first time in an effort to confirm the taxonomic iden-

tion is supported by the estuarine depositional

tity of this specimen.

environment in which the skeleton was found.

Institutional Abbreviations. AMNH: American Museum of Natural History, New York; CM: Carnegie Museum of Natural

Introduction

History, Pittsburgh; CMN: Canadian Museum of Nature, Ottawa; NSM-PV: National Science Museum, Division of Ver-

In the late summer of 1925, Charles M. Sternberg and his team

tebrate Paleontology, Tokyo; PMU: University of Uppsala,

collected the skeleton of a ceratopsid dinosaur from the upper

Sweden; ROM: Royal Ontario Museum, Toronto; TMP: Royal

Horseshoe Canyon Formation (Edmonton Group) of the Red

Tyrrell Museum of Palaeontology, Drumheller.

189

FIGURE 13.1.

Locality map for CMN 8547 modified from Eberth (unpublished). The specimen locality is indicated by an asterisk SW of Rumsey. More detailed locality information is available from the TMP Collections Section. Upper right inset shows field area in relation to the Province of Alberta.

Systematic Paleontology Ornithischia Seeley 1888

Description of CMN 8547 SKULL

Ceratopsia Marsh 1890 Ceratopsidae Marsh 1888

Four flattened frill fragments are associated with the skeleton

Chasmosaurinae indet. Lambe 1915

of CMN 8547 (Fig. 13.3). According to Sternberg’s field notes (1925, CMN), these belonged to the left side of the frill, but

Material. CMN 8547, a nearly complete, articulated post-

based on their curvature, probably pertain to the right side. At

cranial skeleton with associated frill fragments. The ‘‘anterior

their thickest points, usually located along the frill margin,

portion of the beak’’ reported by Sternberg in his field notes

the fragments measure approximately 28–30 mm. At their

(1925, CMN) could not be located. The unguals of the first and

thinnest points, they measure 7–10 mm.

third digits of the left manus depicted in both old field photo-

Two of the four frill fragments (Fig. 13.3A, B) form part of the

graphs and photographs of the original panel mount have

scalloped frill margin. The largest of these, a partial right squa-

since been lost.

mosal 704 mm long, bears two episquamosals (sensu Horner

Locality and Horizon. TMP locality L1508. East side of Red

and Goodwin 2008) on its proximalmost edge immediately

Deer River, 10.3 km SW of Rumsey, AB, in NW1/4 of S6-T33-

adjacent to the jugal notch. These are followed by a series of

R21-W4, 725 masl. Uppermost unit 2 of Horseshoe Canyon

four low, blunt scallops lacking sutures at their bases that indi-

Formation (Fig. 13.2), approximately 26 m below coal seam 11

cate where the more distal episquamosals were positioned.

(Carbon coal zone), 16 m below an extensive oyster bed in the area (part of the Drumheller Marine Tongue).

190 mallon & holmes

One of the frill fragments (Fig. 13.3C) is scored by a few shallow (3 mm), subparallel vascular sulci that run across both

FIGURE 13.2. Stratigraphy of CMN 8547. Specimen was found approximately 26 m below coal seam 11, in the uppermost unit 2 of the Horseshoe Canyon Formation. Numbered horizontal bars indicate the stratigraphic occurrence of coal seams recognized by Gibson (1977) and other authors. Fm.: Formation; B. Fm.: Battle Formation; W. Fm.: Whitemud Formation. Modified from Wu et al. (2007).

FIGURE 13.3. Skull material attributed to CMN 8547. (A) Partial squamosal bearing two episquamosals (i, ii) followed by a series of scallops (iii–vi). Arrow points to jugal notch. (B) Another fragment sporting a single episquamosal; (C) fragment scored on either side by a couple of deep vascular sulci, as in Anchiceratops ornatus, Arrhinoceratops brachyops, and other more derived chasmosaurines; (D) fragment of unknown origin.

faces. Similar features are observed running roughly rostro-

The thoracic vertebral column is gently bowed dorsally, with

caudally on both the dorsal and ventral surfaces of the frills

the neck and tail held out straight on either side of the body.

of other ceratopsids, but are especially pronounced in Anchi-

The forelimbs are thrown back under the chest and the hind-

ceratops ‘‘longirostris’’ (CMN 8535), Arrhinoceratops brachyops

limbs are strongly flexed so that the pedes point caudally.

(ROM 796), and other derived chasmosaurines.

Shortly after it was collected, the skeleton was prepared and displayed as a panel mount in what is now the Talisman Energy Fossil Gallery of the Canadian Museum of Nature, with-

POSTCRANIUM

the cast skull of CMN 8535 (holotype of Anchiceratops ‘‘longi-

The postcranium of CMN 8547 was discovered in the field

rostris’’) in place (Fig. 13.4B). Only the right side of the skele-

lying on its right side, with its left flank exposed (Fig. 13.4A).

ton is available for examination, hampering this and previous

Description of a Complete and Fully Articulated Chasmosaurine Postcranium 191

FIGURE 13.4.

Chasmosaurine skeleton CMN 8547. (A) Exposed in the field; (B) on display at the CMN. The skull is a cast of the type of Anchiceratops ‘‘longirostris’’ (Sternberg 1929).

attempts at providing a more exhaustive osteological descrip-

et al. 1907; Ostrom and Wellnhofer 1986) have defined the

tion. Selected measurements from the postcranial skeleton are

cervicals by the short, straight ribs they support. Most (e.g.,

listed in Table 13.1. A reconstruction is provided (Fig. 13.5) to

Brown 1917; Lull 1933; Brown and Schlaikjer 1942; Dodson et

correct for the minimal taphonomic distortion undergone in

al. 2004) have defined the cervicals as those vertebrae bear-

the postcranium.

ing parapophyses on their centra (rather than on their neural

Vertebral Column. The vertebral column is approximately

arches). We have chosen to use the latter definition both be-

405 cm long, measured across the top of the neural spines.

cause of its more common usage and because ribs are rarely

There are 74 vertebrae in all, close to the 77 reported in Cen-

preserved articulated with their supporting vertebrae. As such,

trosaurus apertus (Brown 1917; Lull 1933). Lull (1933) also re-

the cervical column of ceratopsids typically comprises a syn-

ported 76 vertebrae in Chasmosaurus belli, but no complete

cervical and 6 free cervicals. Some workers (e.g., Lull [in

column is known for this species (Mallon and Holmes 2006).

Hatcher et al.] 1907; Ostrom and Wellnhofer 1986; Dodson et

In general, the intervertebral spacing of CMN 8547 is approxi-

al. 2004), assuming the syncervical comprised 4 vertebrae,

mately 10–20 mm, except in the distal half of the tail where

have described 10 cervical vertebrae in ceratopsids; however,

the vertebrae are more tightly compressed.

more recent research has shown that the syncervical is com-

There has been some disagreement in the literature about

posed of 3 vertebrae (Campione and Holmes 2006; Tsuihiji

how to distinguish between the cervical and thoracic regions

and Makovicky 2007), effectively establishing the typical cera-

of the vertebral column. A minority of workers (e.g., Hatcher

topsid cervical count of 9. CMN 8547 is unique in having a

192 mallon & holmes

FIGURE 13.5. Reconstruction of the postcranium of CMN 8547. To correct for the effects of postburial distortion, the scapula was restored to approximate a 45\ angle from horizontal (Paul and Christiansen 2000; Thompson and Holmes 2007), the ribs were spaced farther apart, and the ischium was rotated ventrally.

Table 13.1. Selected Postcranial Measurements (in mm) for CMN 8547 and Comparable Ceratopsids (measurements of the appendicular skeleton follow the standards of Chinnery 2002 unless otherwise noted)

Parameter Length of syncervical

Chasmosaurinae

Centrosaurus

Chasmosaurus

Pentaceratops

indet.

apertus

belli

sternbergii

Styracosaurus albertensis

(CMN 8547)

(AMNH 5351)1

(CMN 2245)2

(PMU.R268)3

(CMN 344)

221

260

263

300

265

Length of presacral series

1780

—

1524

1700

—

Length of synsacral series

870

—

724

840

—

1400

—

—

1560

—

Length of scapula

560

700

653

730

800

Length of coracoid

135

210

184

250

—

Length of humerus

523

600

503

646

630

Length of caudal series

Minimum shaft width of humerus

91

—

Length of radius

296

350

—

70

—

81

434

370

Length of ulna

407

450

451

577

485

Length of ilium

940

1060

—

1120

1050

Length of prepubic process

304

450

292

—

—

Rectilinear length of ischium

556

—

569

—

830

Length of femur, measured along lateral surface

740

740

758 (est.)

880

810

Length of tibia, including astragalocalcaneum

542

600

587

655

610

Length of fibula, including astragalocalcaneum

549

560

504

600

565

1

Measurements from Brown (1917). Vertebral measurements from Sternberg (1927). 3Measurements from Lull (1933). 2

Description of a Complete and Fully Articulated Chasmosaurine Postcranium 193

FIGURE 13.6.

Appendicular skeleton of CMN 8547. (A) Pectoral girdle and forelimb. The single, identifiable sternal plate is marked by an asterisk. The unidentified elements in the area of the sternum are immediately cranial and caudal to this. (B) Pelvic girdle and hindlimb.

truly quadripartite syncervical and 6 free cervicals, for a total

lateral position to a lateral position distally, and attenuate by

of 10 cervical vertebrae—1 more than in any other known

the twenty-second caudal. The chevrons begin on the third

ceratopsid. Cervicals 5–9 are tall with erect neural spines that

caudal and disappear by the thirty-third caudal.

are subequal in height, as in Triceratops horridus (Hatcher et al. 1907).

Ribs. The quadripartite syncervical bears three pairs of small, v-shaped ribs. Ribs associated with the cranial free cervical

The thoracic vertebrae are defined here as those vertebrae

vertebrae are similar in size, although the ribs attached to cer-

bearing parapophyses on their neural arches. There are nor-

vicals 5–7 each bear a short, cranially projecting process be-

mally 12 thoracic vertebrae in ceratopsids (Dodson et al. 2004;

tween their articulating heads. The shafts abruptly become

Ostrom and Wellnhofer [1986] restored Triceratops horridus

longer in the ribs associated with the eighth cervical, and

with 14 thoracics, but 2 were added erroneously), but CMN

reach a length of at least 844 mm by the tenth presacral verte-

8547 has 13. The thoracic vertebrae are taller than the cer-

bra. Rib tubercula are reduced to low protuberances in those

vicals and their neural and transverse processes are swept

ribs associated with the eleventh and subsequent vertebrae,

caudally. Progressing caudally down the length of the col-

supporting the case for 10 cervical vertebrae. The thoracic ribs

umn, the zygapophyseal articulations rotate from a parasagit-

are swept caudally and are flattened mediolaterally, as in other

tal to a transverse orientation. The last thoracic may be co-

ceratopsids. Most unusual, however, is the thickness of these

ossified to the following vertebra as in Centrosaurus apertus

ribs, on average 39 mm wide craniocaudally at midlength.

(‘‘dorsosacral’’ of Lull 1933), but plaster obscures the articula-

This is in contrast to the widths of 37 mm in Chasmosaurus

tion with the next centrum in this specimen.

russelli (CMN 2280) and 27 mm in Styracosaurus albertensis

The synsacrum is relatively long, providing enough space

(CMN 344), the body masses both of which are restored as 1.25

for the 12 reconstructed neural spines of plausible size. Most

and 1.5 times greater than CMN 8547, respectively (Paul

ceratopsids have 10 synsacrals (Dodson et al. 2004), although

1997). The thickness of the ribs and their tight spacing give the

11 have been noted in Pentaceratops sternbergii (Wiman 1930).

impression of a rigid, compact ribcage (Paul 2000), but when

The synsacrum is braced dorsally by a tendon trellis that runs

viewed from the left side (Fig. 13.4A) it is evident that the ribs

across the neural arches, stretching from the penultimate

exposed on the panel mount have been compressed together,

thoracic vertebra to the second caudal vertebra. Unfortu-

effectively exaggerating the rigid appearance of the ribcage.

nately, much of the trellis has been prepared away, making

The ribs begin curving caudally by about the eighth thoracic

it difficult to compare with those of other ceratopsids (e.g.,

vertebra and the penultimate thoracic rib bows laterally to

Holmes and Organ 2007).

abut the medial surface of the ilium and the lateral surface of

Thirty-nine free caudal vertebrae comprise the tail, a count

the prepubic process distally. This condition may be a result

intermediate between that of Pentaceratops sternbergii (30; Wi-

of crushing, although the caudalmost thoracic ribs appear to

man 1930) and Centrosaurus apertus (46; Brown 1917), the

abut the ilium in Centrosaurus apertus as well (Brown 1917:

only other ceratopsids for which a complete tail is known. The

plate XI; Lull 1933: fig. 18). The last thoracic vertebra bears a

centra bear transverse processes that rotate from a ventro-

small set of ribs that project laterally and brace the ventral

194 mallon & holmes

FIGURE 13.7.

Comparison of ceratopsid humeri. (A) Centrosaurus apertus (after Lehman 1989); (B) Chasmosaurus russelli (CMN 2280); (C) Chasmosaurinae indet. (CMN 8547); (D) cf. Arrhinoceratops (ROM 1493); (E) Pentaceratops sternbergii (after Lehman 1989); (F) Triceratops horridus (after Hatcher et al. 1907).

surfaces of the ilia, much as in Centrosaurus apertus (Lull 1933:

pears to be due to post-mortem breakage because the right

fig. 18) and Chasmosaurus irvinensis (pers. obs.).

element preserves the tubercle. The same cannot be said of

Pectoral girdle. The massive scapula is angled 64\ from hori-

CMN 8547, as the surface of the humerus appears complete in

zontal, although this figure was probably closer to 45\ in life

this area. Field photos of the specimen show the same poorly

(Paul and Christiansen 2000; Thompson and Holmes 2007).

developed medial tubercle on the left humerus.

The scapula resembles that of larger chasmosaurines like Pen-

Although the radius resembles that of other chasmosau-

taceratops sternbergii and Triceratops horridus in outline. The

rines, the ulna does not. In most chasmosaurines, the lateral

scapular spine, more pronounced than in Chasmosaurus, runs

lip of the trochlear notch extends well beyond the ulnar shaft

obliquely across the lateral surface of the scapula, as in most

and forms a near right angle with the olecranon process in

other ceratopsids. The acromion process is virtually absent.

lateral view. In CMN 8547, the lip is not offset; rather, it lies

Both coracoids are preserved, but remain largely embedded

approximately in the same plane, as in centrosaurines (Fig.

in the matrix. The exposed parts are shaped very much like

13.8). Unlike in centrosaurines, however, the olecranon pro-

those of other ceratopsids and have a well-developed caudal

cess is especially pronounced (Table 13.3; Chinnery 2002,

‘‘hook-like’’ process.

2004).

A flattened, bean-shaped element with thickened cranial

As in other ceratopsids, the carpus preserves two round ele-

and caudal borders is located medial to the humerus, which

ments, presumably distal carpals 3 and 4. The individual ele-

we identify as a sternal plate (Fig. 13.6). One of two other

ments of the manus (Fig. 13.9) are robust compared to those of

bones adjacent to this might represent the other plate, but

Centrosaurus apertus, Chasmosaurus belli, or C. irvinensis, being

neither is a particularly convincing candidate. One, similarly

slightly shorter and relatively wider in CMN 8547 (Table 13.4).

flattened, is obscured almost entirely by the right forelimb;

The manus bears the typical ceratopsid phalangeal formula of

the other, lying outside the ribcage near the distal end of the

2-3-4-3-2. Digit III is the longest and the digits on either side of

humerus, is somewhat thicker and ‘‘L-shaped.’’ The latter is

it are increasingly reduced both medially and laterally. Large,

less closely comparable to the sternal plates of other cera-

hoof-like unguals are borne on digits I, II, and III.

topsids, although similar asymmetry is seen in the sternal

Pelvic Girdle. The robust pelvic girdle of this specimen (Fig.

plates of Centrosaurus apertus (Lull 1933: fig. 20) and is not un-

13.6B) agrees in most respects with that of other chasmo-

expected for sternal elements (Hildebrand 1994). Regardless,

saurines. The acetabulum is centered beneath the massive il-

it is likely that at least one of these bones is allochthonous,

ium, at a point below the fourth and fifth synsacral vertebrae.

although we cannot be certain which one.

The lateral overhang of the preacetabular process of the ilium

Forelimb. The forelimb of CMN 8547 is stout relative to that

is deflected strongly ventrally so that its edge is continuous

of other ceratopsids (Table 13.2). The humerus approaches

with the downturned ‘‘lip’’ just caudal the acetabulum. It is

that of most ceratopsids in length (Table 13.1), but it is much

here that the ilium is mediolaterally widest. The prepubic pro-

more robust in other respects. The deltopectoral crest is espe-

cess of the pubis projects cranially into the gut cavity, contrary

cially prominent and most resembles that of the larger Tri-

to the reconstruction of Centrosaurus apertus by Brown (1917)

ceratops horridus, extending 80 mm laterally from the humeral

in which it projects cranioventrally. The process is dorsoven-

shaft. The minimum shaft width exceeds that of comparably

trally expanded at its cranial end and its lateral surface articu-

sized ceratopsids as well (Table 13.1). The humerus is unique

lates with the penultimate thoracic rib (i.e., that associated

among ceratopsids in lacking a distinct medial tubercle (Fig.

with the twelfth thoracic vertebra), rather than the last rib as

13.7). One specimen of Chasmosaurus belli, CMN 2245, also

in Agujaceratops mariscalensis (Lehman 1989). The ventrally

lacks a medial tubercle on its left humerus, although this ap-

curved ischium is more robust than in other ceratopsids. In

Description of a Complete and Fully Articulated Chasmosaurine Postcranium 195

Table 13.2. Vertebral Column and Limb Measurements for CMN 8547 and Selected Ceratopsids (forelimb length = humerus + radius + McIII; hindlimb length = femur + tibia + Mt III. (est.) signifies estimated values)

Specimen

Vertebral

Forelimb

Hindlimb

Forelimb/

Hindlimb/

column length

length

length

Vertebral

Vertebral

(mm)

(mm)

(mm)

column

column

Forelimb/Hindlimb

Chasmosaurinae indet. (CMN 8547)

4,050

946

1,435

0.234

0.3543

0.659

Centrosaurus apertus (AMNH 5351)1

4,460

1,080

1,555

0.2422

0.3487

0.6945

Chasmosaurus belli (CMN 2245)2

4,115 (est.)

1,060

1,676

0.2576 (est.)

0.4073 (est.)

0.6325

Pentaceratops sternbergii (PMU.R268)3

4,100

1,080

1,535

0.2634

0.3744

0.7036

Triceratops horridus (NSM PV 20379)

5,300 (est.)

1,214

1,881

0.2291 (est.)

0.3549 (est.)

0.6454

1

Measurements from Brown (1917). Measurements from Sternberg (1927). 3 Measurements from Lull (1933). 2

Table 13.3. Antebrachial Measurements for CMN 8547 and Other Selected Ceratopsids (olecranon length = length from sigmoid fossa rim to proximal end of olecranon (Chinnery 2002); despite its small size, the forelimb of CMN 8547 possesses a large mechanical advantage)

Taxon

Olecranon length

Length of radius

Index

(mm)

(mm)

(O/R*100) 59.66

Chasmosaurinae indet. (CMN 8547)

174.4

292.3

Styracosaurus albertensis (CMN 341)

146

380

38.4

Chasmosaurus belli (ROM 839)

105

337.5

31.1

Triceratops horridus (AMNH 5880)

225.8

492

45.9

Chasmosaurus belli, Agujaceratops mariscalensis, and Pentacera-

The pes is comprised of four functional digits I–IV. The indi-

tops sternbergii, the ischium remains relatively slender along

vidual elements are comparatively robust (Table 13.5), being

most of its length, expanding slightly at the distal end. In

both shorter and relatively wider than in Styracosaurus alber-

CMN 8547, the ischium is dorsoventrally thickest at mid-

tensis (TMP 1989.97.1). Digit III is the longest and the digits on

length, and tapers distally.

either side of it become increasingly reduced both medially

Hindlimb. The hindlimb (Fig. 13.6B) of CMN 8547 is short

and laterally. The lengths of the individual digits correlate

relative to the length of the vertebral column, but not to the

with the lengths of their metatarsals rather than with the

extent seen in Centrosaurus apertus (Table 13.2). The femur

number of their constituent phalanges. The metatarsals are

approximates that of other ceratopsids in length (Table 13.1),

approximately hourglass-shaped. On the left pes, a small, tri-

but the shaft is relatively thicker in the sagittal plane. The

angular splint of bone is evident on the lateral surface of meta-

fourth trochanter is only weakly developed, as is typical of

tarsal IV which we interpret as a vestigial metatarsal V. The pes

ceratopsids. Unfortunately, because of the nature of the panel

bears the typical ceratopsid phalangeal formula of 2-3-4-5-0.

mount, the medial half of the femoral shaft is not exposed.

All functional digits terminate in a large, hoof-like ungual.

The tibia and fibula are preserved close together, with the proximal end of the fibula articulating with the proximocaudal end of the tibia, and the distal end of the fibula articulating

Discussion

with the distocranial end of the tibia. Their proportions are

TAXONOMIC AFFINITIES OF CMN 8547

similar to those of other large ceratopsids (Table 13.1). The astragalocalcaneum is firmly attached to the distal end of the

To date, four ceratopsid species have been described from the

tibia and spans the entire mediolateral width of the element.

Horseshoe Canyon Formation, based almost exclusively on

The astragalocalcaneum adds approximately 4% to the length

cranial material. Pachyrhinosaurus canadensis, the only cen-

of the crus.

trosaurine ceratopsid present, is found in unit 1 of the forma-

196 mallon & holmes

FIGURE 13.8.

Comparison of ceratopsid ulnae. (A) Centrosaurus apertus (after Lehman 1989); (B) Chasmosaurus belli (CMN 2245); (C) Chasmosaurinae indet. (CMN 8547); (D) Pentaceratops sternbergii (after Lehman 1989); (E) Triceratops horridus (after Hatcher et al. 1907).

FIGURE 13.9.

Reconstructed left manus of CMN 8547 based on field photographs and panel mount.

tion. Chasmosaurine ceratopsids are more widely distributed

(e.g., Langston 1959; Russell and Chamney 1967; Russell

throughout the formation. Anchiceratops ornatus and Arrhino-

1977; Dodson 1996; Glut 1997; Paul and Christiansen 2000;

ceratops brachyops have been recovered from units 1 and 2

Dodson et al. 2004; Senter 2007).

(D. Eberth pers. com.), and Eotriceratops xerinsularis was recently discovered in the middle of unit 5 (Wu et al. 2007).

The elongate squamosal, the ventral curvature of the ilium, and the robust postcranial skeleton of CMN 8547 are all chas-

In his 1925 field notes, Sternberg attributed the skeleton of

mosaurine characters (Chinnery 2002, 2004). The frill frag-

CMN 8547 to ‘‘probably a small Anchiceratops,’’ but did not

ments suggest a morphology distinct from that of Eotriceratops

give reasons for this assignment. Lull (1933) later endorsed

(Wu et al. 2007), but comparable with those of both Anchicera-

Sternberg’s identification and asserted that ‘‘[t]here were . . . a

tops and Arrhinoceratops. The provenance of CMN 8547 from

few diagnostic fragments of the posterior portion of the crest

unit 2 of the Horseshoe Canyon Formation further supports

which determined the genus and probably the species.’’ Al-

either of the latter two identifications. Unfortunately, despite

though no further information was provided, the assignment

Lull’s (1933) endorsement of Sternberg’s original classifica-

of CMN 8547 to Anchiceratops has since gone unquestioned

tion, none of the remaining skull fragments are diagnostic to

Description of a Complete and Fully Articulated Chasmosaurine Postcranium 197

Table 13.4. Lengths of Individual Left Manus Elements from CMN 8547 (in mm) (estimates [est.] were made from old field photographs)

The single provocative exception (ROM 1493) is a specimen previously assigned to both Arrhinoceratops (Tyson 1981) and Torosaurus (Farke 2007) that comprises a skull and associated forelimb elements, including a partial humerus. The humerus

Digit

Digit

Digit

Digit

Digit

I

II

III

IV

V

Metacarpal

73

119

127

93

76

CMN 8547 may, in fact, belong to either of these two genera,

Phalanx 1

51

37

36

42

41

but more conclusive evidence is needed.

Phalanx 2

48 (est.)

25

20

19

18

50 (est.)

15

12

Phalanx 3 Phalanx 4 Total digit length

231 (est.)

238 (est.)

166

parable to that of CMN 8547 (Fig. 13.7). This suggests that

The paucity of skull material attributable to CMN 8547 therefore renders its assignment to either of the two known

40 (est.) 172 (est.)

is badly weathered but bears a large deltopectoral crest com-

135

lower Horseshoe Canyon Formation chasmosaurines nearly impossible. In fact, it is conceivable that Sternberg assigned the skeleton to Anchiceratops, not because of any particularly diagnostic characters, but because it may have been the only Edmontonian ceratopsid known to him at the time. Accord-

Table 13.5. Lengths of Individual Right Pes Elements from CMN 8547 (in mm) (metatarsal V was measured from the left pes)

ing to the scant correspondence on file at the University of Toronto between Parks and W. A. Wallace, a librarian at the university, the article describing Arrhinoceratops (Parks 1925)

Digit

Digit

Digit

Digit

Digit

was not accepted for publication until March 31, 1925, and

I

II

III

IV

V

likely was not published at least until the summer of that same

Metatarsal

84

174

175

138

18

Phalanx 1

91

60

58

53

Phalanx 2

79

36

23

29

new genus, despite a brief announcement of the animal made

80

21

22

the year prior at the annual meeting of the Geological Society

74

13

of America (Parks 1924). None of the correspondence between

351

317

Phalanx 3 Phalanx 4 Phalanx 5 Total digit length

350

tian beds of Alberta and may not have been aware of Parks’s

Sternberg and Parks on file at either the CMN or the ROM

62 254

year. At this time, Sternberg was collecting in the Edmonton-

18

suggests that Sternberg knew of Arrhinoceratops before it was named in print. The centrosaurine Pachyrhinosaurus was not named from the formation until 25 years later by Sternberg. Therefore, it may be that he assigned the skeleton of CMN

the genus level. Parks (1925) argued that the thinness of the

8547 to Anchiceratops simply by default, unaware of any other

frill of Arrhinoceratops (5–10 mm in most areas), and the pres-

ceratopsids from the lower Edmonton Group.

ence of vascular sulci on both the dorsal and ventral surfaces

Despite the difficulties in associating CMN 8547 with any

of the frill, could be used to diagnose the genus (a sentiment

known ceratopsid taxon, it is clearly distinct from any de-

echoed by Dodson 1996). However, these same features are

scribed postcranial skeleton in the robust proportions of many

observed in the type skull of Anchiceratops ‘‘longirostris’’ (CMN

of the limb and girdle elements, and particularly in its unique

8535), and therefore cannot be used to distinguish between

vertebral formula. The presacral column comprises 10 cervi-

these taxa (Sternberg 1929). The size and shape of those squa-

cals and 13 dorsals, for a total of 23 vertebrae, or two more

mosal epoccipitals nearest the jugal notch are similarly indis-

than in other ceratopsids. The synsacrum contains 12 fused

tinguishable between Anchiceratops and Arrhinoceratops, and

vertebrae, two more than the similar-sized Centrosaurus apertus

cannot be used to narrow the affinities of CMN 8547. In fact,

(Lull 1933: fig. 18). The caudal series consists of 39 vertebrae,

as currently diagnosed, very few cranial characters separate

seven less than in Centrosaurus apertus (Brown 1917). This sug-

Arrhinoceratops from Anchiceratops, and many of the autapo-

gests that the unusual count in CMN 8547 may be the result of

morphies ascribed to the former may simply be due to sexual

a caudal shift in the sacrum during development, effectively

or individual variation, as has been argued for other cera-

releasing two vertebrae to the presacral column, and capturing

topsid taxa (e.g., Ostrom and Wellnhofer 1986; Dodson 1990;

at least four proximal caudals, resulting in a longer sacrum and

Ostrom and Wellnhofer 1990; Lehman 1990; Godfrey and

shorter tail (a similar pattern, albeit less exaggerated, may also

Holmes 1995). We are currently redescribing these genera in

account for the reduced tail and increased presacral count in

an effort to amend their diagnoses.

Pentaceratops sternbergii ). Whether this vertebral formula is

The identification of CMN 8547 is further hampered by the

consistent within the taxon (whatever its identity), or whether

fact that very few ceratopsid specimens from the Horseshoe

it is unique to this individual, cannot be determined until

Canyon Formation are associated with postcranial material.

more articulated skeletons are collected.

198 mallon & holmes

SEMI-AQUATIC CERATOPSIDS

in terrestrial animals. Thus, independently, each feature offers little support for semi-aquatic habits, but together they may

Based on depositional environment and field observations,

substantiate more convincing conclusions (Mitchell and Ted-

Langston (1959) argued that Anchiceratops was common only

ford 1973).

where other ceratopsians were not, and surmised that it was

Most of the adaptations listed in Table 13.6 are present in

restricted to ‘‘low-lying, even marshy habitats where reduc-

CMN 8547. Despite the specimen’s small size, the deltopec-

ing phenomena predominated’’ (but see Russell and Cham-

toral crest and olecranon process are developed well beyond

ney 1967). Anticipating further description of the postcranial

those of similar-sized ceratopsids. Limb length is likewise un-

skeleton of CMN 8547, and assuming that it pertained to An-

derdeveloped. It is unlikely that these features would be so

chiceratops, Langston noted several adaptations of the species

strongly expressed simply to facilitate terrestrial locomotion.

that may have permitted a semi-aquatic lifestyle. The elongate

We suggest that the inferred strong pectoral and brachial mus-

face was hypothesized to have permitted breathing while the

culature, together with the animal’s stout proportions, may

animal was wading in shallow water, counterbalanced by the

have allowed CMN 8547 to maneuver through the muddy,

weight of the heavy frill. The short tail, massive body, and

lowland swamps of the early Maastrichtian, as Langston

stocky limbs were also said to ‘‘bespeak a sluggish nature that

(1959) first suggested. This interpretation is corroborated by

might be expected in animals that enjoyed the relative se-

the fact that the specimen was found in the highest estuarine

clusion and protection of a swampy environment’’ (Langston

facies of the Drumheller Marine Tongue (Horseshoe Canyon

1959: 10).

Formation; D. A. Eberth pers. com.), in association with ‘‘a

The idea that Anchiceratops might have been semi-aquatic

few’’ oysters (C. M. Sternberg field notes 1925, CMN). The

hearkens back to earlier suggestions that ceratopsians were

implication by Tereschenko (2008) that the tails of ceratopsids

adapted to life in the water (W. D. Matthew quoted in Hatcher

might also exhibit adaptations for a semi-aquatic mode of life

et al. 1907; Matthew 1915; Gregory and Mook 1925; Feduccia

(e.g., heterocoelous proximal caudal vertebrae and relatively

1973; Russell 1977). While this interpretation has long since

tall neural spines of the mid-caudals for sculling through wa-

fallen out of favor (Colbert 1948; Bakker 1968, 1986; Coombs

ter) is not supported with reference to the anatomy of CMN

and Deméré 1996; Paul and Christiansen 2000), it is worth

8547.

briefly revisiting in light of recent work suggesting some cer-

It is possible that other ceratopsids were also adapted to

atopsids favored wetland environments (Brinkman 1990;

semi-aquatic life, as the clade shares several features in com-

Eberth and Brinkman 1997; Retallack 1997; Brinkman et al.

mon with the hippo. It is important to limit speculation via

1998; Brinkman et al. 2004; Bykowski and Retallack 2007;

reference to circumstantial sedimentology, however, as sev-

Lehman 2007; Tereschenko 2008; Eberth this volume).

eral ceratopsids are known from well-drained alluvial settings

The most appropriate surviving outgroup of reptiles with

(Eberth this volume).

which to address the likelihood of semi-aquatic ceratopsids is probably the Crocodylia. Unfortunately, crocodilians are highly modified in their own right for an axial subundulatory

SIGNIFICANCE OF CMN 8547

mode of swimming including an elongate, streamlined body

With its robust build, unique vertebral count, and distinc-

(Massare 1988), laterally compressed tail (Bakker 1986), sec-

tive forelimb morphology, the postcranium of CMN 8547 is

ondarily sprawling limbs (Parrish 1987), and webbed feet

unique among ceratopsids. This is noteworthy because Chin-

(Pooley and Gans 1976). With a few rare exceptions croco-

nery (2002, 2004) previously demonstrated that the post-

dilians also do not attain body sizes comparable to those of

crania of most ceratopsids do not exhibit diagnostic characters

ceratopsids (Ross and Magnusson 1989). For these reasons, a

at the species level. Assuming that it is not simply anomalous,

more suitable living analogue might be the mediportal (sensu

CMN 8547 is a likely exception, although the specimen can-

Coombs 1978) common hippo (Hippopotamus amphibious). It

not be positively assigned to any of the known chasmosau-

can be found wading in the rivers and estuaries of sub-Saharan

rines from the Horseshoe Canyon Formation because of a lack

Africa, venturing onto land only at night to forage (Owen-

of sufficient overlapping material and because of the diffi-

Smith 1988). To be sure, dinosaurs are not mammals (Hotton

culties with distinguishing between Anchiceratops and Arrhino-

1994); ceratopsids and hippos differ in many significant ways.

ceratops as currently diagnosed. A careful rediagnosis of these

Yet the role of homoplasy in the evolution of functional mor-

species, a more precise assessment of their biostratigraphy, and

phology cannot be discounted, and several authors have iden-

the discovery of more fossil material should help to resolve

tified a number of semi-aquatic adaptations in the hippo

some of these issues. In the meantime, cladistic analyses cod-

postcranium that can be compared with CMN 8547 (Table

ing postcranial characters for Anchiceratops on the basis of this

13.6). Caution should be taken when relating these features to

single specimen (e.g., Dodson and Currie 1990; Dodson et al.

ceratopsids, however, as many of them have parallel functions

2004) should be regarded with added caution.

Description of a Complete and Fully Articulated Chasmosaurine Postcranium 199

Table 13.6. Possible Osteological Adaptations for Semi-aquatic Habits in Ceratopsids with Reference to the Common Hippo (adaptations marked by an asterisk are disputed by Mihlbachler et al. 2004)

Osteological adaptation

Presence in

Presence in

CMN 8547

other ceratopsids

Function in aquatic forms

Function in terrestrial forms

Thickened cortex of limb bones1

?

?

Reduced buoyancy

Increased structural support

Dorsal placement of nares2,3

?

No

Facilitate breathing while

—

submerged Dorsal placement of orbits2,3

?

Yes5

Facilitate vision above water

—

while submerged Reduced olfaction3 Reduced lacrimal (or lacrimal

?

No6,7,8

Unnecessary

—

Yes

Yes

Unnecessary

—

Yes

No9

Weight of head supported by

—

opening)3 Reduced anterior thoracic neural spines

water rather than neck muscles

3

Expansive ribcage3

Yes

Yes

Enlarged gut supported by water

Enhanced hindgut fermentation

Large olecranon relative to length

Yes

Yes

Mechanical advantage for

Scratch-digging

of radius*,3

walking in mud

Well-developed pectoral muscles4 Retention of long digital flexor

Yes

Yes10,11

Swimming

Support semi-erect posture

?

?

Locomotion through soft

—

tendons4

substrate

Shortened limbs*,2,3

Yes

Yes

Mechanical advantage for

Facilitate grazing

walking in mud 1

7

2

8

Wall 1983. Howell 1930. 3Wall and Heinbaugh 1999. 4Fisher et al. 2007. 5Bykowski and Retallack 2007. 6Brown 1914.

Lull 1933. Witmer 2001. 9Paul and Christiansen 2000. 10Dodson 1996. 11Adams 1991.

The postcranial morphology of CMN 8547 also lends qualitative support to the idea that at least some ceratopsids might have been semi-aquatic, as suggested by an increasing body of sedimentological, microsite, and palaeosol data. Acknowledgments

We thank K. Shepherd and M. Feuerstack (CMN) for providing access to specimens in their care. A. Gerdung allowed us to revisit the quarry on his property, and D. Tanke served as our guide. D. Eberth kindly provided stratigraphic context. S.-I. Fujiwara provided measurements for NSM-PV 20379. R. Sakowski of the University of Toronto library provided historical information concerning Parks’s description of Arrhinoceratops. M. Lipman provided photographs of CMN 8547, and R. Cuthbertson photographed ROM 1493. B. ChinneryAllgeier and A. Farke kindly gave constructive criticism of the manuscript. Finally, thanks to the editors of this volume for their tireless efforts in putting it together and for soliciting our contribution. References Cited Adams, D. A. 1991. The significance of sternal position and orientation to reconstruction of ceratopsid stance and appearance. Journal of Vertebrate Paleontology 11(3, Suppl.): 14A.

200 mallon & holmes

Bakker, R. T. 1968. The superiority of dinosaurs. Discovery 3: 11–22. ———. 1986. The Dinosaur Heresies. New York: Kensington Publishing Corp. Brinkman, D. B. 1990. Paleoecology of the Judith River Formation (Campanian) of Dinosaur Provincial Park, Alberta, Canada: Evidence from vertebrate microfossil localities. Palaeogeography, Palaeoclimatology, Palaeoecology 78: 37–54. Brinkman, D. B., A. P. Russell, D. A. Eberth, and J. Peng. 2004. Vertebrate palaeocommunities of the lower Judith River Group (Campanian) of southeastern Alberta, Canada, as interpreted from vertebrate microfossil assemblages. Palaeogeography, Palaeoclimatology, Palaeoecology 213: 295–313. Brinkman, D. B., M. J. Ryan, and D. A. Eberth. 1998. The palaeogeographic and stratigraphic distribution of ceratopsids (Ornithischia) in the Upper Judith River Group of Western Canada. Palaios 13: 160–169. Brown, B. 1914. Anchiceratops, a new genus of horned dinosaurs from the Edmonton Cretaceous of Alberta. With discussion of the origin of the ceratopsian crest and the brain casts of Anchiceratops and Trachodon. Bulletin of the American Museum of Natural History 33: 539–548. ———. 1917. A complete skeleton of the horned dinosaur Monoclonius, and description of a second skeleton showing skin impressions. Bulletin of the American Museum of Natural History 37: 281–306. Brown, B., and E. M. Schlaikjer. 1942. The skeleton of Leptoceratops

with the description of a new species. American Museum Novitates 1169: 1–15. Bykowski, R., and G. Retallack. 2007. Was Triceratops like a bison, rhino or hippo? Implications for lifestyle and habitat. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts and Programs, pp. 11–16. Drumheller: Royal Tyrrell Museum of Palaeontology. Campione, N. E., and R. B. Holmes. 2006. The anatomy and homologies of the ceratopsid syncervical. Journal of Vertebrate Paleontology 26: 1014–1017. Chinnery, B. J. 2002. Morphometric analysis of evolution and growth in the ceratopsian postcranial skeleton. Ph.D. diss., Johns Hopkins University School of Medicine, Baltimore. ———. 2004. Morphometric analysis of evolutionary trends in the ceratopsian postcranial skeleton. Journal of Vertebrate Paleontology 24: 591–609. Colbert, E. H. 1948. Evolution of the horned dinosaurs. Evolution 2: 145–163. Coombs, W. P., Jr. 1978. Theoretical aspects of cursorial adaptations in dinosaurs. Quarterly Review of Biology 53: 393–418. Coombs, W. P., Jr., and T. A. Deméré. 1996. A Late Cretaceous nodosaurid ankylosaur (Dinosauria: Ornithischia) from marine sediments of coastal California. Journal of Paleontology 70: 311– 326. Dodson, P. 1990. On the status of the ceratopsids Monoclonius and Centrosaurus. In K. Carpenter, and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 231–243. Cambridge: Cambridge University Press. ———. 1996. The Horned Dinosaurs: A Natural History. Princeton: Princeton University Press. Dodson, P., and P. J. Currie. 1990. Neoceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 593–618. Berkeley: University of California Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A. 2010. A review of ceratopsian paleoenvironmental associations and taphonomy. In M. J. Ryan, B. J. ChinneryAllgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 428– 446. Bloomington: Indiana University Press. Eberth, D. A., and D. B. Brinkman. 1997. Paleoecology of an estuarine, incised-valley fill in the Dinosaur Park Formation ( Judith River Group, upper Cretaceous) of southern Alberta, Canada. Palaios 12: 43–58. Farke, A. A. 2007. Cranial osteology and phylogenetic relationships of the chasmosaurine ceratopsid Torosaurus latus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 235–257. Bloomington: Indiana University Press. Feduccia, A. 1973. Dinosaurs as reptiles. Evolution 27: 166–169. Fisher, R. E., K. M. Scott, and V. L. Naples. 2007. Forelimb myology of the pygmy hippopotamus (Choeropsis liberiensis). Anatomical Record 290: 673–693. Gibson, D. W. 1977. Upper Cretaceous and Tertiary coal-bearing

strata in the Drumheller-Ardley region, Red Deer River Valley, Alberta. Geological Survey of Canada, Paper 76–35: 1–41. Glut, D. F. 1997. Dinosaurs: The Encyclopedia. Jefferson: McFarland & Company, Inc. Godfrey, S. J., and R. Holmes. 1995. Cranial morphology and systematics of Chasmosaurus (Dinosauria: Ceratopsidae) from the Upper Cretaceous of western Canada. Journal of Vertebrate Paleontology 15: 726–742. Gregory, W. K, and C. C. Mook. 1925. On Protoceratops, a primitive ceratopsian dinosaur from the Lower Cretaceous of Mongolia. American Museum Novitates 156: 1–9. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49: 1–300. Hildebrand, M. 1994. Analysis of Vertebrate Structure, 4th ed., New York: John Wiley & Sons. Holmes, R., and C. Organ. 2007. An ossified tendon trellis in Chasmosaurus (Ornithischia: Ceratopsidae). Journal of Paleontology 81(2): 411–414. Horner, J. R., and M. B. Goodwin. 2008. Ontogeny of cranial epiossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Hotton, N., III. 1994. Why dinosaurs were not mammals and vice versa. In G. D. Rosenberg and D. L. Wolberg, eds., Dinofest: Proceedings of a Conference for the General Public, pp. 39–59. Paleontological Society Special Publication No. 7. Howell, A. B. 1930. Aquatic Mammals. Springfield: Charles C. Thomas. Lambe, L. M. 1915. On Eoceratops canadensis, gen. nov., with remarks on other genera of Cretaceous horned dinosaurs. Canada Geological Survey Museum Bulletin 12: 1–49. Langston, W., Jr. 1959. Anchiceratops from the Oldman Formation of Alberta. National Museum of Canada Natural History Papers 3: 1–11. Lehman, T. M. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. ———. 1990. The ceratopsian subfamily Chasmosaurinae: Sexual dimorphism and systematics. In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 211–229. Cambridge: Cambridge University Press. ———. 2007. Growth and population age structure in the horned dinosaur Chasmosaurus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 259–317. Bloomington: Indiana University Press. Lull, R. S. 1933. A revision of the Ceratopsia or horned dinosaurs. Peabody Museum of Natural History Memoirs 3: 1–175. Mallon, J. C., and R. Holmes. 2006. A reevaluation of sexual dimorphism in the postcranium of the chasmosaurine ceratopsid Chasmosaurus belli (Dinosauria: Ornithischia). Canadian FieldNaturalist 120: 403–412. Marsh, O. C. 1888. A new family of horned Dinosauria, from the Cretaceous. American Journal of Science 36: 477–478. ———. 1890. Additional characters of the Ceratopsidae, with notice of new Cretaceous dinosaurs. American Journal of Science 39: 418–429 Massare, J. A. 1988. Swimming capabilities of Mesozoic marine

Description of a Complete and Fully Articulated Chasmosaurine Postcranium 201

reptiles: Implications for method of predation. Paleobiology 14: 187–205. Matthew, W. D. 1915. Climate and evolution. Annals of the New York Academy of Sciences 24: 171–318. Mihlbachler, M. C., S. G. Lucas, R. J. Emry, and B. Bayshashov. 2004. A new brontothere (Brontotheriidae, Perissodactyla, Mammalia) from the Eocene of the Ily Basin of Kazakstan and a phylogeny of Asian ‘‘horned’’ brontotheres. American Museum Novitates 3439: 1–43. Mitchell, E., and R. H. Tedford. 1973. The Enaliarctinae, a new group of extinct aquatic Carnivora and a consideration of the origin of the Otariidae. Bulletin of the American Museum Natural History 151: 201–284. Ostrom, J. H., and P. Wellnhofer. 1986. The Munich specimen of Triceratops with a revision of the genus. Zitteliana 14: 111–158. ———. 1990. Triceratops, an example of flawed systematics.In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 245–254. Cambridge: Cambridge University Press. Owen-Smith, N. 1988. Megaherbivores. Cambridge: Cambridge University Press. Parks, W. A. 1924. A new genus and species of horned dinosaur from the Cretaceous of Alberta. Geological Society of America Preliminary Lists 1924: 38. ———. 1925. Arrhinoceratops brachyops a new genus and species of Ceratopsia from the Edmonton Formation of Alberta. University of Toronto Geological Series 19: 5–15. Parrish, M. J. 1987. The origin of crocodilian locomotion. Paleobiology 13: 396–414. Paul, G. S. 1997. Dinosaur models: The good, the bad, and using them to estimate the mass of dinosaurs. In D. L. Wolberg, E. Stump, and G. Rosenberg, eds., Dinofest International: Proceedings of a Symposium Held at Arizona State University, pp. 129–154. Philadelphia: Academy of Natural Sciences. ———. 2000. Restoring the life appearances of dinosaurs. In G. S. Paul, ed., The Scientific American Book of Dinosaurs, pp. 78–106. New York: St. Martin’s Press. Paul, G. S., and P. Christiansen. 2000. Forelimb posture in neoceratopsian dinosaurs: Implications for gait and locomotion. Paleobiology 26: 450–465. Pooley, A. G., and C. Gans. 1976. The Nile crocodile. Scientific American 234: 114–124. Retallack, G. J. 1997. Dinosaurs and dirt. In D. L. Wolberg, E. Stump, and G. Rosenberg, eds., Dinofest International: Proceedings of a Symposium Held at Arizona State University, pp. 34–359. Philadelphia: Academy of Natural Sciences. Ross, C. A., and W. E. Magnusson. 1989. Living crocodilians. In

202 mallon & holmes

C. A. Ross and S. Garnett, eds., Crocodiles and Alligators, pp. 58– 73. New York: Facts on File, Inc. Russell, D. A. 1977. A Vanished World: The Dinosaurs of Western Canada. Ottawa: National Museums of Canada. Russell, D. A., and T. P. Chamney. 1967. Notes on the biostratigraphy of dinosaurian and microfossil faunas in the Edmonton Formation (Cretaceous), Alberta. National Museum of Canada Natural History Papers 35: 1–22. Seeley, H. G. 1888. On the classification of the fossil animals commonly named Dinosauria. Proceedings of the Royal Society of London 43: 165–171. Senter, P. 2007. Analysis of forelimb function in basal ceratopsians. Journal of Zoology 273: 305–314. Sternberg, C. M. 1927. Horned dinosaur group in the National Museum of Canada. Canadian Field-Naturalist 41: 67–73. ———. 1929. A new species of horned dinosaur from the Upper Cretaceous of Alberta. National Museum of Canada Bulletin 54: 34– 37. Tereschenko, V. S. 2008. Adaptive features of protoceratopsoids (Ornithischia: Neoceratopsia). Paleontological Journal 42: 273– 286. Thompson, S., and R. Holmes. 2007. Forelimb stance and step cycle in Chasmosaurus irvinensis (Dinosauria: Neoceratopsia). Palaeontologia Electronica 10: 5A: 1–17. Tsuihiji, T., and P. J. Makovicky. 2007. Homology of the neoceratopsian cervical bar elements. Journal of Paleontology 81: 1132–1138. Tyson, H. 1981. The structure and relationships of the horned dinosaur Arrhinoceratops Parks (Ornithischia: Ceratopsidae). Canadian Journal of Earth Sciences 18: 1241–1247. Wall, W. P. 1983. The correlation between high limb-bone density and aquatic habits in Recent mammals. Journal of Paleontology 57: 197–207. Wall, W. P., and K. L. Heinbaugh. 1999. Locomotor adaptations in Metamynodon planifrons compared to other amynodontids (Perissodactyla, Rhinocerotoidea). National Park Service Paleontological Research 4: 8–17. Wiman, C. 1930. Über Ceratopsia aus der oberen Kreide in New Mexico. Nova Acta Regiae Societatis Scientiarum Upsaliensis, Series 4, 7: 1–19. Witmer, L. M. 2001. Nostril position in dinosaurs and other vertebrates and its significance for nasal function. Science 293: 850– 853. Wu, X.-C., D. B. Brinkman, D. A. Eberth, and D. R. Braman. 2007. A new ceratopsid dinosaur (Ornithischia) from the uppermost Horseshoe Canyon Formation (upper Maastrichtian), Alberta, Canada. Canadian Journal of Earth Sciences 44: 1243–1265.

14 A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation, Northwest South Dakota, United States: A Preliminary Description CHRISTOPHER J. OTT AND PETER L. LARSON

a fragmentary skull and skeleton of a small cera-

variation of Triceratops (Goodwin et al. 2006). Other work

topsian dinosaur from the Hell Creek Formation of

on neoceratopsian ontogeny has demonstrated that certain

South Dakota represents a new taxon of horned dino-

characteristics, such as development of ornamentation, adult

saur. This specimen is estimated to have a skull length of

bone surface texture, fusion of sutures in cranial and post-

approximately 1 m and an approximate total length of 3

cranial elements, and horncore orientation, are consistent in-

m. All preserved elements of the skeleton show charac-

dicators of adult status (Lehman 1990; Sampson et al. 1997;

teristic adult features such as rugose bone surface textur-

Goodwin et al. 2006).

ing, well-developed dendritic veination patterns, and

A small, fragmented, partial skeleton of a ceratopsid (BHI

fusion of sutural contacts to the point of obscurity. The

6226), consisting of at least 47 identifiable skeletal elements

adult status, diminutive size, and significant mor-

and over 2,000 fragments, was collected from exposures of

phological differences between the new animal and con-

the Hell Creek Formation near Buffalo, South Dakota, begin-

temporaneous ceratopsians support the designation of a

ning in 1997. BHI 6226 is similar in size to, and shares some

new taxon for this specimen.

characteristics with, the enigmatic Judithian centrosaurine Avaceratops, but is clearly separated temporally and taxonomi-

Introduction

cally. This specimen has an estimated skull length of approximately 1 m, but shows fully developed adult characteristics.

The neoceratopsian dinosaur fauna of the Hell Creek Forma-

With an approximate skull length of 1 m, the estimated total

tion is currently represented by four taxa: Triceratops horridus

length of BHI-6226 would have been 3–3.5 m, based upon the

(Marsh 1889; Forster 1996b), Triceratops prorsus (Marsh 1890;

proportions of Triceratops horridus (adult skull length approxi-

Forster 1996b), Torosaurus latus (Marsh 1891), and the basal

mately 2 m). Based on the adult features of BHI-6226 and

neoceratopsian Leptoceratops gracilis (Brown 1914; Ott 2006).

morphological differences from other Hell Creek ceratopsians,

The Hell Creek Formation is one of the most intensively col-

a new taxon, Tatankaceratops sacrisonorum, is proposed for this

lected and studied dinosaur-bearing formations in the world,

animal (Fig. 14.1).

and the neoceratopsian fauna within it has been studied for

Institutional Abbreviations. ANSP: Academy of Natural Sci-

over 100 years, though significant work remains. Recent work

ences, Philadelphia; BHI: Black Hills Institute of Geologic Re-

on the advanced neoceratopsian taxa includes description

search, Hill City; LACM: Los Angeles County Museum, Los

of very young individuals and discussion of the ontogenetic

Angeles; MOR: Museum of the Rockies, Bozeman; ROM: Royal

203

Holotype skull of Tatankaceratops sacrisonorum (BI-6226). (A) Known cranial material; (B) reconstruction of the skull of Tatankaceratops sacrisonorum (BHI-6226) superimposed on the reconstructed skull of Triceratops horridus (TCM 2001.93.1; in grey). Scale bar is 10 cm. FIGURE 14.1.

Ontario Museum, Toronto; SMNH: Royal Saskatchewan Mu-

and honor the Lakota Sioux Tribe, who are the prior inhabi-

seum, Regina; TCM: Children’s Museum, Indianapolis;

tants of the area. ‘‘ceratops’’: (Latin) horned face.

UCMP: University of California Museum of Paleontology,

Diagnosis. As for type species, by monotypy.

Berkeley.

Distribution. As for type species, by monotypy. Tatankaceratops sacrisonorum gen. et. sp. nov.

Systematic Paleontology Ornithischia Seeley 1887 Neoceratopsia Sereno 1986

Holotype. BHI-6226, a skull and partial skeleton from the Hell Creek Formation in Harding County, South Dakota. Diagnosis. The new species is distinguishable from other late

Ceratopsidae Marsh 1888

Maastrichtian ceratopsids based on its small adult size, lack of

Chasmosaurinae Lambe 1915

development of cornual sinuses, proportionally massive nasal

Tatankaceratops gen. nov.

horncore, nasal horn the same length as post-orbital horns, nasal horncore recurved, proportionally large orbit compared

Type species. Tatankaceratops sacrisonorum n. sp. by monotypy. Etymology. ‘‘Tatanka’’: Lakota Sioux word for the Ameri-

to skull size, origination of postorbital horncores caudal to the orbit, no overlap of orbit by postorbital horncore, and no evidence of epijugal ossifications.

can Bison (Bison bison), in reference to this specimen being

Type Locality. BHI-6226 was collected on the Niemi Ranch in

roughly the size of an American Bison, its locality being very

northwestern South Dakota, approximately 10 miles north-

close to the town of Buffalo, South Dakota, and to recognize

northwest of the town of Buffalo, Harding County. Precise

204 ott & larson

locality data is available from Black Hills Institute upon re-

a maximum width of 90 mm. The proportionally large orbit

quest. The specimen was surface collected from an area of

size may be a retained juvenile trait, and indicates that het-

approximately 10 square m, and a quarry of about 4 square m.

erochrony, namely progenesis, may be a potential speciation

Many of the surface collected fragments were found to attach

mechanism for this animal (Carpenter et al. 1994; Weisham-

to bones excavated in situ, and no elements were duplicated,

pel and Horner 1994; McNamara 1986). Several additional

so we believe this to be a single individual. Sediments at

fragments (not figured) are thought to be from the postorbi-

and below the site were systematically screen-washed over a

tals and frontals.

period of several years to find and collect many additional

The horncore of BHI-6226 originates behind the caudal margin of the orbit, and the rostral edge of the horncore does not

fragments. Geology and Age. BHI-6226 is from the upper third of the Hell

expand forward to overlap any portion of the orbit. This condi-

Creek Formation, approximately 20 m below the K-T bound-

tion is different from that seen in MOR 692, a specimen tenta-

ary. This formation is late Maastrichtian (latest Cretaceous) in

tively identified as Avaceratops, in which the horncore origi-

age, and falls in the Lancian North American Land Mammal

nates directly above the orbit (Penkalski and Dodson 1999: fig.

Age. The Hell Creek Formation in the area is approximately

11). In Triceratops, the center of the base of the horncore is

100 m thick, and overlies the near-shore marine Fox Hills For-

usually directly above the posterior edge of the orbit, and the

mation and deep-water Pierre Shale (Waage 1968).

horncore generally expands forward such that it overlaps most

The specimen was collected from the base of a dark grey

of the orbit (Hatcher et al. 1907; Dodson et al. 2004; Lehman

overbank mudstone, which represents a typical floodplain de-

1990). Very young individuals of Triceratops, notably UCMP

posit from a meandering fluvial system. The bones were lying

154452 and BHI-4954 (Fig. 14.3), show origination of the

directly upon unconsolidated (point-bar) sandstone. The Hell

horncore directly over the orbit (Goodwin et al. 2006).

Creek Formation is a succession of stacked meandering fluvial

Neither postorbital horn shows development of a cornual

systems, prograding generally west-to-east in a very low-

sinus, which in adult Triceratops and Torosaurus extends well

gradient, fresh-water dominated environment, with intermit-

into the horncore (Hatcher et al. 1907; Forster 1996a; Farke

tent marine influence derived from the persisting Western In-

2006). The development of cornual sinuses is seen in very

terior Sea to the east. (Brown 1907; Kennedy et al. 1998; Hart-

young chasmosaurines (Farke 2006; Horner and Goodwin

man and Kirkland 2002; Murphy et al. 2002).

2006), and the lack of development in this full-grown animal

Etymology. ‘‘sacrisonorum’’: in honor of Stan and Steven Sacrison, the twin brothers from Buffalo, South Dakota, who dis-

is a diagnostic feature clearly separating this taxon from contemporary chasmosaurines. The surfaces of the orbital horncores of BHI-6226 show the

covered and collected the specimen.

deeply incised dendritic veination pattern and rugose surface texture typical of adult ceratopsids (Sampson et al. 1997). This

Description

is in contrast to the relatively smooth surfaces seen in juvenile

SKULL

Triceratops horncores (Fig. 14.3) that are both smaller (BHI4954) and larger (BHI-6271) than those of BHI-6226. The ad-

At present, 27 skull elements of BHI-6226 have been positively

vanced development of the veination patterns and the rugose

identified, representing approximately 50% of the skull by

surface textures in BHI-6226 show that this animal has at-

bone count. Several other skull bones are tentatively identi-

tained an ontogenetic stage beyond that of Triceratops speci-

fied but are so fragmentary that they are not included in the

mens that are equivalent size or larger.

bone count. Fig. 14.1A is a schematic reconstruction of the

The horncores of BHI-6226 begin trending vertically just

appearance of BHI-6226 scaled from preserved elements and

behind the orbit, and then incline rostrally, a condition noted

using Triceratops horridus (TCM 2001.93.1) as a guide.

in very young specimens of Triceratops (UCMP 154452 (Good-

Postorbital. Both postorbitals are preserved, though incom-

win et al. 2006); SMNH P2299, SMNH P26131.1 (Tokaryk

plete (Fig. 14.2). The right postorbital is represented predomi-

1997); BHI-4954) and adult specimens of Triceratops (Hatcher

nantly by the horncore, which measures 145 mm from the tip

et al. 1907). Juvenile chasmosaurines the size of BHI-6226 gen-

to the base of the horncore, as defined by a reversal of curva-

erally show an incline to vertical and often caudal recurvature

ture between the horncore and the basal portion of the post

(Lehman 1990; Goodwin and Horner 2001).

orbital. The basal circumference of the horncores is approxi-

Frontal. The frontals are fused to the postorbitals to such a

mately 210 mm on both sides. The left postorbital is repre-

degree that the suture has been obliterated. Portions of the

sented by most of the basal portion, and 68 mm of the shaft of

frontals are preserved on the postorbitals and as fragments.

the horncore, as well as a weathered fragment that represents

The partial frontals appear similar in form to those of other

the tip. The orbit is very large for the size of this individual,

chasmosaurines. Small sinuses appear between the roof and

when compared to that of other adult ceratopsian skulls, with

floor of the frontals, as seen in Triceratops (pers. obs.).

A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation 205

FIGURE 14.2.

Postorbital and fused prefrontals of BHI-6226. (A) Left medial view; (B) left lateral view; (C) right lateral view; (D) right medial view. Scale bar is 5 cm.

Prefrontal. The prefrontals are fused to the postorbitals. In

and left nasals and a possible epinasal (Lambe 1915; Gilmore

Triceratops such fusion occurs as the individual nears maturity

1917). Most of the nasal horncore is preserved, though the

(Hatcher et al 1907; pers. obs.). Juvenile horncores exhibit

remainder of the nasals is in disjointed fragments (Fig. 14.4).

an unfused suture for the prefrontals (Goodwin et al. 2006;

The total height of the horn is 145 mm as measured from the

Goodwin and Horner 2001; Tokaryk 1997) as may be seen in

upper arch of the nares perpendicular through the tip of the

Fig. 14.3. This suture is fully fused and remodeled in BHI-

horn. This is the same total length as the brow horncores. This

6226, and is no longer visible.

nasal horn is proportionally larger in comparison to the brow

Nasal Horncore. The nasal horncore is composed of the right

206 ott & larson

horns than in any other known contemporary chasmosau-

FIGURE 14.3.

Postorbitals of Tatankaceratops. (A) BHI-6226, left side; (B) BHI6271, right side reversed view; (C) BHI-4954, left side. prf: fused prefrontal; prfa: prefrontal articular surface. Scale bar is 5 cm.

rine. This condition is somewhat similar to that inferred for

(Forster 1996a) this specimen indicates that the arcuate blood

the Judithian Avaceratops lammersi (Dodson 1986), though no

vessel may be found in other taxa. The presence of this charac-

nasal horn is known for that taxon.

ter may indicate a close relationship between BHI-6226 and

The nasal horncore in Triceratops begins as a separate os-

Triceratops.

sification (the epinasal) quite early in ontogeny, and is often

Premaxilla and Rostrum. Both premaxillae are represented, as

seen separated from the nasals in juvenile animals and occa-

is the rostrum (Fig. 14.5). As with many other elements, the

sionally even in adult-sized individuals (Hatcher et al. 1907;

dorsal margin of the premaxillae and the rostrum show deeply

Brown and Schlaikjer 1940a; Goodwin et. al 2006; Horner and

developed dendritic patterns, indicative of adult status. The

Goodwin 2008; pers. obs.). In BHI-6226, it is not currently

premaxillae appear to be fused to each other as well as to the

possible to determine whether the nasal horncore originates

nasals and the rostrum, although some of the sutures are still

as an outgrowth of the nasals or is formed by a separate epi-

discernable. This condition is often seen in adult specimens of

nasal ossification. The advanced state of fusion of the nasal

Triceratops (TCM 2001.93.1; BHI-6220) as well. It is not pos-

elements precludes a firm diagnosis of this character at this

sible to discern the anterior premaxilla-rostrum suture due

time, though further radiographic analysis may eventually

to apparent fusion and development of rugose surface tex-

provide a concrete answer. The closure of the nasal midline

ture. The ventral premaxilla-rostrum suture is still visible on

suture is another indication of the advanced age of BHI-6226

a portion of the right premaxilla but is not apparent else-

(Horner and Goodwin 2006).

where. Observations of an adult cf. Torosaurus skull (BHI-

As in Triceratops, BHI-6226 possesses the impressed trace

4772—skull length 2.3 m) and a subadult Triceratops horridus

of an arcuate blood vessel at the base of the nasal horncore.

(ROM 55380—skull length 1.6 m) show no fusion of the pre-

While this has been used as an autapomorphy for Triceratops

maxillae to the rostrum, the nasal, or to each other, thus it is

A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation 207

FIGURE 14.4.

Nasal horn of BHI-6226. (A) Right lateral view; (B) left lateral view. Scale bar is 5 cm.

Triceratops prorsus (Forster 1996b). The reconstructed depth of the snout is approximately 150 mm, measured from the insertion point of the premaxillae into the nasals to the inferred ventral surface of the premaxillae. Maxilla. Fragments of one or both maxillae are present, though neither appears very complete. They are of typical ceratopsid form, but are too fragmentary to reveal diagnostic features. No maxillary teeth have yet been identified. Parieto-Squamosal Frill. Portions of the fused parietals and the left and right squamosals are preserved, though all are fragmented (Fig. 14.6, and also refer to Fig. 14.1B). The portions present (more than 100 fragments) show characteristics that have been interpreted as advanced age (Sampson et al. 1997) indicating that, despite its diminutive size, BHI-6226 is a mature animal. Approximately one-quarter of the left squamosal is preserved as a single piece with a maximum length of 148 mm. The entire bone is estimated to have been perhaps 350 mm long and 200 mm deep, although not enough of the element is present to provide an exact length. The preserved portion is a maximum of 38 mm thick at the squamosal-exoccipital buttress, and it thins caudally to a minimum of 8 mm. It appears that the bulk of the squamosal lies behind the quadrate/ exoccipital buttress, which is similar to the condition of the squamosal of Triceratops, but differs from that of Avaceratops Rostrum of BHI-6226 fused to the premaxillae. Scale bar is 5 cm. FIGURE 14.5.

(Penkalski and Dodson 1999). The squamosal is highly ornamented with rugose surface texture and dendritic veination patterns on the external surface. This is interpreted to be an adult feature (Sampson et al. 1997). A smaller fragment

likely that fusion of elements in the snout is again an indicator

of squamosal exhibits attached epoccipitals with the sutures

of advanced age for BHI-6226 (Horner and Goodwin 2008).

fused to the point of obscurity, again an adult feature (Leh-

Total length of the snout from the base of the nasal horn to

man 1990).

the tip of the rostrum is approximately 160 mm. The shape of

The fused parietals are represented by several fragments, the

the snout is reminiscent of T. horridus in that it is fairly long

largest of which is 80 mm wide by 100 mm long. The estimated

and curved, instead of relatively short and more vertical as in

total length of the parietals is 42 cm. Enough of the parietals

208 ott & larson

FIGURE 14.6.

Epoccipitals of BHI-6226 fused to parietals (A, B) and squamosals (C, D). (A) Dorsal view; (B) ventral view; (C) lateral view; (D) medial view. fes: fused epoccipital suture. Scale bar is 5 cm.

FIGURE 14.7. Occipital condyle of BHI-6226. (A) Posterior view; (B) ventral view; (C) dorsal view. Scale bar is 5 cm.

are preserved to conclude that they do not appear to have

plete fusion of the condyle (Marsh 1889; pers. obs.). In

been fenestrated, a condition known previously only in Tri-

BHI-6226 the junctions between the three components are

ceratops and Avaceratops. Rugose surface texture and dendritic

completely fused and remodeled so that no trace of the suture

veination patterns are again well-developed, and epoccipitals

is present (Fig. 14.7). This is again interpreted as an adult fea-

are fused.

ture, although it does not necessarily mean that growth has

A total of five epoccipitals are currently represented on the

completely ceased (Dodson 1996). Unfortunately the rest of

squamosal and parietal fragments and all show fused, remod-

the braincase (including the wings of the exoccipitals) is frag-

eled, and obliterated sutures. The epoccipitals are triangular in

mented, so no other information is available at this time.

shape on the parietals and the caudal portion of the squamo-

Jugal. Portions of both jugals are preserved and both show

sals, but on the rostral portion of the squamosals they are

rugose surface texture consistent with adulthood (Sampson et

somewhat bar-shaped. They are all flattened dorsoventrally, a

al. 1997). The quadratojugal-jugal suture is smooth and shows

condition that has been interpreted as an adult characteris-

no evidence of fusion of the two elements. Neither jugal has

tic (Horner and Goodwin 2006). The largest epoccipital pre-

any indication of an epijugal horn, which is an adult charac-

served measures 60 mm wide, with a base to vertex height of

teristic of Triceratops (Hatcher et al. 1907; Horner and Good-

29 mm. This epoccipital is likely from the median of the fused

win 2008). The estimated length measured from the lower

parietals due to its large size, parallel orientation with the pa-

edge of the eye orbit to the ventral tip of the jugal is approxi-

rietals, and nearly symmetrical shape. No estimate of total

mately 200 mm (Fig. 14.8).

number of epoccipitals can be made at this time. Many adult

Quadrate. Both quadrates are present (Fig. 14.9). The pre-

Triceratops specimens lack fused epoccipitals, and fused epoc-

served length of the left quadrate is 144 mm and total es-

cipitals are generally thought to represent advanced age (Leh-

timated length is 200 mm. The lateral portion of the distal

man 1990; Horner and Goodwin 2006, 2008).

articular surface is preserved. The quadrate shows no indica-

Occipital Condyle. In ceratopsids the occiput is formed by the

tion of fusion with the quadratojugal, although that articular

junction of the basioccipital and exoccipitals. In many cases,

surface is somewhat scalloped, limiting movement and pro-

including animals of much larger size, the three elements

viding rigid support for the jaw articulation.

remain distinguishable throughout ontogeny due to incom-

Lacrimal. The left lacrimal is missing the anterior margin,

A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation 209

FIGURE 14.10. Right lacrimal of BHI-6226. (A) Posterior view; (B) lateral view. asp: articular surface for the prefrontal; asj: articular surface of the jugal. Scale bar is 5 cm.

but retains the posterior portions of the sutures for articulation with both the postorbital and the jugal (Fig. 14.10). The left lacrimal is somewhat rectangular shaped in lateral view and measures 25 mm dorsoventrally (between the postorbital and jugal sutures) and 20 mm mediolaterally. FIGURE 14.8. Jugals of BHI-6226. (A) Left medial view; (B) left lateral view; (C) right lateral view; (D) right medial view. Scale bar is 5 cm.

MANDIBLE The majority of the left dentary is preserved, along with right and left articulars, surangulars, angulars, and parts of both splenials. The predentary is also present in several fragments. Two small (51 — 32 mm) nearly complete elements were also collected and have been identified as the right and left coronoids. All lower jaw elements show development of dense adult-style bone surface texture, as well as a deeply impressed dendritic pattern on the external surface of the surangulars. Predentary. The estimated total length of the predentary is approximately 150 mm. The anterior portion is present and shows the curvature of the bone almost to the tip of the beak (Fig. 14.11). Posteriorly, parts of the predentary-dentary contact are preserved. Length was estimated by extrapolation using chasmosaurine predentaries for guides. As with the other skull elements described, the predentary shows well-developed dendritic veination patterns indicative of adulthood. Dentary. The left dentary (Fig. 14.12) has a maximum length of 345 mm and height through the coronoid process is 72 mm. This dentary is 100 mm longer than the type of Avaceratops, but is morphologically similar. The supradental plate is preserved, but is separated from the dentary. Several dozen dentary teeth

Right quadrate of BHI-6226. (A) Posterior view; (B) lateral view; (C) anterior view. Scale bar is 5 cm. FIGURE 14.9.

210 ott & larson

are present and are of typical ceratopsid morphology. The coronoid process twists sharply laterally. The anterior

FIGURE 14.11.

Left lateral view of predentary of BHI-6226. asd: articular surface for the dentary. Scale bar is 5 cm.

projection of the process forms a nearly 90 degree angle with

although both preserve most of the articular surface that

the ascending ramus of the coronoid process. The postero-

they share with the dentary (Fig. 14.16). The right splenial

dorsal margin is truncated by a sutured facet that receives

measures 151 mm in length, although the total maximum

the coronoid. Because Triceratops lacks a separate coronoid

length is estimated to be nearly 240 mm by measurement of

BHI-6226 has a coronoid process with a very different shape

the ventral portion of the dentary. The portion that is pre-

then that of Triceratops.

served shows some similarities to that of Triceratops (TCM

Coronoid. Both coronoids are virtually complete and are

2001.93.1). However, the articular surface between the splenial

mirror images. They are triangular in lateral view and wedge

and dentary of BHI-6226 is much broader proportionally than

shaped mediolaterally (Fig. 14.13). The right coronoid mea-

that of TCM 2001.93.1, whose skull length measures in excess

sures 51 mm in length, 29 mm in height, and 12 mm medi-

of 2 m. Both specimens have a maximum articular surface

olaterally. The left coronoid articulates with the facet found

width of 20 mm, although the length of the splenial in TCM

on the posterio-dorsal margin of the coronoid process of the

2001.93.1 is 447 mm, compared to a maximum length in

left dentary (see Fig. 14.12). Coronoids have not been found

BH-6226 of no more than 240 mm.

in even the youngest of the Triceratops specimens (Goodwin et

The remainder of the skull is represented by well over 1,000

al. 2006) therefore they must have either fused with the coro-

fragments, many of which probably represent more of the

noid process in that taxon very early in ontogeny, or be ab-

elements listed above as well as the palate and braincase.

sent. The presence of coronoids in Tatankaceratops appears to be a primitive character because within the Ceratopsia it has been reported only in psittacosaurs (Fig. 14.14) and primi-

POSTCRANIAL SKELETON

tive neoceratopsians (Brown and Schlaikjer 1940b; Makovicky

Portions of three cervical and nine dorsal vertebrae are rep-

and Norell 2006; Chinnery and Horner 2007).

resented (Fig. 14.17A). The three cervicals are likely C-6, 7,

Surangular. Both surangulars are partially preserved. Both

and 8, and the dorsal vertebrae are likely D-4 through D-12

preserve a portion of the articular surfaces connecting the sur-

(Ostrom and Wellnhofer 1986). The neural arches of the cervi-

angular with the dentary, angular, and articular. Morphologi-

cals and two dorsals are broken, though they may be present

cally the surangular of BHI-6226 appears to be similar to that

as fragments. The remainder of the dorsals all possess neural

of Triceratops.

arches and varying portions of the transverse processes. All

Articular. The articulars are preserved as virtually complete

vertebrae that preserve the base of the neural arches show

elements and show the entire double-cupped articulation sur-

complete fusion and obliteration of the centrum/neural arch

face that receives the distal quadrate (Fig. 14.15). The articula-

suture. This again supports the interpretation of adult status

tions with the surangulars and angulars are smooth and show

of this individual (Chinnery 2004; Ikejiri et al. 2005; Chinnery

no signs of fusion. They measure 98 mm rostrocaudally, 46

and Horner 2007).

mm dorsoventrally and 61 mm mediolaterally.

Two nearly complete cervical ribs and portions of at least 4

Angular. The right and left angulars are preserved as frag-

others are present (Fig. 14.17B). At least 10 dorsal ribs are rep-

ments. Some of the articular surfaces are preserved showing

resented (Fig 17C), 2 of which are nearly complete. The distal

gross morphology consistent with that of other chasmo-

end of the right ulna, fragments of a scapula, 1 manus ungual

saurines.

(missing the tip), and 1 phalanx are also present. Other post-

Splenial. The left splenial is more complete than the right,

cranial bones are represented, but their fragmentary nature

A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation 211

FIGURE 14.12. Left dentary and coronoid of BHI-6226. (A) Lateral view; (B) medial view. c: coronoid; cp: coronoid process. Scale bar is 5 cm.

212 ott & larson

this specimen codes a ‘‘1’’ for 14 of Dodson et al.’s (2004) 29 Ceratopsidae characters with no disagreement. Unfortunately we were only able to code 9 sub-Ceratopsidae characters (Table 14.2). Of these 9, 2 unambiguous characters used by Dodson et al. (2004) to separate centrosaurine from chasmosaurine ceratopsids show conflicting results when applied to this specimen. Possession of triangular squamosal epoccipitals indicates a chasmosaurine affinity for Tatankaceratops, but presence of postorbital horns less than 15% of basal skull length indicates a centrosaurine affinity. This ambiguity undoubtedly was instrumental in the assignment of Tatankaceratops as basal to the Chasmosaurinae (Fig. 14.18). Additionally, the fact that BHI-6226 retains coronoids, a character not shared with any other member of Ceratopsomorpha (Wolfe and Kirkland 1998), seems to support this analysis. Although Wolfe and Kirkland (1998) use the term ‘‘coronoid’’ in their description of Zuniceratops, they are clearly referring to the coronoid process, which lacks a separate coronoid element. In this analysis the Tree Length is 87, Consistency Index (CI) is 0.8851, CI excluding uninformative characters is 0.8387, Homoplasy Index (HI) is 0.1149, HI excluding uninformative characters is 0.1613, Retention Index (RI) is 0.9291, and Rescaled Consistency Index (RC) is 0.8223. Because this analysis is based upon a fragmentary and partially prepared specimen, we expect that the phylogenetic position of Tatankaceratops will be clarified as more characters are revealed. Accordingly, this analysis should be regarded as preliminary.

Discussion The overall small size and consistent appearance of adult features on all elements of BHI-6226, along with morphological differences between this specimen and other contemporaneous ceratopsids, support the erection of the new taxon Tatankaceratops sacrisonorum for this specimen. The adult Right coronoid of BHI-6226. (A) Medial view; (B) anterior view with facet that articulates to the coronoid process of the dentary; (C) lateral view. asd: articular surface for the dentary. Scale bar is 5 cm. FIGURE 14.13.

characteristics include rugose surface texture and dendritic veination patterns on the nasal horn, postorbitals, frontals, prefrontals, parietals, squamosals, rostrum, premaxillae, jugals, and surangulars. In addition the forward-directed postorbital horns and the fusion of many skull and postcranial elements indicate the adult status of Tatankaceratops. Au-

precludes certain identification at this time. Further prepa-

tapomorphies include a nasal horn larger than the postorbital

ration should allow these fragments to be identified in the

horns, postorbital horns originating behind the orbits, ab-

future.

sence of a corneal sinus in postorbital horn, coronoid process truncated caudodorsally and faceted to receive the coronoid,

Preliminary Cladistic Analysis

and the presence of a coronoid. We examined a number of alternate hypotheses to erecting

Because preparation is incomplete, we were only able to con-

a new taxon. It is possible that this animal is closely related to

duct a very preliminary cladistic analysis using Dodson et al.’s

and possibly descended from Triceratops. If that is the case, this

(2004) characters (Table 14.1). Although the fragmentary na-

could represent an example of paedomorphosis (progenesis)

ture of BHI-6226 limited the number of codeable characters,

within that taxon (McNamara 1986; Weishampel and Horner

A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation 213

FIGURE 14.14.

Medial view of the posterior half of the right mandible of Psittacosaurus sp. (BHI-6274) showing the coronoid in place. Scale bar is 5 cm.

1994). However, the current preliminary cladistic analysis does not support this hypothesis. The analysis also discounts the possibility of BHI-6226 being an example of dwarfism in Triceratops as occurred in an isolated population of German sauropods (Sander et al. 2006). This hypothesis seems to have little relevance to BHI-6226, because no evidence of isolated island ecosystems exists within the Hell Creek Formation, especially in the area where this animal was collected. In addition, we considered the alternative that BHI-6226 could represent a late Maastrichtian descendant of the late Campanian Avaceratops. The type specimen of Avaceratops, ANSP 15800 (Dodson 1986), is somewhat smaller than BHI6226, though ANSP 15800 is a juvenile individual. The possible second Avaceratops specimen, MOR 692 (Penkalski and Dodson 1999), compares very favorably in size to BHI-6226. However, both Avaceratops specimens are morphologically and stratigraphically distinct from BHI-6226. Therefore, BHI6226 cannot be referred to Avaceratops. Recent analyses (Dodson et al. 2004) have placed Avaceratops at the base of the ceratopsid tree, or within the Centrosaurinae. In BHI-6226, two unambiguous characters used by Dodson et al. (2004) to separate Centrosaurinae from Chasmosaurinae are preserved. One, triangular squamosal epoccipitals, indicates a chasmosaurine affinity for Tatankaceratops, and the second, postorbital horns less than 15% of basal skull length, indicates a centrosaurine affinity. Therefore, an unambiguous subfamily Articulars of BHI-6226. (A) Right dorsal view; (B) right ventral view; (C) left dorsal view; (D) left ventral view. asq: articular surface for the quadrate; ass: articular surface for the surangular. Scale bar is 5 cm. FIGURE 14.15.

214 ott & larson

designation is impossible at this time, although our opinion, supported by BHI-6226’s affinities to contemporary chasmosaurines, is that Tatankaceratops is most likely a chasmosaurine. Unfortunately, due to the fragmentary nature of

FIGURE 14.16. Left splenial of BHI-6226. (A) Lateral view; (B) medial view. asd: articular surface for the dentary. Scale bar is 5 cm.

BHI-6226, further preparation will be required and the discovery of another specimen may be required before a more meaningful phylogenetic analysis can clarify the relationships between Tatankaceratops and other neoceratopsians. The most conservative approach in this case would be to erect a new species of Triceratops. However, we feel that the morphological and size differences between this animal and Triceratops, as well as BHI-6226’s superficial similarity to the non-contemporaneous Avaceratops, are sufficient to separate this individual into a new taxon. Additionally, this specimen falls well outside of the ontogenetic series that has been described for Triceratops. Referring this specimen to a new species of Triceratops or to Avaceratops would do nothing more than confuse the issue. We feel that as the tedious preparation of this specimen proceeds more characters will emerge, creating the necessity for a second report. It is also quite likely that more specimens of this small animal have been discovered, and have been identified as juveniles of Triceratops. Given the current interest in the Hell Creek and Lance Formations, it is only a matter of time for new specimens to be located and the relationship between this animal and other ceratopsians to be clarified. Acknowledgments

The authors thank Stan and Steven Sacrison for the discovery, collection, and (in part) donation of the specimen to the Black Hills Institute; the Niemi family for continued access to the site; the staff of the Black Hills Institute for collection assistance, preparation, and curation of this specimen; N. Larson, T. Larson, and L. Shaffer for illustration preparation; A. Farke for running the PAUP analysis; R. Bakker, P. Dodson, A. Farke, R. Farrar, L. Hutson, J. MacQuaker, P. Manning, K. Seymour,

FIGURE 14.17. Representative postcranial elements of BHI-6226. (A) Lateral view of anterior dorsal vertebra, showing fused neural arch (inset); (B) left seventh cervical rib; (C) right second dorsal rib. Scale bar is 5 cm.

and D. Tanke for information, ideas, and suggestions; and Brenda Chinnery-Allgeier and Ken Carpenter for reviewing this manuscript.

A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation 215

FIGURE 14.18. Preliminary cladogram showing our current understanding of the phylogenetic position of Tatankaceratops based upon characters listed in Table 14.2.

Table 14.1. Taxon Character Matrix Used for Ceratopsid Inter-relationships for Tatankaceratops sacrisonorum (after Dodson et al. 2004) 10

20

30

40

50

60

70

80 000

Protoceratops

000?0? ? ?0?

0?00000000

0? ? ? ?00000

?000000000

00? ? ? ? ? ? ? ?

?000000000

0000000000

Zuniceratops

??????????

??????????

10000? ? ? ? ?

??????????

??????????

? ? ? ? ? ? ?010

00? ? ? ? ? ? ? ?

???

Achelosaurus

01100? ?00?

0?110120?1

1101? ?1111

0111111100

1111010001

1111111111

1111011110

111

Anchiceratops

11111101?0

1001011101

10101?0121

2010001100

1100100000

0111110111

111? ?11111

211

Arrhinoceratops

111111010?

1001011101

10101?01?1

?010001100

1? ? ? ?00000

01?111? ? ?1

??????????

???

Avaceratops

?1100? ?00?

0?11? ? ? ? ? ?

? ? ? ? ? ? ?1? ?

? ?10101? ? ?

??????????

?1?111?111

111? ? ?1? ? ?

?11

Centrosaurus

01100? ?00?

0?11011011

1001011111

0111111100

1111011111

0111111111

1111011110

111

Chasmosaurus

11111011?0

1000111?01

100?00?111

1010001110

0100100100

0111110111

1111111111

211

Diceratops

1111100111

1101011101

10101?0121

2010001101

110?100000

01?111? ? ?1

11? ? ? ? ? ? ? ?

???

Einiosaurus

01100? ?00?

0?11011011

1001? ?1111

0111111100

1111010001

1111111111

1111011110

111

Pachyrhinosaurus

01100? ?00?

0?110120?1

1101? ?1111

0111111100

1111010011

1111111111

1111011110

111

Pentaceratops

11111011?0

1000111?01

10001?01?1

1010001110

0100100100

0111110111

11111?1111

211

Styracosaurus

01100? ?00?

0?11011011

1001?11111

0111111100

1111011111

1111111111

1111011110

111

Torosaurus

1111100111

? ?01011101

1010100121

2010001101

1100100000

0111110111

11? ? ? ? ? ? ? ?

???

Triceratops

1111100111

1101011101

1010100121

2110001100

1110100000

0111110111

1111111111

211

Tatankaceratops

11111? ?1? ?

? ? ? ? ?11? ?1

11?1? ? ?1?1

? ? ? ? ? ?1? ? ?

?11?1? ?1? ?

? ? ? ? ?1?1? ?

11? ? ? ?1? ? ?

???

216 ott & larson

Table 14.2. Ceratopsidae and Sub-Ceratopsidae Characters of Tatankaceratops sacrisonorum (after Dodson et al. 2004) Parietal epoccipital at locus 1 caudally directed

1

Greatly enlarged external nares

1

Prominent premaxillary septum

1

Antorbital fenestra greatly reduced

1

Presence of a nasal horncore

1

Reduced lachrimal

1

Frontal elimination from orbital margin

1

Dentary articulation set well below the alveolar margin

1

Marginal ungulations on frill augmented by epoccipitals

1

Tooth row extended caudal to coranoid

1

Laterally stacked quadrat-quadratojugal-jugal

1

Elongate groove on squamosal to receive quadrate

1

More than two replacement teeth in each vertical series

1

Loss of subsidiary ridges on teeth

1

Hoof-like pedial unguals

1

Postorbital horns 15% of basal skull length

0

Epoccipital crosses the patietal squamosal joint

1

Rostral enlarged with deeply concave caudal margin and hypertrophied dorsal and ventral process

1

Premaxillary septum

1

Premaxillary narial strut

1

Presence of inter-premaxillary fossa

1

Triangular squamosal epoccipitals

1

Large parietal fenestrae

0

References Cited Brown, B. 1907. The Hell Creek beds of the Upper Cretaceous of Montana. Bulletin of the American Museum of Natural History 23: 823–845. ———. 1914. Leptoceratops, a new genus of Ceratopsia from the Edmonton Cretaceous of Alberta. Bulletin of the American Museum of Natural History 33: 549–558. Brown, B., and E. M. Schlaikjer. 1940a. The origin of ceratopsian horn-cores. American Museum Novitates 1065: 1–7. ———. 1940b. The structure and relationships of Protoceratops. Annals of the New York Acadamy of Sciences 40: 133–266. Carpenter, K., K. F. Hirsch, and J. R. Horner. 1994. Summary and prospectus. In K. Carpenter, K. F. Hirsch, and J. R. Horner, eds., Dinosaur Eggs and Babies, pp. 366–370. Cambridge: Cambridge University Press. Chinnery, B. J. 2004. Description of Prenoceratops pieganensis gen. et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Chinnery, B. J., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Dodson, P. 1986. Avaceratops lammersi: A new ceratopsid from the

Judith River Formation of Montana. Proceedings of the Academy of Natural Sciences of Philadelphia 138: 305–317. ———. 1996. The Horned Dinosaurs. Princeton: Princeton University Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Farke, A. A. 2006. Morphology and ontogeny of the cornual sinuses in chasmosaurine dinosaurs (Ornithischia: Ceratopsidae). Journal of Paleontology 80: 780–785. Forster, C. A. 1996a. New Information on the skull of Triceratops. Journal of Vertebrate Paleontology 16: 246–258. ———. 1996b. Species resolution in Triceratops: Cladistic and morphometric approaches. Journal of Vertebrate Paleontology 16: 259–270. Gilmore, C. W. 1917. Brachyceratops, a ceratopsian dinosaur from the Two Medicine Formation of Montana. U.S. Geological Survey Professional Paper 103: 1–45. Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian. 2006. The smallest known Triceratops skull: New observations on ceratopsid cranial anatomy and ontogeny. Journal of Vertebrate Paleontology 26: 103–112. Goodwin, M. B., and J. R. Horner. 2001. How Triceratops got its horns: New information from a growth series on cranial morphology and ontogeny. Journal of Vertebrate Paleontology 21(3, Suppl.): 56A. Hartman, J. H., and J. I. Kirkland. 2002. Brackish and marine mollusks of the Hell Creek Formation of North Dakota: Evidence for a persisting Cretaceous seaway. In J. H. Hartman, K. R. Johnson, and D. J. Nichols, eds., The Hell Creek Formation and the Cretaceous-Tertiary Boundary in the Northern Great Plains: An Integrated Continental Record of the End of the Cretaceous, pp. 271–296. Geological Society of America Special Paper 361. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49. Horner, J. R, and M. B. Goodwin. 2006. Major cranial changes during Triceratops ontogeny. Proceedings of the Royal Society Biological Sciences 273: 2757–2761. ———. 2008. Ontogeny of cranial epi-ossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Ikejiri, T., V. Tidwell, and D. L. Trexler. 2005. New adult specimens of Camarasaurus lentus highlight ontogenetic variation within the species. In V. Tidwell and K. Carpenter, eds., Thunder-Lizards: The Sauropodomorph Dinosaurs, pp. 154–179. Bloomington: Indiana University Press. Kennedy, W. J., N. H. Landman, W. K. Christensen, W. A. Cobban, and J. M. Hancock. 1998. Marine Connections in North America during the late Maastrichtian: Paleogeographic and paleobiogeographic significance of Jeletzkyties nebrascensis zone cephalopod fauna from the Elk Butte Member of the Pierre Shale, SE South Dakota and NE Nebraska. Cretaceous Research 19: 745–775. Lambe, L. M. 1915. On Eoceratops canadaensis, gen. nov., with remarks on other genera of Cretaceous horned dinosaurs. Geological Survey of Canada Museum Bulletin 12: 1–49.

A New, Small Ceratopsian Dinosaur from the Latest Cretaceous Hell Creek Formation 217

Lehman, T. M. 1990. The ceratopsian subfamily Chasmosaurinae: Sexual dimorphism and systematics. In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 211–229. Cambridge: Cambridge University Press. Makovicky, P. J., and M. S. Norrell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates No. 3530. Marsh, O. C. 1888. A new family of horned dinosaurs from the Cretaceous. American Journal of Science 36: 477–478. ———. 1889. Notice of gigantic horned Dinosauria from the Cretaceous. American Journal of Science 38: 173–175. ———. 1890. Description of new Dinosaurian reptiles. American Journal of Science 39: 82–86. ———. 1891. Notice of new vertebrate fossils. American Journal of Science 42: 265–269. McNamara, K. J. 1986. A guide to the nomenclature of heterochrony. Journal of Paleontology 60: 4–13. Murphy, E. C., J. W. Hoganson, and K. R. Johnson. 2002. Lithostratigraphy of the Hell Creek Formation in North Dakota. In J. H. Hartman, K. R. Johnson, and D. J. Nichols, eds., The Hell Creek Formation and the Cretaceous-Tertiary Boundary in the Northern Great Plains: An Integrated Continental Record of the End of the Cretaceous, pp. 9–34. Geological Society of America Special Paper 361. Ostrom, J. H., and P. Wellnhofer. 1986. The Munich specimen of Triceratops with a revision of the genus. Zitteliana 14: 111–158. Ott, C. J. 2006. Cranial anatomy and biogeography of the first Leptoceratops gracilis (Dinosauria: Ornithischia) specimens from the Hell Creek Formation, Southeast Montana. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 213–233. Bloomington: Indiana University Press. Penkalski, P., and P. Dodson, 1999. The morphology and system-

218 ott & larson

atics of Avaceratops, a primitive horned dinosaur from the Judith River Formation (Late Campanian) of Montana, with the description of a second skull. Journal of Vertebrate Paleontology 19: 692–711. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnaean Society of London 121: 293–337. Sander, P. M., O. Mateus, T. Laven, and N. Knotschke. 2006. Bone histology indicates insular dwarfism in a new Late Jurassic sauropod dinosaur. Nature 441: 739–741. Seeley, H. G. 1887. On the classification of the fossil animals commonly called Dinosauria. Proceedings of the Royal Society of London 43: 221–228. Sereno, P. C. 1986. Phylogeny of the Bird-hipped Dinosaurs (Order Ornithischia). National Geographic Research 2: 234–256. Tokaryk, T. T. 1997. First evidence of juvenile ceratopsians (Reptilia: Ornithischia) from the Frenchman Formation (Late Maastrichtian) of Saskatchewan. Canadian Journal of Earth Science 34: 1401–1404. Waage, K. M. 1968. The Type Fox Hills Formation, Cretaceous (Maastrichtian), South Dakota: Part 1. Stratigraphy and Paleoenvironments. Peabody Museum of Natural History, Yale University, Bulletin 27. Weishampel, D. B., and J. R. Horner. 1994. Life history syndromes, heterochrony, and the evolution of Dinosauria. In K. Carpenter, K.F. Hirsch, and J. R. Horner, eds., Dinosaur Eggs and Babies, pp. 229–243. Cambridge: Cambridge University Press. Wolfe, D. G., and J. I. Kirkland. 1998. Zuniceratops christopheri n. gen. & n. sp., a ceratopsian dinosaur from the Moreno Hill Formation (Cretaceous, Turonian) of west-central New Mexico. New Mexico Museum of Natural History and Science Bulletin 14: 303–317.

PART THREE ANATOMY, FUNCTIONAL BIOLOGY, AND BEHAVIOR

15 Comments on the Basicranium and Palate of Basal Ceratopsians P E T E R D O D S O N , H A I - L U Y O U , A N D K Y O TA N O U E

the anatomy of basal ceratopsians has been studied

in position at the midlength of the skull and the median

for more than eight decades, but certain anatomical re-

joint is short, while the central plate of this element in

gions of the skull have been neglected due to problems of

basal neoceratopsians is caudally positioned with an

preservation and especially of preparation. In the past

elongate median joint. In psittacosaurids the vomer is

decade many new specimens of basal ceratopsians, in-

nearly horizontal. The conspicuous choanae are very

cluding both psittacosaurids and basal neoceratopsians,

large, and open ventrally into the roof of the oral cavity.

have been discovered in China, and a number of new

Air flow through the dorsally positioned naris into the

taxa have been described. Although the Jehol fauna is

nasal cavity is predominantly vertical. In basal neo-

justifiably famous for its feathered theropods and birds,

ceratopsians the vomer is inclined caudodorsally. The

it is also the source of important psittacosaurid material,

choanae are reduced to slits at the rostral end of the oral

especially Psittacosaurus lujiatunensis and

cavity, and are difficult to view due to their steep inclina-

Hongshanosaurus houi, and basal neoceratopsians such as

tion. The air flow through the nasal cavity is much more

Liaoceratops yanzigouensis. The excellent preservation

nearly horizontal. Psittacosaurids exhibit a large number

and preparation of this fossil material now allows for

of autapomorphies in the nasal cavity. These and other

new detailed studies. We present the first detailed de-

related features add characters that provide us with

scriptions of the palate and basicranium of selected basal

new data that help us to understand the evolution of the

ceratopsians. A striking contrast is found between psit-

Ceratopsia.

tacosaurids and basal neoceratopsians in these regions. In caudal view the basioccipital of psittacosaurids is bilobate, while that of basal neoceratopsians is quadrilateral.

Introduction

The basioccipital-basisphenoid synchondrosis is exposed

Ceratopsian dinosaurs are distinctive herbivorous dinosaurs

ventrally in psittacosaurids and is covered by the

of the Cretaceous of Asia and North America. The roots of

pterygoids in basal neoceratopsians. In psittacosaurids

their radiation may be sought in the Jurassic. Recently a puta-

the basipterygoid processes of the sphenoid are long and

tive basal ceratopsian, Yinlong, was reported from the Late Ju-

directed rostrally, while those of basal neoceratopsians

rassic of northwestern China (Xu et al. 2006). Yinlong joins

are shorter and directed ventrally or ventrolaterally. The

Chaoyangsaurus and Xuanhuaceratops (Zhao et al. 1999, 2006)

central plate of the pterygoid of psittacosaurids is rostral

as basalmost Ceratopsia. For many years, only three taxa gen-

221

FIGURE 15.1.

Phylogenetic tree of the Ceratopsia compiled from You et al. (2003, 2005); Chinnery (2004); Makovicky and Norell (2006); and Xu et al. (2006).

erally characterized basal ceratopsians: Leptoceratops from Al-

only a single paper in recent years has focused on the palate of

berta (Brown 1914), and Psittacosaurus (Osborn 1923) and Pro-

a basal ceratopsian (Osmólska 1986). However, the compara-

toceratops (Granger and Gregory 1923), both from Mongolia.

tive perspective from hadrosaurs (Heaton 1972) is also useful.

Apart from the fragmentary Microceratops described by Bohlin

It was the recent fortuitous discovery of a juvenile specimen

(1953) from Gansu Province, no new basal ceratopsian taxon

of Liaoceratops (CAGS-IG-VD-002) that inspired this paper. In

was described until Bagaceratops was described by Maryanska ´

this specimen, the cranium had been removed, possibly by a

and Osmólska in 1975. The past decade has revealed a pleth-

small predator (You et al. 2007), revealing the palate from

ora of new basal ceratopsians including Archaeoceratops (Dong

above. The present paper is offered with the goal of extend-

and Azuma 1997; You and Dodson 2003); Chaoyangsaurus

ing Osmólska’s excellent study to include other taxa of basal

(Zhao et al. 1999); Liaoceratops (Xu et al. 2002); Hongshanosau-

ceratopsians.

rus (You and Xu 2005; You et al. 2003); Auroraceratops (You et

We present here a comparative study of the basicranium

al. 2005); Xuanhuaceratops (Zhao et al. 2006); and Yinlong (Xu

and palate of skulls (with special emphasis on the pterygoids,

et al. 2006), all from China. In recent years, there has also been

palatines, and vomer) of several taxa of both psittacosaurids

renewed interest in Psittacosaurus, in including the descrip-

and basal neoceratopsians that have been collected and/or

tion of new species: P. xinjiangensis (Sereno and Chao 1988); P.

described during the past decade: from the Lujiatun village

meileyingensis (Sereno et al. 1988); P. neimongoliensis (Russell

site of the Yixian Formation in Liaoning, northeastern China,

and Zhao 1996) and P. ordosensis (Russell and Zhao 1996),

one specimen each of Psittacosaurus major and P. lujiatunensis,

both from Inner Mongolia; P. mazongshanensis (Xu 1997) from

one specimen of Hongshanosaurus houi, and three specimens

Gansu; P. sibiricus from Siberia (Averianov et al. 2006); P. lu-

of Liaoceratops yanzigouensis. One specimen each of Archaeo-

jiatunensis (Zhou et al. 2006) from Liaoning; and P. major (Se-

ceratops oshimai and Auroraceratops rugosus come from the late

reno et al. 2007), also from Liaoning. A general understanding

Early Cretaceous Xinminpu Group of Gansu, northwestern

of the phylogeny of basal ceratopsians is shown in Fig. 15.1.

China (Table 15.1). Certain details can be confirmed in Ar-

Ceratopsians have distinctive cranial morphology, and

chaeoceratops, but lateral crushing has fractured the palate,

readily yield characters that permit phylogenetic analysis (Fig

rendering it less anatomically useful than the preceding speci-

15.1.; Sereno 2000; Makovicky 2001; You and Dodson 2003,

mens. A common feature of all these specimens is that nearly

2004; Xu et al. 2006). Not only have new materials been de-

all sediment has been removed from internal spaces, and ex-

scribed, but in many cases, new preparation practices have

cept for Hongshanosaurus, one or both of the jaws have been

provided access to structures such as the palate and the basi-

prepared free, providing optimal access to the basicranium

cranium that previously could not be viewed. The new mate-

and palate. As a practical matter, specimens in which the cra-

rial invites reexamination of some areas of anatomy that have

nium or cheek have been damaged and/or removed by tapho-

been less thoroughly studied than is desirable. For example,

nomic or other processes often provide access to deep struc-

222 dodson, you, & tanoue

Table 15.1. Specimen List Taxon

Repository

Origin

Basal length

Psittacosauridae Psittacosaurus major

CAGS-IG-VD-004

Yixian Fm., Lujiatun

P. lujiatunensis (juvenile)

CAGS-IG-VD-005

Yixian Fm., Lujiatun

192 mm 77 mm

Hongshanosaurus houi*

IVPP V 12617

Yixian Fm., Lujiatun

147 mm

Neoceratopsia Liaoceratops yanzigouensis (juvenile)

CAGS-IG-VD-002

Yixian Fm., Lujiatun

Liaoceratops yanzigouensis

IVPP V 12633

Yixian Fm., Lujiatun

56 mm** 77 mm

Liaoceratops yanzigouensis*

IVPP V 12738

Yixian Fm., Lujiatun

109 mm

Auroraceratops rugosus*

CAGS-IG-VD-2004-001

Xinminpu Group, Mazongshan

170 mm

Archaeoceratops oshimai*

IVPP V 11114

Xinminpu Group, Mazongshan

145 mm

*Designates type specimen. **Estimated basal length.

BASICRANIUM

tures that are otherwise difficult or impossible to see. This was the case with a small specimen of Liaoceratops (You et al. 2007) and with the large specimen of Psittacosaurus major (You et al.

The basicranium consists primarily of the basioccipital, basi-

in press). Additional observations have been made on speci-

sphenoid, and parasphenoid—elements that support the brain

mens of Protoceratops and Psittacosaurus at the American Mu-

from below. The basioccipital forms the central element of the

seum of Natural History and of Leptoceratops at the Canadian

occipital condyle, and forms a pendant structure ventral to the

Museum of Nature in Ottawa. It has long been appreciated

condyle that contacts the basisphenoid rostrally through the

that psittacosaurids are very distinct from the rest of cera-

basispheno-basioccipital synchondrosis.

topsians (Maryanska ´ and Osmolska 1975; Sereno 1990). This

Basioccipital. The basioccipital may be viewed either cau-

study will confirm this conclusion based on the structure of

dally or ventrally. In caudal (occipital) view, the basioccipi-

the basicranium and the palate.

tal of Psittacosaurus (Fig. 15.2A; see also Hongshanosaurus

Institutional Abbreviations. AMNH: American Museum of

Fig. 15.2B) presents as paired processes projecting ventrally

Natural History, New York; CAGS, IG: Institute of Geology,

slightly rostral to the occipital condyles. A broad median cleft

Chinese Academy of Geological Sciences, Beijing; CMN: Ca-

separates the two processes, which are rounded ventrally. By

nadian Museum of Nature, Ottawa; IVPP: Institute of Verte-

contrast, in the neoceratopsians Liaoceratops (Fig. 15.3A), Au-

brate Paleontology and Paleoanthropology, Beijing.

roraceratops (Fig. 15.3B), and Archaeoceratops (Fig. 15.3C), the

Anatomical Abbreviations. boc: basioccipital; bpt: basiptery-

basioccipital presents as a broad quadrilateral plate with

goid process of basisphenoid; bsp: basisphenoid; ch: choana;

the form of an inverted keystone, approximately twice the

cpr: crista prootica; crqp: cranioquadrate passage; ect: ectop-

breadth of the occipital condyle. A similar pattern is found in

terygoid; exo: exoccipital; mx: maxilla; oc: occipital condyle;

the neoceratopsians Protoceratops (Fig. 15.4A) and Lepto-

p: parietal; pl: palatine; pmx: premaxilla; ps: parasphenoid;

ceratops (Fig. 15.4B). The bottom of the plate is flat, and the

pt: pterygoid; ptc: central plate of pterygoid; ptm: mandi-

ventrolateral corners are angled obliquely. The sides of the

bular ramus of pterygoid; ptp: palatine ramus of pterygoid;

plate are directed dorsomedially towards the base of the par-

ptq: quadrate ramus of pterygoid; q: quadrate; qc: quad-

occipital processes dorsal to the occipital condyle. In ven-

rate condyle; qpt: pterygoid ramus of quadrate; r: rostral;

tral view in Psittacosaurus (Fig. 15.5A) and Hongshanosaurus

v: vomer.

(Fig. 15.5B), the transverse spheno-occipital synchondrosis is clearly visible. Basisphenoid. The occipital face of the basisphenoid is con-

Materials

gruent with the contiguous face of the basioccipital, and the

Specimens examined: Hongshanosaurus houi, IVPP V12617;

median groove is continued rostrally. In Liaoceratops (Fig.

Psittacosaurus lujiatunensis, CAGS-IG-VD-005; P. major, CAGS-

15.6) and Auroraceratops (Fig. 15.7A), the central plate of the

IG-VD-004; Archaeoceratops oshimai, IVPP V11114; Auroracera-

pterygoid covers the synchondrosis, which may not be ob-

tops rugosus, IG-2004-VD-001; Leptoceratops gracilis, NMC8889;

served. In Liaoceratops, paired caudal processes from the cen-

Liaoceratops yanzigouensis, IVPP V12633, IVPP V12738, CAGS-

tral plate of the pterygoid rest on the ventral aspect of the

IG-VD-002; Protoceratops andrewsi, AMNH6425. See Table 15.1.

basioccipital plate on either side of the midline; in Aurora-

Comments on the Basicranium and Palate of Basal Ceratopsians 223

FIGURE 15.2. Bilobate basioccipital of psittacosaurids. (A) Psittacosaurus major (CAGS-IG-VD-004) in caudal view; (B) Hongshanosaurus houi (IVPP V12617) in caudodorsal view. Scale bars are 5 cm.

ceratops, the paired processes from the pterygoid project caudally as free processes ventral to the occipital condyle, as they also do in Leptoceratops (Fig. 15.7B). Due to its fractured state in Archaeoceratops (Fig. 15.7C) it is not clear if in life the central plate of the pterygoids covers the spheno-occipital synchondrosis In any case, the central plate appears to be shorter than that of either Liaoceratops or Auroraceratops. The occipital and sphenoid are tightly appressed at the synchondrosis, and the faces of the two bones are congruent. In Psittacosaurus, the basipterygoid processes of the basisphenoid are elongate stout processes that span the space between the body of the basisphenoid and pockets at the rostral base of the quadrate wings of the pterygoids. Their orientation is largely horizontal, but they curve ventrally as they approach the pterygoids giving the basicranium of psittacosaurids an elongate, attenuated appearance. In a very large specimen of Psittacosaurus, P. major (CAGS-IG-VD-004), the basipterygoid processes appear to diverge only modestly from each other but this is due to the lateral compression of these processes. Between the processes is a deep midline cleft, which is continuous caudally with the

224 dodson, you, & tanoue

FIGURE 15.3. Caudal views of the occipital surfaces of the skulls of basal neoceratopsians showing the broad quadrilateral basioccipital. (A) Liaoceratops yanzigouensis (IVPP V12738); (B) Auroraceratops rugosus (IG-2004-VD-001); (C) Archaeoceratops oshimai (IVPP V11114). Scale bars are 5 cm.

PALATE The palate consists of pterygoids, ectopterygoids, palatines, vomers, premaxillae, and maxillae. In all taxa under consideration, the horizontal palatine processes of the premaxillae and maxillae meet their counterparts on the midline to form a short secondary palate rostral to the choanae. This surface is continued rostrally by the rostral bone. The secondary palate is situated dorsal to the cutting edge of the premaxilla and rostral. There are numerous differences among taxa. For example, the secondary palate is broader and rounded in psittacosaurids compared to narrow and more pointed in basal neoceratopsians, and is long relative to basal skull length in Archaeoceratops and short in juvenile Psittacosaurus lujiatunensis (CAGS-IG-VD-005). The vomer is the caudal component of the secondary palate in basal neoceratopsians, wedged in the midline between the maxillae (Osmólska 1986), but does not participate in the secondary palate in psittacosaurids. The ectopterygoid is applied to the caudodorsal aspect of the maxilla and forms a brace between the pterygoid and the medial surface of the jugal. However, the ectopterygoids are not well preserved in the psittacosaur specimens at hand, and hence will be omitted from further FIGURE 15.4. Caudal views of the occipital surfaces of the skulls

consideration.

of basal neoceratopsians showing the broad quadrilateral basioccipital. (A) Protoceratops andrewsi (AMNH6245). Note the strongly pendant basipterygoid processes of the basisphenoid, which project ventral to the basioccipital, an apomorphic feature of Protoceratops. (B) Leptoceratops gracilis (NMC 8889). Scale bars are 5 cm.

Pterygoid. The paired pterygoids, palatines, and the vomer represent the core of this comparative study. The pterygoid is a complex three-dimensional bone that consists of a flat central plate and three elongate rami: the quadrate ramus, the mandibular ramus, and the palatine ramus. In order to appreciate this complex bone, it needs to be viewed both caudally and ventrally. The central plate forms a midline joint of varying size, depending on the taxon. The differences between the pterygoids of psittacosaurids and those of basal neoceratop-

cleft between the basal tubera and rostrally with a cleft be-

sians are striking, although both present the same fundamen-

tween the pterygoids. In the small specimen of P. lujiatunensis

tal features. In the basal neoceratopsians, Liaoceratops (Fig.

(Fig. 15.5C, D; CAGS-IG-VD-005), the basipterygoid processes

15.6) and Auroraceratops (Fig. 15.7A), the central plate forms

diverge from each other at an angle of 60\, forming a broad

a long, horizontal midline joint between left and right coun-

central space between the basipterygoid processes and be-

terparts that extends as far caudally as the basispheno-

tween adjacent regions of the pterygoid. Basal neoceratop-

basioccipital synchondrosis, rostral to the occipital condyle.

sians lack this space as the basipterygoid processes can only be

In Archaeoceratops (Fig. 15.7C), the poorly preserved central

seen with difficulty, if at all, as they are covered ventrally by

plate may have been shorter, but was more caudally posi-

the central plate of the pterygoid. In Archaeoceratops (IVPP

tioned. By contrast, in psittacosaurids (Fig. 15.5), the central

V11114) the basipterygoid processes are visible (Fig. 15.7C)

plate forms a very short median joint between left and right

owing to fracture through the pterygoids. The processes are

counterparts that is very rostral in position, at approximately

remarkable for their small size, project rostroventrally from

the midpoint of the skull, at the level of the ventral rim of the

the body of the basisphenoid, constricted neck, and short,

orbit. In the juvenile Psittacosaurus lujiatunensis specimen (Fig.

divergent terminal prongs. As preserved, they do not appear

15.5C, D; CAGS-IG-VD-005), the central plate presents as a

to be in contact with the pterygoids. In Protoceratops the basi-

transverse bar in front of a wide interpterygoid-intersphenoid

pterygoid processes may be seen in caudal view as stout pegs

vacuity. The central plate lacks a flat, horizontal surface but

directed rostroventrally far beneath the basioccipital plate,

has the form of a transverse bar; the caudal face of the trans-

evidently reaching the pterygoids (Fig. 15.4A).

verse bar slopes rostroventrally, and the rostral face slopes ros-

Comments on the Basicranium and Palate of Basal Ceratopsians 225

FIGURE 15.5.

Basicranium and palate of psittacosaurids in ventral view. (A) Psittacosaurus major (CAGSIG-VD-004); (B) Hongshanosaurus houi (IVPP V12617); (C) Psittacosaurus lujiatunensis juvenile (CAGS-IG-VD-005); (D) reconstruction of P. lujiatunensis juvenile, with jaws removed and asymmetry caused by crushing eliminated. Scale bars are 5 cm. Visible are the bilobate basioccipitals (except for the juvenile Psittacosaurus lujiatunensis), elongate basipterygoid processes, short and centrally located interpterygoid joint, and prominent choanae separated by the vomer.

trodorsally towards the choanae. In the large specimen of

tween the quadrate ramus and its contact with the basioccipi-

P. major (Fig. 15.5A; CAGS-IG-VD-004), the interpterygoid-

tal. In Archaeoceratops (Fig. 15.7C; IVPP V11114), there may be

intersphenoid vacuity is narrow, and the central plate forms a

a small space between the quadrate ramus and the basioccipi-

ventrally convex tab rather than a bar. In Auroraceratops

tal, due to the modest development of the basipterygoid pro-

(IG-2004-VD-001; basal length 170 mm), the interpterygoid

cesses of the basisphenoid. In Liaoceratops (Fig. 15.6), it ap-

joint is approximately 30 mm long, while in the large Psit-

pears that there is a small separation between the basioccipital

tacosaurus major (CAGS-IG-VD-004; basal length 192 mm), the

and the quadrate ramus. In psittacosaurids (Fig. 15.5) there is a

length of the interpterygoid joint is 14 mm.

wide space between the quadrate ramus of the pterygoid and

The quadrate and mandibular rami of the pterygoid are gen-

the basioccipital that has no counterpart in basal neoceratop-

erally similar among all species of basal ceratopsians studied,

sians. This is a consequence of the fact that the quadrate rami

although proportions vary. The quadrate ramus projects

diverge rostral to the basioccipital. This arrangement is seen

caudolaterally, and overlaps broadly a similar lamina, the

clearly in the juvenile specimen of P. lujiatunensis (Fig. 15.5C,

pterygoid ramus, from the quadrate. The ventral edge of the

D; CAGS-IG-VD-005). Dorsally, however, the pterygoid rami

quadrate ramus is thickened, and in palatal view diverges

of the quadrates approach the side of the braincase, caus-

caudolaterally from the central plate towards the shaft of the

ing the gap between them to narrow. The pterygoid wing is

quadrate. In basal neoceratopsians, the quadrate ramus is

bowed slightly laterally, creating a gap between it and the

shorter in palatal view than in psittacosaurids. In Aurora-

prootic-opisthotic, called the cranioquadrate passage by Se-

ceratops (Fig. 15.7A; IG-2004-VD-001), there is no space be-

reno (1987), clearly visible in Figs. 15.5A, B, and C. The

226 dodson, you, & tanoue

FIGURE 15.6.

Basicranium and highly vaulted palate of Liaoceratops yanzigouensis (IVPP V12738) in ventral view. (A) Specimen is rotated slightly along its long access, exposing the slit-like right choana (left side of photograph); (B) drawing of the same specimen in true ventral view with the effects of postmortem distortion removed. The choanae are scarcely visible in true ventral view. Scale bar is 5 cm.

pterygoid wing contacts the braincase on a small thickening or

The palatine processes of the pterygoid rise as a pair of

prominence at the rostral end of the horizontal crista prootica

closely appressed rods oriented rostrodorsally from the cen-

on the opisthotic (Fig. 15.8A). The pterygoid wing of the quad-

tral plate of the pterygoid and terminating at approximately a

rate and its complement on the pterygoid form a curtain of

point halfway to the skull roof and at the midpoint of the

bone that obscures the ventral portions of the braincase in

orbit, through which they may be viewed. The palatine pro-

lateral view (Fig. 15.8B), including most of the cranial nerve

cesses are frequently broken somewhere along this trajectory.

foramina. Neither Auroraceratops nor Archaeoceratops preserve

In the juvenile Psittacosaurus lujiatunensis (CAGS-IG-VD-005)

the caudodorsal region of the skull well enough to determine

and in Auroraceratops (IG-2004-VD-001), the angle of rise of

whether a cranioquadrate passage is present.

the palatine process is shallow (Fig. 15.9B), roughly 30\,

The mandibular ramus of the pterygoid projects caudo-

whereas in the adult Psittacosaurus (CAGS-IG-VD-004) and in

ventrally from the caudal edge of the maxilla, and is typically

adult Liaoceratops (IVPP V12738) the angle is 60\ or more (Fig.

reinforced by the ectopterygoid, which also runs across the

15.9D). The juvenile specimen of Psittacosaurus (CAGS-IG-

caudodorsal edge of the maxilla and is braced against the me-

VD-005) shows the full complexity of the palatine process

dial surface of the jugal. The mandibular process is very short

through its relatively large orbit. The palatine process contacts

and poorly developed in juvenile P. lujiatunensis (Fig. 15.5C,

the vomer at a level near the rostral midpoint of the left orbit

D), broad and tab-like in the large P. major (Fig. 15.5A; CAGS-

(Fig. 15.9A, B). However, paired processes of the vomer rise

IG-VD-004) and Hongshanosaurus (Fig. 15.5B; IVPP V12617),

several mm above the vomer, and form paired laminae that

short and pointed in Archaeoceratops (Fig. 15.7C; IVPP

are directed caudally, forming an interorbital lamina that

V11114), and long and decurved in Auroraceratops (Fig. 15.7A;

stands above the central plate of the pterygoid. The caudal

IG-2004-VD-001). No pattern of phylogenetic significance is

ends of these laminae embrace the rostral end of the elongate

evident. Two important observations can be made: the lateral

parasphenoid. Contact between the pterygoids and the para-

edge of the mandibular ramus defines the medial edge of man-

sphenoid also occurs in Hongshanosaurus (IVPP V12617), and

dibular adductor fossa leading to the lower jaw, and there is a

is reported in P. lujiatunensis (Zhou et al. 2006). In these larger

significant embayment or passage between the quadrate

specimens, the orbit is relatively smaller and more dorsally

ramus and the mandibular ramus. This passage probably re-

positioned, and the palatine process is near the ventral margin

lates to the course of pterygoideus rostralis muscle that origi-

of the orbit. In the juvenile Liaoceratops, the pterygoids are

nates caudally from the dorsal surface of the palate, and con-

confined to the caudal end of the oral cavity, as is reported in

tinues ventrally over the mandibular process to reach the

Bagaceratops (Osmólska 1986). The palatine process does not

caudoventromedial surface of the mandible, as in modern

extend very far rostrally. In Auroraceratops, the palatine pro-

crocodilians. A comparison of the size of the mandibular ad-

cess extends rostrally to the level of the middle of the tooth-

ductor fossa with the pterygoid passage suggests that the vol-

row, but is broken rostral to this point. It cannot be deter-

ume of the pterygoid muscles was considerably inferior to that

mined in Archaeoceratops.

of the mandibular adductors in all basal ceratopsians.

Palatine. The palatines form lateral elements of the palate

Comments on the Basicranium and Palate of Basal Ceratopsians 227

FIGURE 15.7.

Basicranium and palate of basal neoceratopsians in ventral view. (A) Auroraceratops rugosus (IG-2004-VD-001); (B) detail of Leptoceratops gracilis (NMC 8889); (C) Archaeoceratops oshimai (IVPP V11114). The interpterygoid joint is long, caudally situated, and covers the basioccipitalbasisphenoid synchondrosis. Fracture of the interpterygoid central plate reveals the short, ventrally directed basipterygoid processes in Archaeoceratops. The complex structure of the pterygoid, with quadrate, mandibular, and palatine rami, is particularly evident in Auroraceratops, although the latter is broken rostrally. Scale bars are 5 cm.

and span the suborbital space between the central plate and

of the palatine coincides with the rostral rim of the orbit. A

the palatine process of the pterygoid and the maxilla. In basal

saddle-shaped transverse ridge defines the caudal border of the

neoceratopsians, the palatines contact the vomers on the mid-

choana (fenestra exochoanalis of Osmólska 1986; internal

line. In psittacosaurids, the palatines do not extend dorsally

nostril of Heaton 1972). The thickened lateral process, termed

far enough to contact the vomers, which are embraced instead

the transverse palatine wing by Osmólska (1986), contacts the

by the pterygoids. The palatine is clearly visible in both the

medial surface of the lacrimal, near the lacrimal-jugal junction

juvenile and adult specimens of Liaoceratops, and in the ju-

adjacent to the rostroventral corner of the orbit. The upturned

venile of Psittacosaurus lujiatunensis (CAGS-IG-VD-005). Por-

medial process of the saddle is termed the longitudinal pala-

tions of the palatine are also visible in Hongshanosaurus and

tine wing. Externally, the antorbital fossa is very prominent in

Auroraceratops. The palatines may also be seen in palatal view,

basal neoceratopsians but is absent in psittacosaurids. In Ar-

although they can be difficult to view in basal neoceratopsians

chaeoceratops and Auroraceratops, the antorbital fossa is fenes-

with closely positioned maxillae. In juvenile P. lujiatunensis

trated, and the transverse palatine wing roofs the canal that

the palatine is only visible dorsally through the orbit. On

leads to the fenestra. In the juvenile Liaoceratops (CAGS-IG-

the juvenile specimen of Liaoceratops (Fig. 15.10; CAGS-IG-

VD-002), the antorbital fossae are prominent, but evidence for

VD-002), the cranium has been removed perfectly exposing

actual fenestrae is ambiguous due to deficient preservation. In

the palatines in dorsal view. In all specimens, the rostral limit

the juvenile Psittacosaurus lujiatunensis (CAGS-IG-VD-005), a

228 dodson, you, & tanoue

FIGURE 15.8.

Skull of Psittacosaurus major 5(CAGS-IG-VD-004). (A) Detail of the left occipital region viewed caudodorsolaterally and showing the crista prootica, the cranioquadrate passage, and adjacent structures; (B) left lateral view of the same skull showing the fully developed pterygoid ramus of the quadrate hanging beneath the crista prootica and obscuring the braincase in lateral view.

prominent pocket or recess is visible beneath the transverse

counterpart near the midline to embrace the caudal end of the

palatine wing precisely where the antorbital fenestra would be

vomer, as figured in Bagaceratops (Fig. 15.1; Osmólska 1986).

expected, if one were present. It aligns well with the maxillary

Viewed ventrally, the palatines of basal neoceratopsians ap-

depression, stated by Sereno (2000) to be a neomorphic struc-

pear narrow and steeply inclined (longitudinally vaulted).

ture unrelated to the true antorbital fossa. The palatines ap-

Viewed dorsally however, the palatines of Liaoceratops and Au-

pear to be longer rostrocaudally in basal neoceratopsians

roraceratops give an entirely different impression. The pala-

(Liaoceratops, Auroraceratops) than they are in psittacosaurids,

tines, coupled with the dorsal surface of the maxilla (dorsal

consistent with relatively longer toothrows in the former. The

surface of the ‘‘box’’ enclosing the tooth roots), appear to

ventromedial surface of the maxilla is vertical in all taxa under

be predominantly horizontal, with only the medial edge

study. It forms an approximately rectangular box enclosing

rolled dorsally (Fig. 15.10). The dorsalmost point on the pal-

the roots of the teeth. Dorsally the maxilla forms a horizontal

atine of Liaoceratops is the vomerine process of the longi-

shelf in the suborbital space between the medial face of the

tudinal palatine wing. A prominent feature at the caudal

maxilla and the jugal. Viewed dorsomedially, the palatine

end of the palatine in Liaoceratops and Auroraceratops is the

contacts the medial surface of the maxilla as far ventrally as

postpalatine foramen, bounded by the ectopterygoid with or

the line of ‘‘special’’ dental foramina and as far caudally as the

without a contribution from the pterygoid. This foramen is

caudal end of the maxilla. Approximately four tooth positions

reported in several species of Psittacosaurus, including P. mei-

rostral from the caudal margin of the maxilla, the rostral bor-

leyingensis (Sereno et al. 1988), and P. lujiatunensis (Zhou et

der of the palatine bends directly dorsally to the top of the

al. 2006), but was not confirmed in the specimens in this

maxilla. From there the palatine bends arcs dorsomedially

study, and is perhaps less prominent and/or more variable in

forming the longitudinal palatine wing and approaches its

psittacosaurids.

Comments on the Basicranium and Palate of Basal Ceratopsians 229

FIGURE 15.9. Comparison of the palate and basicranium in (A, B) Psittacosaurus lujiatunensis (CAGS-IG-VD-005; juvenile) and (C, D) Liaoceratops yanzigouensis (IVPP V12738), a basal neocerotopsian. (A) In ventrolateral oblique view showing the palatine, palatine process of pterygoid, and parasphenoid; (B) schematic sagittal section of skull. Note the horizontal vomer, enlarged choana, limited rostrocaudal extent of the pterygoid, restricted palatine, elongated horizontal basipterygoid process, and ventrally exposed basioccipital. (C) Photograph in reversed right view; (D) schematic sagittal section of same. Note the sloping vomer, restricted size of the choana, elevated palatine, caudally expanded pterygoid, short basipterygoid process, and nearly covered basioccipital. Scale bars are 5 cm.

Vomer. Vomers are very often the casualty of poor preser-

2006) but the term is never usefully explained, nor are fig-

vation or of poor preparation. Vomers are preserved in the

ures presented in palatal view sufficient to confirm vaulting.

juvenile P. lujiatunensis (Fig. 15.5C, D, Fig. 15.9B; CAGS-IG-

Osmólska (1986) described the palate in lateral view of Baga-

VD-005), Hongshanosaurus (Fig. 15.5B; IVPP V12617), and Lia-

ceratops as vaulted both longitudinally and transversely. She

oceratops (Fig. 15.6, Fig. 15.9C, D; IVPP V12738; type speci-

indicated that the longitudinal vaulting of the palate in Baga-

men), but not in the large specimen of Psittacosaurus major

ceratops is produced by the caudodorsal inclination of the

(CAGS-IG-VD-004), small Liaoceratops (CAGS-IG-VD-002), Ar-

vomer coupled with the rostrodorsal inclination of the longi-

chaeoceratops (IVPP V11114), or in Auroraceratops (IG-2004-

tudinal palatine wing, This configuration is also found in Liao-

VD-001). The vomers fuse to form the median bar of the palate

ceratops (Fig. 15.9D; CAGS-IG-VD-002); rostrally the vomer is

(vomerine bar of Osmólska 1986) on the midline between the

at a level equivalent to the external maxillary ridge, rises

palatine process of the pterygoids (psittacosaurids) or the lon-

caudodorsally to a vertical level equivalent to the middle of

gitudinal palatine wing of the palatine (basal neoceratop-

the orbit. In the psittacosaur specimens examined for this

sians) and the median intermaxillary joint of the secondary

study, the vomer is not vaulted as previously claimed, but

palate. The vomer separates the choanae from each other. The

appears to be horizontally positioned in both Hongshanosau-

vomer is very gracile in Liaoceratops, but the bar is more robust

rus (IVPP V12617) and in juvenile Psittacosaurus lujiatunensis

(up to 7 mm in breadth), in Hongshanosaurus. The vomer ex-

(Fig. 15.9A, B; CAGS-IG-VD-005). In Hongshanosaurus (IVPP

pands at each end, contacting the median intermaxillary joint

V12617), the rostral end of the vomer rests on the dorsal sur-

rostrally, and the pterygoid or the palatine caudally. In psit-

face of the median intermaxillary joint at a level dorsal to the

tacosaurids the vomer is invariably characterized as vaulted

external maxillary ridge in the ventral part of the circumnarial

(Sereno 1987; Sereno et al. 1988; You and Xu 2005; Zhou et al.

depression. The vomer extends caudally through the horizon

230 dodson, you, & tanoue

lujiatunensis (Fig. 15.5C, D; CAGS-IG-VD-005) the choanae are fully exposed in palatal view as large elliptical openings into the oral cavity. In these specimens the rostral end of the oral cavity is much broader relative to tooth row length than in basal neoceratopsians. In Hongshanosaurus, the relatively enormous choanae measure 30–32 mm in length and 13–17 mm in breadth, and essentially comprise the roof of the entire oral cavity (Fig. 15.5B). In juvenile Psittacosaurus lujiatunensis (Fig. 15.5C, D) the choanae are also relatively large, but they do not extend as far caudally as they do in Hongshanosaurus. In basal neoceratopsians, the course of airflow from the nares (fenestra exonarina of Osmólska 1986) through the nasal cavity to the choanae was at a relatively low angle to the horizontal, 30\ or less. In the juvenile P. lujiatunensis (CAGS-IG-VD-005) the axis of airflow was 45\, whereas in large P. major (CAGS-IG-VD-004) and other specimens of Psittacosaurus with deep skulls and dorsally positioned nares, airflow followed a vertical dorsal to ventral path through the nasal cavity.

Discussion Basal ceratopsians have been known to science for nearly a century. The field remained rather static until 1975, with the description of Bagaceratops rozhdestvenskyi and of other basal neoceratopsian materials from Mongolia by Maryanska ´ and A very small juvenile specimen of Liaoceratops yanzigouensis (CAGS-IG-VD-002) in dorsal view. Taphonomic removal of the skull roof reveals the long horizontal palatine and adjacent palatal elements. Rostrally the transverse palatine wing defines the caudal border of the choana.

Osmólska (1975). In the past decade the field has literally ex-

FIGURE 15.10.

ploded, particularly, but not exclusively, with discoveries of new materials from China. New taxa have been discovered as well as new material from previously known taxa. Moreover, the anatomical and taxonomic value of specimens have been increased by improved techniques of skilled preparation that

tal plane and terminates rostroventral to the orbit. In the juve-

have minimized or eliminated the problem of matrix obscur-

nile P. lujiatunensis, the vomer is likewise horizontal, but is

ing critical details. The resulting specimens are fragile and del-

more dorsal in position, and terminates within the orbit adja-

icate, yet are rich in informative anatomical detail unavailable

cent to its rostral margin. In both cases, the rostral end of the

to previous workers.

vomer does not contribute to the formation of the secondary

In this preliminary report we review the anatomy of several recently collected specimens of psittacosaurids and of basal

palate but lies well dorsal, in a subnarial position. Choanae. The choanae (internal nostrils) connect nasal and

neoceratopsians (Table 15.1). The palate has only rarely been

oral cavities. The choanae are bounded rostrally and laterally

the topic of focus in dinosaur studies, the important excep-

by the maxillae, medially by the vomers, caudally by pala-

tions being Heaton’s (1972) study of the hadrosaur palate and

tines,

the

Osmólska’s (1986) study of the palate of ‘‘protoceratopsids.’’

pterygoids. The most complete basal neoceratopsian speci-

In describing anatomical regions that have been poorly sur-

men for the rostral palate in this study is the adult Liaoceratops

veyed in the past such as the basicranium and the palate, we

(Fig. 15.6; IVPP V12738). When viewed in palatal view, the

have found significant differences between psittacosaurids

choanae are visible as narrow slits, because their orientation is

and basal neoceratopsians (Table 15.2). These underscore the

steep (20\ from the vertical in Bagaceratops). When the skull is

autapomorphies of psittacosaurids. Studies of basal ornitho-

rotated to a dorso-oblique view, the choanae appear as tri-

pods such Hypsilophodon (Galton 1974) and of pachycephalo-

angular openings opening into the roof of the oral cavity situ-

saurians (Maryanska ´ et al. 2004) demonstrate the polarity of

ated between the first and sixth maxillary teeth and measur-

relevant character states of basal ceratopsians. Such psittaco-

ing 14 mm (left) to 17 mm (right) in length. By contrast, in

saurid characters as bilobate basioccipital; elongate, hori-

Hongshanosaurus (Fig. 15.5B; IVPP V12617) and Psittacosaurus

zontal basipterygoid processes of basisphenoid, horizontal

and,

in

psittacosaurids,

caudomedially

by

Comments on the Basicranium and Palate of Basal Ceratopsians 231

Table 15.2. Summary of Some of the Major Differences in Basicranium and Palate between Psittacosaurids and Basal Neoceratopsians Psittacosauria

Basal Neoceratopsia

Basioccipital bilobate

Basioccipital quadrilateral

Basioccipital—basisphenoid syn-

Basioccipital—basisphenoid syn-

chondrosis exposed ventrally

chondrosis covered by pterygoids

Basipterygoid processes of

Basipterygoid processes of

sphenoid long and directed

sphenoid short and ventrally

rostrally

directed

Central plate of pterygoid rostral, interpterygoid suture

Central plate of pterygoid caudal, interpterygoid suture long

an elaborate inverted-U course for the air passages in the nasal vestibule (a sand particle trap?). Much remains to be studied. These issues may be explored further on another occasion. Acknowledgments

We dedicate this paper to the memory of Halszka Osmólska (1930–2008), who died March 31, 2008. Dr. Osmolska was in every sense a pioneer in the study of basal neoceratopsians. Her 1986 paper on the palate was the frank inspiration for the present work. She was a true friend of dinosaurs and is sorely missed. We are grateful to Ji Qiang and Xu Xing for allowing access to specimens at Institute of Geology Chinese Academy of Geo-

short Vomer horizontal, rostral end

sideration is that a ventrally located fleshy nostril would entail

Vomer inclined caudodorsally,

logical Sciences (CAGS-IG) and the Institute of Vertebrate

terminates dorsal to secondary

rostral end contributes to sec-

Paleontology and Paleoanthropology, Beijing (IVPP). Peter

palate

ondary palate

Dodson thanks his chairman, Dr. Narayan Avadhani, for sup-

Choanae very large, occupy

Choanae reduced, situated at the

much of oral cavity, easily vis-

rostral end of oral cavity, diffi-

ible ventrally due to horizon-

cult to see ventrally due to

tal orientation

steep inclination

port. Funding was provided by the Ministry of Science and Technology of China (973 Project: 2006CB701405), the National Natural Science Foundation of China (40672007), and the Hundred Talents Project of Ministry of Land and Resources of China to You Hai-Lu. Kyo Tanoue was funded by Summer Research Stipends in Paleontology (University of

vomer, and enlarged choanae are unique to psittacosaurids

Pennsylvania), School of Arts and Sciences Dissertation

and highlight an evolutionary radiation parallel to that of the

Research Fellowship (University of Pennsylvania), and Juras-

neoceratopsians.

sic Foundation Research Grant. We thank reviewers Michael

As Witmer (2001) notes, the nasal region plays a critical role

Ryan, Matt Vickaryous, and Larry Witmer for helpful reviews,

in respiration, olfaction, thermal physiology, and even repro-

granted that deficiencies remain. We are grateful to sym-

ductive biology of animals. Soft tissues play a crucial role in

posium organizers and volume editors Michael Ryan, David

these functions. Unfortunately there is as of yet no evidence

Eberth, and Brenda Chinnery-Allgeier. Finally, we thank bibli-

of turbinal structures or conchae in basal ceratopsians. Of sig-

ographer extraordinaire Jerry Harris for his help.

nificance is the nasal region of psittacosaurids. Witmer (2001) has made a strong case for the near universality of the placement of the fleshy nostrils in a rostroventral position near the oral margin in vertebrates generally, and in dinosaurs in particular. He notes that many dinosaurs, including macronarian sauropods, hadrosaurine hadrosaurids, and ceratopsids, show enlargement of the bony nares, which were possibly occupied in part by cavernous soft tissues of the nasal vestibular vascular complex. However, psittacosaurids seem more difficult to reconcile with Witmer’s conclusions. The nostrils are generally small, often circular, and always are dorsal in position near the level of the skull roof. There would seem to be relatively little space for cavernous tissue within the bony naris, but perhaps some could fit in the shallow, poorly defined circumnarial depression. In any case, it seems unlikely that the fleshy nostril would have been projected downward towards the oral margin for two reasons. One is that the extensive beak presumably was keratinized (Russell and Zhao 1996; Sereno 1990) and would not have been tunneled by a fleshy nostril. Avoiding the keratinous beak, the fleshy nostril would then be improbably situated caudal to the bony naris. The other con-

232 dodson, you, & tanoue

References Cited Averianov, A. O., A. V. Voronkevich, S. V. Leshchinskiy, and A. V. Fayngertz. 2006. A ceratopsian dinosaur Psittacosaurus sibiricus from the Early Cretaceous of West Siberia, Russia and its phylogenetic relationships. Journal of Systematic Palaeontology 4: 359–395. Bohlin, B. 1953. Fossil reptiles from Mongolia and Kansu. Reports from the Scientific Expedition to the North-western Provinces of China Under Leadership of Dr. Sven Hedin. Publication 37: 1–113. Brown, B. 1914. Leptoceratops, a new genus of Ceratopsia from the Edmonton Cretaceous of Alberta. American Museum of Natural History Bulletin 33: 567–580. Chinnery, B. J. 2004. Description of Prenoceratops pieganensis gen. et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Dong, Z., and Y. Azuma. 1997. On a primitive neoceratopsian from the Early Cretaceous of China. In Z. Dong, ed., SinoJapanese Silk Road Dinosaur Expedition, pp. 68–89. Beijing: China Ocean Press. Galton, P. M. 1974. The ornithischian dinosaur Hypsilophodon

from the Wealden of the Isle of Wight. Bulletin of the British Museum (Natural History) Geology 25: 1–152. Granger, W., and W. K. Gregory. 1923. Protoceratops andrewsi, a pre-ceratopsian dinosaur from Mongolia. American Museum Novitates 72: 1–9. Heaton, M. J. 1972. Palatal structure of some Canadian Hadrosauridae (Reptilia: Ornithischia). Canadian Journal of Earth Sciences 9: 185–205. Makovicky, P. J. 2001. A Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) braincase from the Horseshoe Canyon Formation of Alberta. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 243–262. Bloomington: Indiana University Press. Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Maryanska, ´ T., R. E. Chapman, and H. Osmólska. 2004. Pachycephalosauria. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 464–477. Berkeley: University of California Press. Maryanska, ´ T., and H. Osmólska. 1975. Protoceratopsidae (Dinosauria) of Asia. Palaeontologia Polonica 33: 133–182. Osborn, H. F. 1923. Two Lower Cretaceous dinosaurs of Mongolia. American Museum Novitates 95: 1–10. Osmólska, H. 1986. Structure of nasal and oral cavities in the protoceratopsid dinosaurs (Ceratopsia, Ornithischia). Acta Palaeontologica Polonica 31: 145–157. Russell, D. A., and X. Zhao. 1996. New psittacosaur occurrences in Inner Mongolia. Canadian Journal of Earth Sciences 33: 637– 648. Sereno, P. C. 1987. The ornithischian dinosaur Psittacosaurus from the Lower Cretaceous of Asia and the relationships of the Ceratopsia. Ph.D. diss., Columbia University, New York. ———. 1990. Psittacosauridae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 579–592. Berkeley: University of California Press. ———. 2000. The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia. In M. J. Benton, M. A. Shishkin, D. M. Unwin, and E. N. Kurochkin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. 480–516. Cambridge: Cambridge University Press Sereno, P. C., and S. Chao. 1988. Psittacosaurus xinjiangensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northwestern China. Journal of Vertebrate Paleontology 8: 353–365. Sereno, P. C., S. Chao, Z. Cheng, and C. Rao. 1988. Psittacosaurus meileyingensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northwestern China. Journal of Vertebrate Paleontology 8: 366–377. Sereno, P. C., X. Zhao, L. Brown, and T. Lin. 2007. New psittaco-

saurid highlights skull enlargement in horned dinosaurs. Acta Paleontologica Polonica 52: 275–284. Witmer, L. M. 2001. Nostril position in dinosaurs and other vertebrates and its significance for nasal function. Science 293: 850–853. Xu, X. 1997. A new psittacosaur (Psittacosaurus mazongshanensis sp. nov.) from Mazongshan area, Gansu Province, China. In Z. Dong, ed., Sino-Japanese Silk Road Dinosaur Expedition, pp. 48–67. Beijing: China Ocean Press. Xu, X., C. A. Forster, J. M. Clark, and J. Mo. 2006. A basal ceratopsian with transitional features from the Late Jurassic of northwestern China. Proceedings of the Royal Society of London B 273: 2135–2140. Xu, X., P. J. Makovicky, X. Wang, M. A. Norell, and H. You. 2002. A ceratopsian dinosaur from China and the early evolution of Ceratopsia. Nature 416: 314–317. You, H., and P. Dodson. 2003. Redescription of neoceratopsian dinosaur Archaeoceratops and early evolution of Neoceratopsia. Acta Paleontologica Polonica 48: 261–272. ———. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press. You, H., D. Li, Q. Ji, M. C. Lamanna, and P. Dodson. 2005. On a new genus of basal neoceratopsian dinosaur from the Early Cretaceous of Gansu Province, China. Acta Geologica Sinica 79: 593–597. You, H., K. Tanoue, and P. Dodson. 2007. A new specimen of Liaoceratops yanzigouensis (Dinosauria: Neoceratopsia) from the Early Cretaceous of Liaoning Province, P. R. China. Acta Geologica Sinica 81: 898–904. ———. In press. A new cranial specimen of the ceratopsian dinosaur Psittacosaurus major from the Early Cretaceous Yixian Formation of Liaoning Province, China. Acta Paleontologica Polonica 53. You, H., and X. Xu. 2005. An adult specimen of Hongshanosaurus houi (Dinosauria: Psittacosauridae) from the Lower Cretaceous of western Liaoning Province, China. Acta Geologica Sinica 79: 168–173. You, H., X. Xu, and X. Wang. 2003. A new genus of Psittacosauridae (Dinosauria: Ornithopoda) and the origin and early evolution of marginocephalian dinosaurs. Acta Geologica Sinica 77: 15–20. Zhao, X., Z. Cheng, and X. Xu. 1999. The earliest ceratopsian from the Tuchengzi Formation of Liaoning, China. Journal of Vertebrate Paleontology 19: 681–691. Zhao, X., Z. Cheng,, X. Xu, and P. J. Makovicky. 2006. A new ceratopsian from the Upper Jurassic Houcheng Formation of Hebei, China. Acta Geologica Sinica 80: 467–473. Zhou, C., K. Gao, R. C. Fox, and S. Chen. 2006. A new species of Psittacosaurus (Dinosauria: Ceratopsia) from the Early Cretaceous Yixian Formation, Liaoning, China. Palaeoworld 15: 100–114.

Comments on the Basicranium and Palate of Basal Ceratopsians 233

16 Mandibular Anatomy in Basal Ceratopsia K Y O TA N O U E , H A I - L U Y O U , A N D P E T E R D O D S O N

numerous discoveries of new basal ceratopsian taxa

Introduction

with superb skulls in the past 10 years now make it possible to document the evolution of ceratopsian jaw struc-

The Ceratopsia is known for its unique forms of mastication.

ture. Further preparation of the specimens after the

Derived ceratopsians (ceratopsids) have jaws with large num-

initial descriptions, especially involving the detachment

bers of teeth that are compressed for close packing in dental

and removal of mandibles from the skulls, has also re-

batteries (Ostrom 1964; Dodson 1996). The basal ceratop-

vealed new information that was not accessible in initial

sians, on the other hand, are known to lack dental batteries,

studies. In this study, the anatomy of mandibles in basal

with one or two replacement teeth erupting in a line. To grasp

ceratopsians has been examined. Some basal cera-

the evolution of the jaw structure of ceratopsians, detailed

topsians have ventrally convex ventral mandibular mar-

observation of the skulls of basal ceratopsians is necessary. Un-

gins and labially concave tooth rows unlike ceratopsids,

til recently, only a few genera of basal ceratopsians included

which implies diversity in the early evolution of the

well-preserved skulls, including Leptoceratops (Brown 1914),

Ceratopsia. Within basal Ceratopsia, three groups can be

Protoceratops (Granger and Gregory 1923) and Psittacosaurus

recognized based on the morphology of the mandible:

(Osborn 1923). Recent discoveries of the skulls of basal cera-

basalmost ceratopsians including Chaoyangsaurus and

topsians, often in superb preservation, include the new gen-

Yinlong, psittacosaurids, and basal neoceratopsians.

era Archaeoceratops (Dong and Azuma 1997), Chaoyangsaurus

These groups conform to recent cladistic analyses. Struc-

(Zhao et al. 1999), Liaoceratops (Xu et al. 2002), Hongshano-

tures of the jaw joint differ among the three groups. The

saurus (You et al. 2003), Magnirostris (You and Dong 2003),

portion of the mandible caudal to the coronoid process is

Lamaceratops, Platyceratops (Alifanov 2003), Prenoceratops

relatively longer in basalmost ceratopsians and psit-

(Chinnery 2004), Auroraceratops (You et al. 2005), Yinlong (Xu

tacosaurids than in basal neoceratopsians. This mor-

et al. 2006), Xuanhuaceratops (Zhao et al. 2006), Yamaceratops

phological difference reflects a smaller contribution of

(Makovicky and Norell 2006), and Cerasinops (Chinnery and

surangular and angular to the lateral surface of the man-

Horner 2007). All of these genera have been described since

dible in the latter. Transverse thickening of the mandible

1997, making it possible now to study the evolution of cera-

in basal neoceratopsians is associated with medial dis-

topsian jaw structure. However, characters listed in previous

placement of the tooth row, indicating different feeding

descriptions of basal ceratopsian taxa are insufficient for the

adaptations within basal Ceratopsia.

detailed study of the masticatory apparatus, since such char-

234

Description

acters are often chosen for phylogenetic analysis and not for functional morphology. Articulated skulls and mandibles of

CHAOYANGSAURUS

type specimens are seldom separated when they are first documented. Further preparation of the specimens after the ini-

Chaoyangsaurus youngi comes from the Lower Cretaceous

tial descriptions, especially involving the detachment and

Tuchengzi Formation in Liaoning Province, China (Zhao et al.

removal of mandibles from the skulls, has revealed new in-

1999; Swisher et al. 2002). The holotype (IGCAGS V371) is the

formation that was not previously accessible. In this study,

only known specimen of the species. Sutures between the ele-

the anatomy of the mandibles in basal ceratopsians has been

ments are obliterated. Although the caudal half of the left mandible is missing,

examined. Institutional Abbreviations. AMNH: American Museum of

the right mandible is mostly complete (Fig. 16.1A, B). The

Natural History, New York; CAGS, IG, IGCAGS: Institute of

coronoid and prearticular are not preserved. The left and

Geology, Chinese Academy of Geological Sciences, Beijing;

right mandibles diverge caudally at an angle of approximately

IVPP: Institute of Vertebrate Paleontology and Paleoanthro-

60\ (Fig. 16.1A). In dorsal view, the right mandible is nearly

pology, Beijing; NMC: Canadian Museum of Nature, Ottawa;

straight with very little change in width except for at the cau-

PKUP: Peking University Paleontological Collections, Beijing.

dal end. The tooth row is horizontal, and the straight ventral

Anatomical Abbreviations. an: angular; ar: articular; cor: cor-

border of the mandible caudal to the rostral extends slightly

onoid; d: dentary; pd: predentary; pra: prearticular; sa: sur-

caudoventrally (Fig. 16.1B). The dorsally convex ventral bor-

angular; sp: splenial; vlf: ventrolateral flange of mandible.

der of the mandible ascends caudodorsally. The caudodorsal border of the mandible is nearly straight and slopes slightly caudally from the apex of the coronoid process. The lateral

Materials and Methods SPECIMENS

ridge of the mandible is undeveloped. A flange extends ventrally in the mandible of Chaoyangsaurus (Zhao et al. 1999). It extends from the middle of the ventral border to the middle of

Chaoyangsaurus youngi, IGCAGS V371; Yinlong downsi, IVPP

the caudoventral border. The rostral end of the flange is con-

V14530; Hongshanosaurus houi, IVPP V12617; Psittacosaurus

fluent with the ventral border of the dentary. The rostral half

lujiatunensis, PKUP 1053, PKUP 1054, PKUP 1060; P. major,

of the flange runs on the medial side of the mandible, but it

CAGS-IG-VD-004; P. mongoliensis, AMNH 6254; P. neimon-

curves dorsolaterally to reach the lateral side of the mandible

goliensis, IVPP V12-0888-2; P. sinensis, IVPP V738; Archaeo-

caudally. M. pterygoideus (MPT) would have attached on the

ceratops oshimai, IVPP V11114; Auroraceratops rugosus, IG-

medial surface of this flange. On the contrary, the ventrolat-

2004-VD-001; Leptoceratops gracilis, NMC 8889; Liaoceratops

eral flange of psittacosaurid mandibles develops along the lat-

yanzigouensis, IVPP V12738, IVPP V12633, CAGS-IG-VD-002;

eral border of the mandible in ventral view, obliterating the

Protoceratops andrewsi, AMNH 6418, AMNH 6425, AMNH

rostral border of the attachment of the MPT. The mandible of

6429, AMNH 6441, AMNH 6460, AMNH 6467.

Chaoyangsaurus lacks an external mandibular fenestra (Zhao et al. 1999). The right mandibular fossa is 41 mm long and

MEASUREMENTS

18 mm high. In dorsal view, the dorsal margin of the predentary is semi-

Mandibles and associated elements of nine basal ceratopsian

circular as in Psittacosaurus (Fig. 16.1A). Unlike in Psittacosau-

genera were observed and measured. In this study, Chaoyang-

rus, however, the dorsal border of the Chaoyangsaurus preden-

saurus and Yinlong are attributed to basalmost Ceratopsia.

tary ascends rostrodorsally (Fig. 16.1B). Although partially

Hongshanosaurus and Psittacosaurus comprise Psittacosauri-

missing, the ventral border of the predentary appears to curve

dae. Archaeoceratops, Auroraceratops, Leptoceratops, Liaocera-

slightly in lateral view. The preserved rostral tip is a few milli-

tops, and Protoceratops represent basal Neoceratopsia. The

meters above the level of the dentary teeth. In ventral view,

lengths of the dentary and associated elements were mea-

the preserved bilobate portion of the ventral process, 13 mm

sured along the long axis of each mandible, excluding the

wide on each side, stretches a few millimeters caudally.

predentary. Widths were measured in the horizontal plane

Although the outline is unclear, the dentary seems to cover

perpendicular to the long axis of each element. Heights

more than half of the lateral surface of the mandible (Fig.

were measured perpendicular to the first two dimensions. The

16.1B). The mid-portion of the dentary is convex laterally. The

measurements of the mandibles, mandibular elements, and

coronoid process is located immediately caudolateral to the

tooth rows are shown in Table 16.1. Detailed description of

last tooth row. The apex, which is part of the dentary, is al-

the various mandibles provides the information necessary

most at the same level with the dorsalmost point of the sur-

for this and future studies, and is the primary purpose of

angular, if not higher. This feature is similar to that of Yinlong

this paper.

(Fig. 16.1C–E) and different from those of psittacosaurids and

Mandibular Anatomy in Basal Ceratopsia 235

Table 16.1. Selected Measurements of Basal Ceratopsian Specimens Cy Element

l

lf mandible

w

Yd h

l

w

175

Hh h

l

w

Ao h

l

w

141 32

60 135 35

r mandible

120 22 38 165 35 ]45 141 33

60 141 29

predentary

27 26

48 24

Ar h

l

w

Lg h

l

58 171 45 62

h

78 126

54 170 49 59 25

w

Ly w

h

56

l

w

Pa2 h

l

w

h

95 23 46 ]232 51 111 272 58 112

77 117 112 23 46 180 84

lf dentary

290 67 122

25 19 25

78

r dentary

77

lf surangular

83

48

r surangular

87

78

lf angular

96

50

r angular

92

43

lf articular

21

r articular

20

36 31

lf coronoid

13 ]13

48 35

14 11 10

48

14

22

10

22

r coronoid

]15

]16

50

lf splenial

]62

34

167

68

186

67

r splenial

l

Pa1

]54

7

92

lf prearticular

]53

r prearticular

]62

]20

17 18

]26

lf tooth row

39

38

55

62

154

50

93

r tooth row

46

41

53

63

152

49

95

Note: Measurements in millimeters. See Materials and Methods for dimensions. Abbreviations: Ao: Archaeoceratops oshimai (IVPP V11114); Ar: Auroraceratops rugosus (IG-2004-VD-001); Cy: Chaoyangsaurus youngi (IGCAGS V371); Hh: Hongshanosaurus houi (IVPP V12617); Lg: Leptoceratops gracilis (NMC8889); Ly: Liaoceratops yanzigouensis (IVPP V12738); Pa1: Protoceratops andrewsi (AMNH6429); Pa2: P. andrewsi (AMNH6460); Yd: Yinlong downsi (IVPP V14530); lf: left; r: right. l = length; w = width; h = height.

basal neoceratopsians in that in the latter two groups the sur-

saurus type specimen. This long thin bone on the medial sur-

angular is generally lower than the apex of the coronoid pro-

face of the mandible extends from below the 6th dentary

cess. The ventral border of the dentary in Chaoyangsaurus

tooth to the middle of the ventral border of the mandibu-

forms the rostral half of the ventral flange.

lar fossa. The rostral end is 13 mm caudal to the mandibular

The surangular contributes to the caudal portion of the cor-

symphysis.

onoid process in Chaoyangsaurus (Fig. 16.1B). The caudodorsal

The tooth rows are concave labially as in basal neoceratop-

border slopes only slightly caudally for about 55 mm unlike in

sians, running along the medial border of the dentary in dorsal

other basal ceratopsians except for Yinlong. The surangular-

view (Fig. 16.1A). The right tooth row is composed of 11 teeth,

angular sutures are obliterated and the outline of the surangu-

whereas there are only 9 teeth preserved in the left tooth row.

lar is unclear. The angular forms the caudoventral part of the mandible.

YINLONG

The ventral margin ascends caudodorsally at an angle of approximately 140\ from the ventral margin of the dentary (Fig.

Yinlong downsi is the oldest currently known ceratopsian. It

16.1B). The rostroventral portion of the angular possibly com-

comes from the Upper Jurassic Oxfordian of the Shishugou

prises most of the ventral flange of the mandible, unlike in

Formation in Xinjiang Province, China (Xu et al. 2006). The

psittacosaurids.

holotype (IVPP V14530) is the complete skeleton of a single

The articular contributes to the caudal end of the right man-

individual.

dible. The articular is wider than long, unlike that of other

The mandible is articulated with the skull. The right man-

basal ceratopsians observed (Fig. 16.1A). It extends medial to

dible is almost complete and the left mandible is missing the

the mandibular ramus as in basal neoceratopsians. In dorsal

lateral surface and apex of the coronoid process (Fig. 16.1C).

view, medial and lateral cotyles are 5 mm and 8 mm in width,

Caudal to the mandibular symphysis, the mandibular rami

respectively. The short retroarticular process is only a few mil-

are relatively wide, expanding both laterally and medially and

limeters long.

reaching maximum widths of 24 mm in the left mandible and

The splenial is partially preserved on the left mandible and

21 mm in the right, slightly rostral to the coronoid process.

seems almost complete in the right mandible of the Chaoyang-

Caudally from this point, the mandible tapers until it expands

236 tanoue, you, & dodson

FIGURE 16.1. Mandibles of basalmost ceratopsians. Chaoyangsaurus youngi (IGCAGS V371) in (A) dorsal and (B) right lateral views; Yinlong downsi (IVPP V14530) in (C) ventral and (D, E) right ventrolateral views. (A–D) photographs; (E) interpretive outlines. Scale bar is 5 cm.

medially for the jaw articulation. A possible external mandib-

The angular of Yinlong is about two-thirds the size of the

ular fenestra is located at the dentary-surangular-angular

surangular in lateral view (Fig. 16.1D, E). The rostral half

junction. The following description is mainly based on the

of the lateral surface of the angular is rugose. The rugosity

right mandible.

merges caudally with a lateral ridge extending from the ven-

The predentary of Yinlong is bluntly keeled. In lateral view,

tral margin. The surface of this ridge is smooth. The ventral

the ventral border descending caudoventrally is straight (Fig.

margin extends laterally with the greatest extension at the

16.1D, E). The caudolateral process is narrow, only 3 mm in

middle of the angular, forming a laterally convex ventro-

height. The left caudolateral process is 10 mm long and right

lateral border in ventral view (Fig. 16.1C). A low ridge runs

7 mm long as preserved. The ventral process is partially miss-

caudolaterally on the rostral two-thirds of the ventral sur-

ing and is at least 13 mm wide. Impressions for articulation

face, indicating the attachment of the MPT caudomedial

with the predentary on the rostral ends of dentaries are ap-

to it.

proximately 5 mm wide on both sides (Fig. 16.1C).

The splenial is not fully preserved. The preserved rostral end

The dentary is not as deep as that of other ceratopsians

is 11 mm long, and it is located caudal to the caudoventral end

(Fig. 16.1D, E). Its surface is smooth both on the lateral and

of the mandibular symphysis (Fig. 16.1C). It is about one-

medial sides. The apex of the coronoid process is slightly

third the width of the mandibular ramus.

higher than the dorsalmost point of the surangular. The lateral surface of the surangular is mostly smooth (Fig. 16.1D). A protuberance extends laterally from the jaw joint

The articular is not well preserved and is only observable in caudal view. It extends medially from the caudal end of the surangular.

(Fig. 16.1C). In addition, a second, lower protuberance is lo-

Dentary teeth are exposed on the rostral part of the right

cated about 10 mm rostral to the first one in the right man-

tooth row, and one erupting tooth is present on the left den-

dible (Xu et al. 2006). The dorsal border of the surangular

tary. The dentary tooth rows are inset medially, but it is un-

is slightly convex dorsally and slopes gently caudoventrally

certain if they are running along the medial border of the

(Fig. 16.1D, E). A very low ridge extends along the ventral

dentary. Labially concave maxillary tooth rows indicate the

margin of the surangular.

dentary tooth rows are likewise.

Mandibular Anatomy in Basal Ceratopsia 237

HONGSHANOSAURUS

The splenial is preserved only in the left mandible of Hongshansaurus, and lacks the caudal ramus. Dorsal and rostral

Hongshanosaurus houi comes from the Lower Cretaceous Yi-

rami cover the medial side of dentary, with the dorsal border

xian Formation of Liaoning Province, China (You et al. 2003;

running rostroventrally from just below the most distal tooth.

You and Xu 2005). The holotype (IVPP V12704) is a juvenile

The rostral end reaches the rostroventral corner of the den-

skull and IVPP V12617 is the skull of an adult. In this study,

tary. The ventral border descends slightly.

IVPP V12617 is described.

The prearticular is partially preserved in both mandibles.

The mandible is mostly preserved (Fig. 16.2A, B). The man-

The rostral parts are missing due to the lack of the rostral

dible is deeper than that of Psittacosaurus. The external man-

border of the mandibular fossa. The caudal region is nearly

dibular fenestra is situated at the dentary-surangular-angular

straight in medial view, curving slightly medially at its caudal

junction.

end. It expands dorsally to cover the rostral surface of the

The rostral tip of the predentary of Hongshanosaurus is missing. If complete, the rostral would have been semicircular in dorsal view (Fig. 16.2A). In ventral view, it widens slightly caudally to the point of bifurcation, from which it tapers

articular. The caudal tip of the right prearticular reaches the caudal end of the mandible. Both tooth rows bear 10 teeth (Fig. 16.2A). They are slightly concave labially.

caudally on both sides. In lateral view, the nearly straight ventral border descends caudoventrally at an angle of about 50\

PSITTACOSAURUS

from the roughly horizontal dorsal border (Fig. 16.2B). The caudolateral process is approximately 5 mm long, stretching

Psittacosaurus is one of the most abundant of all dinosaurs.

caudally from the caudodorsal corner of the predentary. The

All specimens come from the Lower Cretaceous of Asia. Most

significantly longer ventral process develops caudoventrally

specimens are discovered from eastern Asia, but some occur-

and bifurcates shortly at its end to articulate with the caudo-

rences are reported from western Siberia, Japan, Thailand, and

laterally diverging dentaries.

Xinjiang Uyghur Autonomous Region of China, indicating

The dentary comprises nearly half of the Hongshanosaurus

the wide distribution of the genus (Osborn 1923; Young 1931,

mandible in lateral view (Fig. 16.2B; You and Xu 2005). It is

1958; Sereno and Chao 1988; Sereno et al. 1988; Buffetaut et

extremely deep relative to the length. Slightly rostral to the

al. 1989; Manabe and Hasegawa 1991; Russell and Zhao 1996;

midpoint of the dentary stretches the rostral end of the well-

Averianov et al. 2006). Recent discoveries of new species such

developed ventrolateral flange, which continues to the mid-

as P. lujiatunensis and P. major (Zhou et al. 2006; Sereno et

dle of the angular. In dorsal view, three neurovascular for-

al. 2007) suggest that our knowledge of this important animal

amina are aligned along the lateral margin (Fig. 16.2A).

is still growing. In this study, mandibles of P. lujiatunensis

The surangular forms the caudodorsal portion of the man-

(PKUP 1053, PKUP 1054, PKUP 1060), P. major (CAGS-IG-

dible including the caudal part of the coronoid process (Fig.

VD-004), P. mongoliensis (AMNH 6254), P. neimongoliensis

16.2B; You and Xu 2005). The dorsalmost point of the sur-

(IVPP V12-0888-2), and P. sinensis (IVPP V738) are observed.

angular is slightly lower than the apex of the coronoid process

The mandible of Psittacosaurus is slender and the mandibu-

on the right mandible and about the same level in the left

lar ramus is nearly straight in dorsal view, except for the rela-

mandible. The dorsal margin is strongly convex. At the caudal

tively wide and caudally expanding mandibular ramus of P.

end, a well-developed lateral protuberance is present lateral to

sinensis (Fig. 16.2C). Caudal to the predentary, the width of

and parallel with the rostral half of the articular, forming the

the mandible shows very little change except at the jaw joint

lateral part of the articular surface (Fig. 16.2A).

as in Chaoyangsaurus. The external mandibular fenestra is

Although the dentary-angular suture is obliterated, the an-

closed in P. sinensis, P. neimongoliensis, P. lujiatunensis, P. sibi-

gular appears to be a little larger than the surangular in lateral

ricus, and P. major (Young 1958; Russell and Zhao 1996; Averia-

view based on the position of surangular-angular suture (Fig.

nov et al. 2006; Zhou et al. 2006; Sereno et al. 2007). The

16.2B; You and Xu 2005). The ventral border ascends caudo-

mandible of Psittacosaurus is deep compared to that of basal

dorsally. The rostrocaudal part contributes to the ventrolat-

neoceratopsians partially due to a distinctive flange extending

eral flange.

ventrolaterally from the ventral part of the mandible (Fig.

The rostral half of the dorsal surface of the articular is rela-

16.2D). The flange is present in P. mongoliensis, P. meileyingen-

tively flat. Caudal to the articular surface is a depressed shelf

sis, P. ordosensis, P. lujiatunensis, P. sibiricus, and P. major (Sereno

for the mandibular depressor muscle on the dorsal surface of

et al. 1988; Sereno 1990; Russell and Zhao 1996; Averianov et

the retroarticular process (Fig. 16.2A).

al. 2006; Zhou et al. 2006; Sereno et al. 2007). Among these

Only the dorsal part of the right coronoid is preserved. As in other basal ceratopsians, it covers the medial side of the coronoid process and it expands caudodorsally.

238 tanoue, you, & dodson

species, the flange is most strongly developed in P. major, a very large species. The predentary caps the dentaries rostrally. It is triangular

FIGURE 16.2. Mandibles of psittacosaurids. Hongshanosaurus houi (IVPP V12617) in (A) dorsal and (B) left lateral views; Psittacosaurus major (CAGS-IG-VD-004) in (C) dorsal, (D) rostral, (E, F) left lateral, and (G, H) left medial views. (A–E, G) photographs; (F, H) interpretive outlines. Scale bar is 5 cm.

in lateral view and semicircular in rostral view (Fig. 16.2D–F).

meet at the symphysis (Fig. 16.2C). The caudodorsal portion

In lateral view, it is relatively short and deep in P. lujiatunensis,

of the dentary rises to form the rostral half of the coronoid

P. major, P. meileyingensis, P. neimongoliensis, and P. sinensis, and

process including its apex (Fig. 16.2E, F). The apex has stria-

relatively long and shallow in P. mongoliensis and P. sibiricus

tions running caudodorsally for M. pseudotemporalis (MPsT)

(Young 1958; Sereno 1987; Sereno et al. 1988; Russell and

attachment. From the ventral margin stretches the ventro-

Zhao 1996; Averianov et al. 2006; Zhou et al. 2006; Sereno et

lateral flange, which is present to some degree in most psit-

al. 2007). In dorsal view, the cutting edge is horseshoe-shaped

tacosaurids. P. sinensis and P. neimongoliensis, however, do not

as in the opposing rostral (Fig. 16.2C). The lateral process ar-

possess this flange (Sereno et al. 1988; Russell and Zhao 1996).

ticulates with the rostrodorsal corner of the right dentary. The

A prominence extending across the lateral surface of the den-

ventral process extends caudoventrally in lateral view taper-

tary from the coronoid process to the rostral corner of the

ing distally (Fig. 16.2E, F). In dorsal view, the caudal margin of

flange is noted in P. meileyingensis, P. mongoliensis, and P. lu-

the surface of the predentary of P. major (CAGS-IG-VD-004) is

jiatunensis (Sereno 1987; Sereno et al. 1988; Zhou et al. 2006).

composed of two concave arches that receive the convex ends of dentaries (Fig. 16.2C).

The surangular is a large bone forming much of the caudal third of the mandible (Fig. 16.2E, F). It is bounded by the

The dentary forms nearly half of the mandible in lateral

dentary rostrally, by the angular ventrally, and by the articular

view (Fig. 16.2E, F). Its rostral end is covered by the caudal mar-

caudally. In lateral view, the surangular is subtriangular, taper-

gin of the predentary, with the predentary-dentary suture run-

ing caudoventrally. The rostrodorsal corner of the surangular

ning caudoventrally. Medially, the rostral ends of the dentaries

contributes to the caudal half of the coronoid process. In dor-

Mandibular Anatomy in Basal Ceratopsia 239

sal view, it is mostly slender (Fig. 16.2C). However, at its caudal

Azuma 1997). The holotype (IVPP V11114) includes a nearly

end it bears a lateral process. Together with the medially proj-

complete skull, the only one presently known for this species.

ecting articular, the mandible reaches its greatest width at the

The sutures between the cranial elements are mostly unclear

jaw joint. The projection is lateral to and co-planar with the

since the surface of the holotype is not well preserved.

glenoid surface of the articular, contributing to the lateral part of the articular surface.

The mandible is mostly preserved except for the prearticular (Fig. 16.3A). The left mandible is slightly compressed in

The angular forms most of the ventral half of the caudal

a rostrocaudal direction. The presence of an external man-

third of the mandible of Psittacosaurus (Fig. 16.2E, F). It con-

dibular fenestra is uncertain due to preservation (Fig. 16.3B).

tacts the dentary rostrally, the surangular dorsally, and the

Description is based primarily on the right mandible of Ar-

articular caudally. In lateral view, it is a subtriangular element

chaeoceratops.

tapering caudally. In medial view it is overlapped by the prear-

The predentary of Archaeoceratops is clearly seen (Fig. 16.3A,

ticular and the splenial (Fig. 16.2G, H). The rostral end forms

B). The rostral tip is pointed slightly dorsally. In lateral view,

the caudal part of the ventrolateral flange of the mandible.

both the dorsal and ventral margins are slightly convex ven-

The splenial covers the midsection of the medial side of the

trally (Fig. 16.3B). In dorsal view, the lateral margins diverge

mandible (Fig. 16.2G, H). It is bounded by the dentary ros-

at approximately 40\ (Fig. 16.3A). Ridges are located on the

trodorsally, the coronoid dorsally, the prearticular caudally,

dorsal surface medial to the lateral margins. The ridges form

and the angular caudoventrally. The rostral end of the splenial

the dorsal margin of predentary in lateral view. These ridges

reaches the caudoventral end of the mandibular symphysis.

diverge caudally, but the space between the ridges and the

From the rostral end, the dorsal margin of the splenial stretches

lateral margin widens caudally. The caudolateral process is

caudodorsally to the coronoid. The ventral margin is ventrally

short and measures 8 mm on the left side. The ventral process

convex. The caudodorsal margin is convex rostroventrally.

of the Archaeoceratops predentary is long as in other basal cera-

The coronoid is the smallest element of the mandible, and

topsians (Fig. 16.3B). It tapers slightly and bifurcates at the

contributes to the medial side of the coronoid process (Fig.

caudal end. In lateral view, grooves and foramina are dis-

16.2G, H). It contacts the dentary rostrally and dorsally, the

tributed on the dorsal region of the predentary. The inter-

splenial ventrally, the prearticular caudoventrally, and the sur-

dentary process, which is 10 mm long, tapers caudoventrally

angular caudally. In medial view it is L-shaped, with its dorsal

(Fig. 16.3A).

portion expanding caudodorsally. The rostroventral process extends to just below the caudal end of the tooth row.

Although the caudoventral end is missing, most of the dentary is preserved (Fig. 16.3A, B). Its width is nearly the same as

The Psittacosaurus prearticular consists of a thin strip on the

the height of the mandibular ramus. Although the ventral

medial side of the caudal portion of the mandible (Fig. 16.2G,

margin of the mandible is convex ventrally, that of the den-

H). It is bordered by the splenial rostrally, the coronoid ros-

tary is nearly straight (Fig. 16.3B). In dorsal view, the lat-

trodorsally, the articular caudodorsally, and the angular ven-

eral margin of the mandible is laterally concave just caudal to

trally. The rostral half of the prearticular is thinner than the

the predentary (Fig. 16.3A). The caudodorsal region forms the

caudal half and stretches rostrodorsally. The caudal half of

rostral half of the coronoid process including its apex (Fig.

the prearticular is horizontal, with a dorsal projection just in

16.3B). The caudal half of the dentary contributes to the lat-

front of the articular in medial view. A narrow prong stretches

eral ridge of the mandible.

caudally from the base of the dorsal projection and ends just before the caudal end of the mandible.

The ventral border of the right surangular is ambiguous (Fig. 16.3B). The dorsalmost point of the surangular is lower than

The articular forms the dorsal side of the caudalmost portion

that of the dentary. The caudodorsal margin of the surangular

of the mandible (Fig. 16.2G, H). The rostral portion of the

descends caudoventrally. The caudal half of the dorsally con-

articular is a highly unusual flat articular surface, which along

vex lateral ridge of the mandible is well developed on the lat-

with the co-planar lateral process of the surangular receives the

eral surface of surangular. A lateral protuberance extends from

quadrate (Fig. 16.2C, G, H; Sereno 1990; Zhou et al. 2006). The

the caudal end of the lateral ridge and lateral to the jaw joint

medial half of the articular surface projects medially. Unlike

(Fig. 16.3A). The surface of the caudal part of the lateral ridge

basalmost ceratopsians and basal neoceratopsians, a long and

including the protuberance is rough as opposed to the rela-

narrow retroarticular process extends caudally from the articu-

tively smooth surface of the rostral portion. Although ambig-

lar surface, reaching the caudal end of the mandible.

uous on the left mandible, the caudal end of the right surangular forms a short, 4 mm long retroarticular process.

ARCHAEOCERATOPS

The outline of the angular is unclear as with other elements. The ventrally convex ventral border ascends caudodorsally

Archaeoceratops oshimai comes from the Lower Cretaceous

(Fig. 16.3B). The caudal end of the left angular is missing, but

of the Mazongshan area, Gansu Province, China (Dong and

it may have reached the caudal end of the mandible.

240 tanoue, you, & dodson

FIGURE 16.3.

Mandibles of basal neoceratopsians. Archaoeceratops oshimai (IVPP V11114) in (A) dorsal and (B, reflected) right lateral views; Auroraceratops rugosus (IG-2004VD-001) in (C) dorsal, (D, E) right lateral, and (F, G) right medial views. (A–D, F) photographs; (E, G) interpretive outline. Scale bar is 5 cm.

The articular is rectangular in dorsal view, with its long axis running caudomedially (Fig. 16.3A). The long side of the right

rostral end of coronoid, about 5 mm rostral to the apex of coronoid process.

articular is 17 mm long and the short side is 12 mm long. In caudal view, the exposed height of the right articular is

AURORACERATOPS

12 mm. The dorsal surface forms the medial cotyle and part of the lateral cotyle of the jaw joint.

Auroraceratops rugosus is known from the Lower Cretaceous

The dorsal part of the left coronoid is preserved on the me-

Xinminpu Group in the Mazongshan area, Gansu Province,

dial side of the coronoid process. It expands caudodorsally to

China (You et al. 2005). The holotype (IG-2004-VD-001), the

articulate with the dentary and the surangular laterally.

only specimen discovered, is the largest Early Cretaceous basal

Although splenials are present in both Archaeoceratops man-

ceratopsian skull from China. Although the preservation of

dibles, the right splenial is better preserved. The ventral border

the teeth is not superb, the complete mandible is preserved

is slightly convex ventrally, with the caudal ramus slightly

(Fig. 16.3C).

higher than the rostral ramus. The caudal end of the splenial

The predentary is relatively long with a horizontal exten-

may have reached the caudal end of the mandible as with the

sion of the rostral end unlike that seen in other basal neo-

angular. The rostral end of the right splenial is located 4 mm

ceratopsians (Fig. 16.3D, E; You et al. 2005). The dorsal border

caudal to the end of the mandibular symphysis.

is roughly horizontal, tilting slightly rostroventrally. The ven-

The labially concave tooth rows are aligned along the me-

tral border, stretching caudoventrally from the rostral tip at

dial borders of the dentaries in dorsal view (Fig. 16.3A). Both

about 30\ from the dorsal border, is long and nearly straight,

tooth rows contain 14 teeth. Two teeth on the rostral end of

unlike the curved ones of other neoceratopsians. The ven-

each dentary are located apart from each other and the pri-

tral surface of the Auroraceratops predentary is bluntly keeled

mary tooth rows. The most distal tooth is located caudal to the

along the midline. In dorsal view, the lateral margins diverge

Mandibular Anatomy in Basal Ceratopsia 241

caudolaterally at an angle of approximately 65\ (Fig. 16.3C).

expands mediolaterally to cover the surangular ventrally. The

The left caudolateral process is approximately 5 mm long. In

caudal end of the angular is just a few millimeters rostral to the

dorsal view, an interdentary process 11 mm long and 5 mm

caudal end of the mandible.

wide stretches along the midline and separates the rostral

The articular is the most caudal element of the mandible

ends of dentaries. In lateral view, the ventral process is longer

along with surangular (Fig. 16.3C, F, G). Although the caudal

and deeper than the caudolateral process (Fig. 16.3D, E). The

end of the left articular is missing, the right articular is com-

ventral process bifurcates at its caudal end to articulate with

plete. The rostral surface of the articular is covered by the sur-

the caudally diverging dentaries (You et al. 2005).

angular and prearticular, the lateral surface by the surangular,

The dentary of Auroraceratops composes nearly half of the

and the ventrolateral surface by the angular. In dorsal view,

mandible in lateral view (Fig. 16.3D, E). In dorsal view, both

the right articular is roughly trapezoidal (Fig. 16.3C). It mea-

medial and lateral margins are laterally concave as in Lepto-

sures 26 mm along the long axis, is 22 mm wide, and tapers

ceratops gracilis (Fig. 16.3C; Sternberg 1951). The curvature is

caudally. The caudal surface is triangular and measures 12 mm

intensified by the presence of a prominent protuberance at

in height. The rostral half of the dorsal surface contributes to

the caudal end of the dentary and rostrodorsal to the external

the small medial cotyle and the medial side of the larger lateral

mandibular fenestra (You et al. 2005). Four neurovascular for-

cotyle. Caudal to the cotyles, the 10 mm long retroarticular

amina are present on the dorsal surface along the lateral mar-

process stretches caudally, with the depression for the man-

gin, the diameters of which become larger caudally. In this

dibular depressor muscle attachment on the dorsal surface.

specimen, the rostral ends of the dentaries are not in contact

The Auroraceratops splenial covers the medial side of the

with each other, due to the interdentary process (see above) or

mandible (Fig. 16.3F, G). The rostral ends of both splenials are

possibly due to preservation. The dorsal margin rostral to the

missing. The axes of rostral and caudal rami are nearly hori-

coronoid process is convex, with a steeper rostral slope and a

zontal. Most of the rostrodorsal margin and the caudodorsal

shallower caudal slope (Fig. 16.3D, E). In medial view, the ros-

margins are concave caudoventrally and rostroventrally, re-

tral half of the medial surface is concave medially (Fig. 16.3F).

spectively, with the latter more strongly curved. The rostral

The caudodorsal portion of the dentary forms the rostral part

part of the rostral ramus tapers rostroventrally. The dorsal

of the coronoid process, located dorsal and slightly rostral to

ramus is almost vertical, tilted slightly caudally. The dorsal

the external mandibular fenestra (Fig. 16.3D, E). The coronoid

end of the ramus reaches the ventral end of the coronoid. The

process extends vertically in rostral view.

caudodorsal border of the splenial is not in contact with the

The surangular occupies the caudodorsal part of the mandi-

mandibular fossa, due to the presence of the prearticular in

ble and contributes to the caudal half of the coronoid process

this area. The caudal end of the splenial is below and slightly

in lateral view (Fig. 16.3D, E). The dorsalmost point of the

rostral to the caudal border of the mandibular fossa.

surangular is lower than the apex of the coronoid process. The

The coronoid contributes to the medial side of the coronoid

dorsal margin of the surangular is slightly convex dorsally

process (Fig. 16.3F, G). The dorsal part expands caudodorsally.

and slopes caudoventrally. Medially, the surangular is a thin

The coronoid tapers ventrally and extends a rostroventral

plate that covers the rostral surface of the articular and forms

ramus below the most distal tooth. Although this ramus is not

the caudal border of the mandibular fossa (Fig. 16.3F). The

fully preserved in the left coronoid, that of the right coronoid

dorsally convex lateral ridge runs across the surangular and

is complete and is horizontal.

reaches the caudal part of dentary, with a boss at each end

The intercoronoid of Auroraceratops can be observed in cau-

(Fig. 16.3D). The lateral ridge is much lower than the bosses

dal view or in ventral view through the external mandibular fe-

(Fig. 16.3C). The boss lateral to the glenoid fossa is very promi-

nestra. It is a small rectangular element between the coronoid

nent. Medial to this boss, the surangular contributes to the

and the surangular. The left intercoronoid measures 9 mm

lateral portion of the lateral cotyle. A thin dorsal process at the

along its long axis and 6 mm in width. The right intercoronoid

caudal end of the dorsal margin separates the lateral cotyle

is 13 mm in length and 5 mm wide. The intercoronoid runs

and the boss, preventing the quadrate from moving laterally.

rostroventrally and slightly medially along the caudoventral

Caudal to the boss, the surangular tapers at its caudal end to

margin of the coronoid. In caudal view, the rostral half of the

form the lateral half of the short retroarticular process.

left intercoronoid is twisted about 90\ clockwise from the mid-

The angular of Auroraceratops is much smaller than the sur-

dle and the right intercoronoid counterclockwise.

angular, less than half the size of the latter in lateral view (Fig.

The prearticular contacts the splenial rostrally and ven-

16.3D, E). Its lateral surface is relatively smooth compared to

trally, and the articular caudally (Fig. 16.3F, G). The rostral half

that of dentary and surangular, due to the absence of pro-

stretches rostrodorsally along the caudodorsal margin of the

tuberances (You et al. 2005). The rostral margin interdigitates

splenial, forming the rostral and ventral border of the man-

with the caudoventral corner of the dentary. The ventral mar-

dibular fossa. Its caudal portion bifurcates into a short dorsal

gin gently ascends caudodorsally. In ventral view, the angular

branch and a long ventral branch. The dorsal branch covers

242 tanoue, you, & dodson

the rostromedial surface of the articular and forms the caudo-

contribution of the dentary to the coronoid process is rela-

ventral border of the mandibular fossa. The ventral branch

tively larger than that of most basal ceratopsians observed.

tapers caudally and ends just before the caudal end of the

The caudal margin interdigitates with the rostral margins of

mandible.

the surangular and angular. A low lateral expansion is present,

The dentary tooth rows are aligned along the medial borders of the dentaries in dorsal view, and are concave laterally

ventral and slightly rostral to the coronoid process. The ventral margin is ventrally convex and expands laterally.

(Fig. 16.3C). Twelve teeth are included in both closely packed

The surangular forms the caudodorsal portion of the man-

tooth rows. Two sockets are present on the dorsal surface of

dible in lateral view, and is about two-thirds the size of the

both dentaries a few millimeters apart from each other and

angular (Fig. 16.4D, E). Its contribution to the coronoid pro-

rostral to the tooth rows. Since the second socket of the left

cess is much smaller than that of the dentary. The dorsal mar-

dentary contains a root-like filling, they may be alveoli for

gin of the surangular runs caudoventrally with a gentle rostro-

unpacked teeth as in Archaeoceratops.

ventrally concave curvature. The ventral margin interdigitates with the dentary and angular, with the rostrocaudal process

LEPTOCERATOPS

received by the dentary and the dorsal process of the angular received by the surangular. Medial to the caudal end of the

Leptoceratops gracilis is one of the few North American basal

dorsal margin, which serves as the lateral wall of the jaw joint,

neoceratopsians. It is, incidentally, the first described basal

the surangular forms the lateral half of the lateral cotyle of the

ceratopsian. Specimens of Leptoceratops come from the upper

jaw joint.

Maastrichtian of the Scollard Formation in Alberta, the Lance

The angular of Leptoceratops contacts the dentary rostrally,

Formation in Wyoming, and the Hell Creek Formation in

the surangular dorsally, the splenial rostromedially, and the

Montana (Brown 1914; Sternberg 1951; Ostrom 1978; Ryan

articular caudomedially (Fig. 16.4D–G). Its dorsal and rostral

and Currie 1998; Ott 2007). Leptoceratops is also among the

margins interdigitate with the surangular and the dentary, re-

youngest basal neoceratopsians. Three complete and several

spectively. The ventral margin is undulating and ascends cau-

partial skulls have been collected along with postcranial skele-

dodorsally (Fig. 16.4D, E).

tons. In this study, the well-preserved skull of NMC 8889 is described (Fig. 16.4A).

The rostral and lateral surfaces of the articular are covered by the surangular and the angular (Fig. 16.4F, G). The rostral

The predentary is disarticulated from the dentaries in this

half of the dorsal surface contributes to the medial cotyle and

specimen (Fig. 16.4B). The ventral margin is composed of

the medial half of the lateral cotyle (Fig. 16.4A). The caudal

two nearly straight ridges, forming an angle of approximately

half of the articular tapers caudally in dorsal view. The con-

155\. The dorsal border is straight caudally, but gently curves

cave surface for the mandibular adductor muscle is located

dorsally toward the rostral tip. Although not fully preserved,

just caudal to the cotyles. A short retroarticular process is

the dorsal ramus is shorter than the ventral ramus. The ventral

formed by the articular medially and the surangular laterally.

ramus is very long, slightly more than half of the length of the

In caudal view, both the lateral and medial sides of the ventral

predentary. In dorsal view, the ventral ramus tapers caudally

surface slope toward the midline.

and bifurcates to articulate with the dentaries diverging cau-

The Leptoceratops coronoid forms the medial side of the cor-

dolaterally. Medial to the dorsal ramus, an unusual process

onoid process (Fig. 16.4F, G). It expands dorsomedially. The

approximately 15 mm long stretches caudally and is received

ventral portion is narrow, stretching and tapering medially to

by the socket on the rostrodorsal surface of the dentary (Fig.

the caudal end of the tooth row.

16.4C; Sternberg 1951). The dorsal third of the ventral surface

The intercoronoid is best observed in caudomedial view. It

is ornamented with neurovascular foramina and grooves (Fig.

is sandwiched between the dentary and the coronoid. It is a

16.4B). Unlike the rough surface of the dorsal portion, the

thin element stretching rostroventrally and slightly medially

ventral two-thirds of the surface are relatively smooth.

(Sternberg 1951).

The Leptoceratops dentary covers a large portion of the lat-

The splenial covers nearly one-third of the medial surface of

eral surface of the mandible (Fig. 16.4D, E). The rostrodorsal

the mandible (Fig. 16.4F, G). The rostral ramus reaches the

corner receives the caudal process of the predentary. When

caudoventral end of the mandibular symphysis. From the ros-

articulated with the ventral process of the predentary, the

tral end, the rostrodorsal margin ascends caudodorsally to the

nearly straight rostroventral margin of the dentary reaches

base of the 12th dentary tooth and then stretches horizontally

approximately 80 mm. Along and caudal to the ventral pro-

to the base of the most distal tooth. The caudodorsal margin

cess of the predentary runs the mandibular symphysis, which

forms the rostral and ventral borders of the glenoid fossa. The

is 70 mm along the midline. The caudodorsal portion of the

caudal ramus reaches the rostral end of the articular. The ven-

dentary forms the coronoid process. The coronoid process of

tral margin of the splenial is ventrally convex, congruent with

Leptoceratops is strongly tilted medially (Sternberg 1951). The

that of the dentary and the angular.

Mandibular Anatomy in Basal Ceratopsia 243

FIGURE 16.4.

Mandibles of basal neoceratopsians. Leptoceratops gracilis (NMC 8889) in (A) dorsal view; (B) predentary in left lateral view; (C) caudal process of predentary (white arrow) and socket of left dentary (black arrow) in dorsal view; (D, E) mandible in right lateral view; (F, G) mandible in right medial view. Liaoceratops yanzigouensis (IVPP V12738) in (H) dorsal and (I, J) right lateral views; Protoceratops andrewsi (AMNH 6460) in (K) dorsal and (L) right lateral views. (A–D, F, H, I, K, L) Photographs; (E, G, J) interpretive outlines. Scale bars are 5 cm.

244 tanoue, you, & dodson

Most of the Leptoceratops prearticular can be observed in

at least five foramina are aligned along the lateral margin: the

dorsal view. It is a thin element only several millimeters in

rostral two are small, the third and fourth of medium size, and

width except at the caudal end where it widens. It runs along

the caudalmost one is large (Fig. 16.4H).

the medial side of the caudal ramus of the splenial, forming

The Liaoceratops surangular is about two-thirds the size of

the ventral border of the mandibular fossa. At the caudal end,

the angular in lateral view (Fig. 16.4I, J). The dorsal border is

the prearticular is twisted and expands dorsally (Fig. 16.4F, G).

not as convex as in other basal ceratopsians except for that of

The dentary tooth rows are greatly inset from the lateral

Chaoyangsaurus. Although the surangular contributes to the

margin, as much as 50 mm from the lateral margin of dentary

caudal portion of coronoid process, the dorsalmost point is

(Fig. 16.4A). Both dentary tooth rows are composed of 16

much lower than the apex, by as much as 6 mm. A dorsally con-

teeth. In dorsal view, they are slightly concave labially.

vex lateral ridge runs across the lateral surface of the surangular. Unlike in Archaeoceratops, this ridge is very low and a lateral

LIAOCERATOPS

process does not develop at the caudal end of the ridge. The caudal part of the surangular forms the lateral side of the lateral

Liaoceratops yanzigouensis is known from the Lower Creta-

cotyle, medial to the caudal end of the dorsal margin. The

ceous Yixian Formation of western Liaoning Province, China

caudal end of the surangular also forms the extremely short

(Xu et al. 2002). The holotype (IVPP V12738) is an almost

retroarticular process, located caudal to the glenoid fossa.

complete adult skull. Two juvenile skulls have also been col-

The angular forms the caudoventral part of the mandible

lected (IVPP V12633, CAGS-IG-VD-002; Xu et al. 2002; You et

(Fig. 16.4I, J). The rostral border receives the caudal prong of

al. 2007). The following description is mostly based on IVPP

the dentary. The ventrally convex ventral border gently as-

V12738, the most mature of the three skulls.

cends caudodorsally. The ventral margin expands laterally

Although the right mandible appears to maintain its original morphology, the left mandible of Liaoceratops is shortened

and forms a low ridge, which continues rostrally to the ventral margin of the dentary.

rostrocaudally (Fig. 16.4H). The caudal end of the left mandi-

The right articular is mostly complete (Fig. 16.4H), but the

ble is missing. The mandibular ramus is low as in other basal

left articular is only partially preserved. In dorsal view, the

neoceratopsians (Fig. 16.4I, J). In the caudal half of the lateral

right articular is roughly rectangular, with long sides run-

side, a shallow depression is present, surrounded by a lateral

ning caudolaterally. The right medial cotyle is slightly deeper

ridge and another ridge along the ventral margin. The surface

than the lateral cotyle. The caudal end of the right articular is

is smooth without any distinctive rugosity. An opening is ob-

missing, but the caudodorsal surface exhibits a small depres-

servable at the junction of the dentary, the surangular, and the

sion, indicating the attachment of the mandibular depressor

angular in the right mandible, but not in the left mandible.

muscle.

Most of the predentary other than the rostral tip is pre-

The coronoid contributes to the apex of the coronoid pro-

served (Fig. 16.4H–J). The rostral tip of the predentary is pre-

cess. The dorsal portion widens caudodorsally to form the me-

served in IVPP V12633 and CAGS-IG-VD-002, both of which

dial side of the coronoid process. Ventrally, the coronoid bi-

are beveled (Xu et al. 2002; You et al. 2007). In dorsal view, the

furcates into rostral and caudal prongs. The rostral prong

dorsal margin is rostrolaterally convex in the rostral half and

descends rostroventrally along the rostrodorsal margin of the

straight in the caudal half (Fig. 16.4H). In lateral view, the

splenial and ends below the second from the last tooth. The

dorsal margin ascends rostrodorsally (Fig. 16.4I, J). The ven-

caudal prong stretches ventrally to meet the rostrodorsal tip of

tral border of the predentary is strongly curved. The caudolat-

the prearticular.

eral processes tapering caudally and slightly dorsally are 9 mm

The splenials are almost complete in both mandibles of

on both sides. The ventral process is relatively wide. The dor-

IVPP V12738. The tip of the dorsal ramus reaches the level of

sal margin of the ventral process is unique among those of

the dorsal border of the mandibular fossa. The rostrodorsal

basal neoceratopsians in that it is dorsally convex. A short

margin descends rostroventrally. Although the rostral tips are

median interdentary process stretches caudoventrally and is

missing on both splenials, they may have reached the caudo-

4 mm long as preserved.

ventral end of the mandibular symphysis. The caudal ramus

The dentary covers about half of the lateral side of mandible (Fig. 16.4I, J). The rostral two-thirds compose the mandibular

stretches caudally between the prearticular and the angular, and ends a few millimeters rostral to the articular.

ramus. The caudodorsal part forms the rostral portion of the

The prearticulars are partially preserved in both mandibles.

coronoid process, including its apex. The low lateral ridge be-

The rostral portion is vertical, contributing to the rostral

gins in the caudal portion. The caudal half of the ventral mar-

border of the mandibular fossa. From the rostroventral cor-

gin expands laterally, which is noted as a ventral flange in Xu

ner, the prearticular runs caudally, covering the lateral and

et al. (2002). Numerous neurovascular foramina are distrib-

dorsal sides of the ventral border of the mandibular fossa.

uted along the rostral half of the lateral surface. In dorsal view,

It expands dorsally just rostral to the articular. Caudally a

Mandibular Anatomy in Basal Ceratopsia 245

thin prong extends, which almost reaches the caudal end of

of the surangular forms the lateral part of the lateral cotyle

the mandible.

and the lateral side of the retroarticular process.

In dorsal view, the labially concave tooth rows are aligned

The angular forms the caudoventral part of the mandible. In

along the medial borders of the dentaries (Fig. 16.4H). The left

AMNH 6429, it is longer than the surangular in lateral view.

tooth row contains 13 teeth, and the right tooth row contains

The ventral margin is strongly convex caudoventrally and as-

15 teeth. The most distal tooth is only a few millimeters rostral

cends caudodorsally up to the caudal end, which is horizon-

to the apex of the coronoid process.

tal. The angular of AMNH 6429 expands medially in ventral view. It does not reach the caudal end of the mandible.

PROTOCERATOPS

The left articular is well preserved in AMNH 6429. The Protoceratops articular is a massive element at the caudal end of the

Protoceratops is one of the largest basal ceratopsians. Abundant

mandible, similar to that of Leptoceratops. It is roughly tra-

specimens in various growth stages are found in Upper Creta-

pezoidal in dorsal and medial views. The rostral half of the

ceous sediments of Mongolia and China, allowing studies of

dorsal surface contributes to the medial cotyle and the medial

ontogeny and sexual dimorphism (Kurzanov 1972; Dodson

part of the lateral cotyle. A short retroarticular process extends

1976). Two species, P. andrewsi and P. hellenikorhinus, have

caudal to the cotyles, with a shallow depression for the man-

been described (Granger and Gregory 1923; Lambert et al.

dibular depressor muscle attachment.

2001). P. andrewsi is observed in this study.

The coronoid is clearly visible in AMNH 6460. The dorsal

The following description is mainly based on the relatively

part expands dorsally to form the medial side of the coronoid

well-preserved mandibles of AMNH 6429 and AMNH 6460,

process. The ventral part is narrow and curves ventromedially,

both of which are disarticulated from the skulls. AMNH 6429

forming the rostrodorsal border of the mandibular fossa. At the

includes a partial left mandible, complete except for the pre-

medial end, the coronoid bifurcates for a short distance and

dentary and the prearticular. The mandible of AMNH 6460 is

receives the dorsal tip of the splenial between the two prongs.

nearly complete (Fig. 16.4K, L). The mandibles of Protoceratops

The splenial is mostly preserved in AMNH 6460. The rostral

have ventrally convex ventral margins. The external mandib-

end reaches the caudoventral edge of the mandibular sym-

ular fenestra is absent.

physis (Brown and Schlaikjer 1940). The rostrodorsal margin

The predentary of AMNH 6460 tapers rostrodorsally in rostral and lateral views to form a sharp rostral tip, which is

stretches caudodorsally to the midpoint of the tooth row. The dorsal margin runs horizontally.

keeled on the midline in rostral view (Fig. 16.4L; Brown and

The prearticular is best observed in the right mandible of

Schlaikjer 1940). The ventral margin is strongly convex ven-

AMNH 6460. The Protoceratops prearticular is narrow and

trally in lateral view. Although the predentary-dentary suture

nearly horizontal except for at the rostral end, which stretches

is mostly obliterated in this specimen, predentaries of AMNH

rostrodorsally. It is attached medially to the splenial along the

6418, AMNH 6441, and AMNH 6467 have short caudolateral

dorsal margin of the caudal ramus, forming the ventral border

processes and long ventral processes of roughly the same

of the mandibular fossa.

height. The ventral process of AMNH 6460 diverges at the caudal end and extends about 10 mm on each side.

The dentary tooth rows of AMNH 6460 are inset from the lateral margin of the dentaries by as much as 30 mm (Fig.

The dentary of AMNH 6429 covers a large portion of the

16.4K). The left tooth row contains 13 teeth and the right

lateral surface of the mandible. The ventral margin is slightly

tooth row between 12 and 14 teeth. The dentary tooth rows

convex ventrally, but this curvature is not as strong as in

are slightly concave lingually in dorsal view.

the ventral margins of the angular and the predentary. The caudodorsal portion forms the rostral part of the coronoid process including the apex. In rostral view, the coronoid

Discussion

process is vertical. The rostral half of the lateral ridge con-

Mandibular morphology is compared among the basal cera-

tinues rostroventrally from the surangular to the middle of

topsians and the ceratopsids. Table 16.2 lists a summary of the

the dentary.

characters discussed below.

The Protoceratops surangular contributes to the caudal part

Some basal ceratopsian mandibles observed in this study

of the coronoid process. In AMNH 6429, the dorsal margin

differ considerably from those of ceratopsids. One distinctive

forms a steep slope tilting caudoventrally. The caudal half

feature in some basal ceratopsian taxa is the convex ventral

of the strong lateral ridge, which is dorsally convex, devel-

border of the mandible. Among the observed specimens, Lep-

ops across the surangular (Brown and Schlaikjer 1940). Speci-

toceratops and Protoceratops have strongly curved ventral bor-

mens with articulated skulls and jaws such as AMNH 6418 and

ders (Fig. 16.4D, L; Brown and Schlaikjer 1940; Sternberg

AMNH 6425 indicate that the caudal half of the lateral ridge is

1951). This feature is also present in Udanoceratops (Kurzanov

congruent with the ventral border of the jugal. The caudal end

1992). In ceratopsids, the ventral border of the mandibular

246 tanoue, you, & dodson

Table 16.2. Summary of Mandibular Morphology Basal

ventral margin tooth rows mandibular width pcpl/ml articular surface

Basalmost Ceratopsia

Psittacosauridae

Neoceratopsia

Ceratopsidae

straight/ventrally

straight/ventrally

straight/ventrally

straight/dorsally

convex

convex

convex

convex

labially concave

straight/labially

labially concave

straight

(Chaoyangsaurus)

concave

narrow/wide

narrow

wide

wide

?.43

?.39

ⱕ.35

[ 0.35

depressed

flat

depressed

depressed

lateral wall of jaw joint

absent

absent

present

present

extension of jaw joint from

medial

lateral/medial

medial

medial

mandibular ramus Note: Outline of ceratopsid mandible modified after Hatcher et al. (1907). Abbreviations: ml: mandibular length; pcpl: postcoronoid process length (length from the apex of coronoid process to the caudal end of mandible).

ramus is either straight or dorsally convex, owing largely to

contribution of the surangular and angular to the lateral sur-

the morphology of the elongated dentary (Hatcher et al. 1907;

face of the mandible in the latter.

Dodson et al. 2004; Ryan 2007).

Basal neoceratopsians also exhibit a transverse thickening

Another interesting feature to discuss is the curvature of the

of the mandible, which is associated with the medial displace-

tooth rows. In dorsal view the tooth rows of ceratopsids are

ment of the tooth row. As a result, the M. adductor mandibu-

straight, whereas in some basal ceratopsians, especially basal

lae externus (MAME) entering the mandibular fossa of basal

neoceratopsians, the tooth rows are concave labially. This fea-

neoceratopsians must have been low and wide, whereas that

ture may have accommodated more teeth in the relatively

of other basal ceratopsians was deep and narrow. In addition,

short mandibular ramus of those basal ceratopsians compared

the tooth row of basal neoceratopsians is more convex and

to that of ceratopsids or may relate to an increased mechanical

relatively longer than that of other basal ceratopsians, reflect-

advantage (Tanoue in prep.).

ing the caudal extension of the tooth row in basal neoceratop-

Basalmost ceratopsians, psittacosaurids, and basal neocera-

sians (Tanoue in prep.).

topsians differ from each other in mandibular morphology,

The structure of the jaw joints differs among the three

which conforms to the cladistic analysis in Xu et al. (2006).

groups. The jaw joint of Chaoyangsaurus shows the lateral and

Chaoyangsaurus has a slender mandible with very little change

medial cotyles clearly (Fig. 16.1A). The two condyles of the

in width except at the glenoid, a condition similar to that of

quadrate are received by the cotyles. The jaw joint of Chao-

psittacosaurids (Fig. 16.1A). However, the predentary of Chao-

yangsaurus appears to have a simple hinge structure with no

yangsaurus is beveled as in neoceratopsians. The mandible of

additional buttressing structures. Cotyles are also present

Yinlong is relatively wide as in basal neoceratopsians, but the

on the articular surface of basal neoceratopsian jaws. More-

curvature of the lateral margin of the mandible just caudal

over, in the latter the caudal end of the dorsal margin of the

to the predentary is not as strong as in neoceratopsians (Fig.

surangular is located lateral to the glenoid fossa, providing a

16.1C). Table 16.3 shows the ratio of the length caudal to the

lateral buttress to the joint. Archaeoceratops and Leptoceratops

apex of the coronoid process to total mandibular length. Two

have an additional dorsal protuberance just lateral to the

basalmost ceratopsians and most psittacosaurids have ratios

jaw joint, which is higher than the caudal end of the dor-

of over 0.4. The exception is P. sinensis, whose right mandible

sal margin of the surangular (Fig. 16.3B, 16.4F). These fea-

has a ratio of 0.39. This ratio in basal neoceratopsians is no

tures prevent the quadrate from sliding laterally. Thus the jaw

higher than 0.35. Thus the basalmost ceratopsians and psit-

joints of basal neoceratopsians could extend medially to the

tacosaurids have a relatively longer portion of the mandible

mandibular ramus, but not laterally. The jaw joints of the

caudal to the coronoid process than seen in the basal neo-

basalmost ceratopsians also extend medially. In contrast, the

ceratopsians. This morphological difference reflects a smaller

mandibles of psittacosaurids show a jaw joint composed of an

Mandibular Anatomy in Basal Ceratopsia 247

Table 16.3. Ratio of the Length Caudal to the Apex of the Coronoid Process to Mandibular Length Species

Specimen no.

lf/r

ml

pcpl

Yinlong downsi

IVPP V14530

r

165

78

Chaoyangsaurus youngi

IGCAGS V371

r

120

52

0.43

Hongshanosaurus houi

IVPP V12617

lf

141

67

0.47

IVPP V12617

r

141

56

0.40

PKUP V1054

lf

159

78

0.49

PKUP V1054

r

157

73

0.46

AMNH6254

lf*

128

52

0.41

AMNH6254

r

125

52

0.41

IVPP V738

lf

105

43

0.41

IVPP V738

r

103

40

0.39

IG-2004-VD-004

lf

188

85

0.45

IG-2004-VD-004

r

193

86

0.45

IVPP V11114

lf

135

40

0.30

Psittacosaurus lujiatunensis P. mongoliensis P. sinensis P. major Archaeoceratops oshimai Auroraceratops rugosus

pcpl/ml 0.47

IVPP V11114

r

141

43

0.30

IG-2004-VD-001

lf

171

60

0.35

IG-2004-VD-001

r

170

53

0.31

NMC8889

lf

367

105

0.29

Leptoceratops gracilis

NMC8889

r

374

94

0.25

Liaoceratops yanzigouensis

IVPP V12738

lf

93

25

0.27

IVPP V12738

r

115

36

0.31

L. yanzigouensis

IVPP V12633

lf

71

22

0.31

IVPP V12633

r

80

26

0.33

CAGS-IG-VD-002

lf

65

20

0.30

CAGS-IG-VD-002

r

55

16

0.29

AMNH6460

r

290

84

0.29

L. yanzigouensis Protoceratops andrewsi

Abbreviations: lf: left mandible; ml: mandibular length in millimeters; pcpl: post-coronoid process length in millimeters (see Table 16.2); r: right mandible. * Measured from cast.

unusual flat dorsal surface of the articular and co-planar lat-

saurids indicates different feeding adaptations within basal

eral protuberance of the surangular. Cotyles are completely

Ceratopsia (Tanoue in prep.).

lacking. Together, they form an articular surface extending not only medial to, but also lateral to the mandibular ramus

Acknowledgments

(Fig. 16.2A, C). The quadrate is as wide as the width of whole

The authors are grateful to Carl Mehling (American Museum

articular surface in psittacosaurids, and the flat jaw joint

of Natural History), Kieran Shepherd (Canadian Museum of

has no bony structure to limit the horizontal movement of

Nature), Xing Xu, Qi Zhao (Institute of Vertebrate Paleon-

the jaw.

tology and Paleoanthropology), and Ke-Qin Gao (Peking University) for providing access to their collections. The au-

Conclusions

thors thank Brenda Chinnery-Allgeier and Xiao-chun Wu for reviewing the manuscript. Kyo Tanoue was funded by the

Recent discoveries of numerous basal ceratopsian specimens

Summer Research Stipends in Paleontology (University of

allow detailed anatomical study of their mandibles. Some gen-

Pennsylvania), the School of Arts and Sciences Dissertation

era have mandibles with curved ventral borders and tooth

Research Fellowship (University of Pennsylvania), and a Ju-

rows. These unique features in basal Ceratopsia, not seen

rassic Foundation Research Grant. Funding was provided

in Ceratopsidae, show diversity in the early evolution of the

by the Ministry of Science and Technology of China (973 Proj-

Ceratopsia. In addition, comparative study of basal ceratop-

ect: 2006CB701405), the National Natural Science Founda-

sian mandibles reveals different trends of morphological

tion of China (40672007), and the Hundred Talents Project

change in psittacosaurids and basal neoceratopsians. The rela-

of Ministry of Land and Resources of China to Hai-Lu You.

tively shorter distance between the coronoid process and the

Peter Dodson thanks his chairman, Narayan Avadhani, for

jaw joint of the basal neoceratopsian mandibles compared to

support.

the condition seen in basalmost ceratopsians and psittaco-

248 tanoue, you, & dodson

References Cited Alifanov, V. A. 2003. Two new dinosaurs of the infraorder Neoceratopsia (Ornithischia) from the Upper Cretaceous of the Nemegt Depression, Mongolian People’s Republic. Paleontological Journal 37: 524–534. Averianov, A. O., A. V. Voronkevich, S. V. Leshchinskiy, and A. V. Fayngertz. 2006. A ceratopsian dinosaur Psittacosaurus sibiricus from the Early Cretaceous of West Siberia, Russia and its phylogenetic relationships. Journal of Systematic Palaeontology 4: 359–395. Brown, B. 1914. Leptoceratops, a new genus of Ceratopsia from the Edmonton Cretaceous of Alberta. American Museum of Natural History Bulletin 33: 567–580. Brown, B., and E. M. Schlaikjer. 1940. The structure and relationships of Protoceratops. Annals of the New York Academy of Sciences 40: 133–266. Buffetaut, E., N. Sattayarak, and V. Suteethorn. 1989. A psittacosaurid dinosaur from the Cretaceous of Thailand and its implications for the palaeogeographical history of Asia. Terra Nova 1: 370–373. Chinnery, B. J. 2004. Description of Prenoceratops pieganensis gen. et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Chinnery, B. J., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Dodson, P. 1976. Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. Journal of Paleontology 50: 929–940. ———. 1996. The Horned Dinosaurs. Princeton: Princeton University Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Dong, Z., and Y. Azuma. 1997. On a primitive neoceratopsian from the Early Cretaceous of China. In Z. Dong, ed., SinoJapanese Silk Road Dinosaur Expedition, pp. 68–89. Beijing: China Ocean Press. Granger, W., and W. K. Gregory. 1923. Protoceratops andrewsi, a pre-ceratopsian dinosaur from Mongolia. American Museum Novitates 72: 1–9. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monographs 49: 1–300. Kurzanov, S. M. 1972. Sexual dimorphism in protoceratopsians. Paleontological Journal 1972: 91–97. ———. 1992. A gigantic protoceratopsid from the Upper Cretaceous of Mongolia. Paleontological Journal 26: 103–116. Lambert, O., P. Godefroit, H. Li, C. Shang, and Z. Dong. 2001. A new species of Protoceratops (Dinosauria, Neoceratopsia) from the Late Cretaceous of Inner Mongolia (P. R. China). Bulletin de l’ Institut Royal des Sciences Naturelles Belgique: Sciences de la Terre 71-Supp.: 5–28. Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia)

from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Manabe, M., and Y. Hasegawa. 1991. The Cretaceous dinosaur fauna of Japan. Fifth Symposium on Mesozoic Terrestrial Ecosystems and Biota. Extended Abstracts 364: 41–42. Osborn, H. F. 1923. Two Lower Cretaceous dinosaurs of Mongolia. American Museum Novitates 95: 1–10. Ostrom, J. H. 1964. A functional analysis of jaw mechanics in the dinosaur Triceratops. Postilla 88: 1–35. ———. 1978. Leptoceratops gracilis from the ‘‘Lance’’ Formation of Wyoming. Journal of Paleontology 52: 697–704. Ott, C. J. 2007. Cranial anatomy and biogeography of the first Leptoceratops gracilis (Dinosauria: Ornithischia) specimens from the Hell Creek Formation, Southeast Montana. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 213–233. Bloomington: Indiana University Press. Russell, D. A., and X. Zhao. 1996. New psittacosaur occurrences in Inner Mongolia. Canadian Journal of Earth Sciences 33: 637– 648. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, Southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., and P. J. Currie. 1998. First report of protoceratopsians (Neoceratopsia) from the Late Cretaceous Judith River Group, Alberta, Canada. Canadian Journal of Earth Sciences 35: 820–826. Sereno, P. C. 1987. The ornithischian dinosaur Psittacosaurus from the Lower Cretaceous of Asia and the relationships of the Ceratopsia. Ph.D. diss., Columbia University, New York. ———. 1990. Psittacosauridae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 579–592. Berkeley: University of California Press. Sereno, P. C., and S. Chao. 1988. Psittacosaurus xinjiangensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northwestern China. Journal of Vertebrate Paleontology 8: 353–365. Sereno, P. C., S. Chao, and C. Rao. 1988. Psittacosaurus meileyingensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northeastern China. Journal of Vertebrate Paleontology 8: 366–377. Sereno, P. C., X. Zhao, L. Brown, and L. Tan. 2007. New psittacosaurid highlights skull enlargement in horned dinosaurs. Acta Palaeontologica Polonica 52: 275–284. Sternberg, C. M. 1951. Complete skeleton of Leptoceratops gracilis Brown from the Upper Edmonton Member on Red Deer River, Alberta. Bulletin of the National Museum of Canada 123: 225– 255. Swisher, C., X. Wang, Z. Zhou, Y. Wang, F. Jin, J. Zhang, X. Xu, F. Zhang, and Y. Wang. 2002. Further support for a Cretaceous age for the feathered-dinosaur beds of Liaoning, China: New 40Ar/ 39Ar dating of the Yixian and Tuchengzi Formations. Chinese ScienceBulletin 47: 135–138. Xu, X., C. A. Forster, J. M. Clark, and J. Mo. 2006. A basal ceratopsian with transitional features from the Late Jurassic of northwestern China. Proceedings of the Royal Society of London B 273: 2135–2140.

Mandibular Anatomy in Basal Ceratopsia 249

Xu, X., P. J. Makovicky, X. Wang, M. A. Norell, and H. You. 2002. A ceratopsian dinosaur from China and the early evolution of Ceratopsia. Nature 416: 314–317. You, H., and Z. Dong. 2003. A new protoceratopsid (Dinosauria: Neoceratopsia) from the Late Cretaceous of Inner Mongolia. Acta Geologica Sinica 77: 299–304. You, H., D. Li, Q. Ji, M. C. Lamanna, and P. Dodson. 2005. On a new genus of basal neoceratopsian dinosaur from the Early Cretaceous of Gansu Province, China. Acta Geologica Sinica 79: 593–597. You, H., K. Tanoue, and P. Dodson. 2007. A new specimen of Liaoceratops yanzigouensis (Dinosauria: Neoceratopsia) from the Early Cretaceous of Liaoning Province, P. R. China. Acta Geologica Sinica 81: 898–904. You, H., and X. Xu. 2005. An adult specimen of Hongshanosaurus houi (Dinosauria: Psittacosauridae) from the Lower Cretaceous of Western Liaoning Province, China. Acta Geologica Sinica 79: 168–173.

250 tanoue, you, & dodson

You, H., X. Xu, and X. Wang. 2003. A new genus of Psittacosauridae (Dinosauria: Ornithopoda) and the origin and early evolution of marginocephalian dinosaurs. Acta Geologica Sinica 77: 15–20. Young, C. C. 1931. On some new dinosaurs from western Suiyan, Inner Mongolia. Bulletin of the Geological Survey of China 11: 259–266. ———. 1958. The dinosaurian remains of Laiyang, Shantung. Palaeontologia Sinica Series C 16: 1–138. Zhao, X., Z. Cheng, and X. Xu. 1999. The earliest ceratopsian from the Tuchengzi Formation of Liaoning, China. Journal of Vertebrate Paleontology 19: 681–691. Zhao, X., Z. Cheng, X. Xu, and P. J. Makovicky. 2006. A new ceratopsian from the Upper Jurassic Houcheng Formation of Hebei, China. Acta Geologica Sinica 80: 467–473. Zhou, C., K. Gao, R. C. Fox, and S. Chen. 2006. A new species of Psittacosaurus (Dinosauria: Ceratopsia) from Early Cretaceous Yixian Formation, Liaoning, China. Palaeoworld 15: 100–114.

17 Histological Evaluation of Ontogenetic Bone Surface Texture Changes in the Frill of Centrosaurus apertus A L L I S O N R . T U M A R K I N - D E R AT Z I A N

fragmentary frill elements (parietal and squamosal)

rectly from material examined here, related findings by

of Centrosaurus apertus were thin-sectioned for histologi-

other authors concerning cranial ontogeny and growth

cal analysis in order to determine the microstructural

dynamics of centrosaurines suggest a possible age of 2

basis of distinctive centrosaurine cranial bone surface

years for the long-grained to mottled transition, followed

textures described by previous authors. The study sample

by attainment of reproductive maturity and develop-

included examples of four textures—long-grained, mot-

ment of cranial ornamentation at approximately age 3.

tled, smooth, and rugose. Long-grained texture is associated histologically with bone depositional processes, including active incorporation of vascular canals into the

Introduction

bone matrix along periosteal surfaces. Mottled texture,

Reliable assignment of ontogenetic status, or differentiation

by contrast, is associated with resorption of surface bone.

of juvenile, subadult, and adult individuals, in the absence of

At both the gross and microstructural levels there is often

absolute age data and soft tissue anatomy is a critical issue in

a sharp demarcation between long-grained (deposi-

paleontological studies. The relative age terms juvenile, sub-

tional) and mottled (resorptive) regions that resembles

adult, and adult, as applied to fossil vertebrates, generally im-

an erosion front. Elements exhibiting smooth and rugose

ply successive steps toward attainment of somatic (specifically

textures are highly remodeled, with abundant secondary

skeletal) maturity. Assignment of an individual to relative age

osteons. Smooth surfaces are generally underlain directly

class based on skeletal maturity does not carry any implicit

by uninterrupted layers of lamellar bone tissue; osteonal

assumptions concerning the animal’s reproductive status.

remodeling is confined to innermost regions of the bone.

This is in sharp contrast to the use of the terms juvenile, sub-

In the case of rugose texture, the zone of active remodel-

adult, and adult in neontological studies, where definitions

ing more commonly extends out to the periosteal sur-

of ontogenetic stages are often based on gonadal develop-

face. The histological features observed in this study lend

ment (e.g., Hanson 1965, 1967; Chaloupka et al. 2004; Chal-

support to previous hypotheses associating changes from

oupka and Limpus 2005), mating behavior (e.g., Voelker 1997;

long-grained to mottled to smooth/rugose textures with

Araújo et al. 2000; Apio et al. 2007) and/or secondary sex

skeletal maturation of cranial elements during cen-

characteristics (e.g., Hanson 1965, 1967; Voelker 1997; Hoyer

trosaurine ontogeny. Although precise chronological

and Stewart 2000; Apio et al. 2007).

ages of surface bone changes cannot be determined di-

Establishing a clear distinction between ontogenetic stages

251

based on somatic versus reproductive criteria is critical, be-

a fundamental distinction between surface textures on the

cause somatic and reproductive maturity are not often coin-

bones of animals of different somatic age classes. Such tex-

cident. Sexual maturity precedes somatic maturity in extant

tural variations are macroscopic reflections of differences in

squamates and crocodylians (Chabreck and Joanen 1979; An-

the pattern of ossification of immature versus mature skeletal

drews 1982; Shine and Charnov 1992; Wilkinson and Rhodes

material. In this context, the term immature indicates bone

1997); the reverse is true in extant birds (Craighead and Stock-

that is actively growing longitudinally and/or appositionally;

stad 1964; Ricklefs 1968; Owen 1980; Tumarkin-Deratzian

the term mature indicates bone that has ceased to grow, or in

et al. 2006). In extant mammals, the pattern is variable:

which growth has slowed to a negligible rate.

medium-to large-sized taxa that take multiple years to reach

Documentation of predictable ontogenetic bone textural

adult size attain sexual maturity in advance of somatic matu-

changes in extant archosaurs (both modern and archaeologi-

rity, whereas smaller taxa follow an avian-like pattern with

cal specimens) has been largely confined to the postcranial

reproduction delayed until after growth ceases (Lee and Wern-

skeleton. Several authors have recognized diagnostic onto-

ing 2008). A few histological studies have suggested dinosaurs

genetic trends in birds (Callison and Quimby 1984; Benecke

followed the reptilian/large mammalian pattern, based on di-

1993; Cohen and Serjeantson 1996; Gotfredsen 1997; Sanz et

rect evidence of reproductive activity such as medullary bone

al. 1997; Serjeantson 1998, 2002; Mannermaa 2002; Tumarkin-

(Lee and Werning 2008) and brooding behavior (Erickson et

Deratzian et al. 2006); two studies ( Johnson 1977; Tumarkin-

al. 2007), or indirect evidence such as a marked slowing of

Deratzian et al. 2007) have addressed the possibility of predict-

growth at smaller than adult body size (Rimblot-Baly et al.

able bone textural changes in crocodylians. Johnson (1977)

1995; Curry 1999; Sander 2000). Such data are as yet available

postulated, but did not test for, a pitted texture on the bone of

for a limited number of taxa, and therefore most consider-

young individuals, and a smooth surface in older animals with

ations of relative age classes and ontogenetic status in fos-

reduced growth rates. Tumarkin-Deratzian et al. (2007) found

sil vertebrates relate only to somatic maturity, even in cases

no evidence of predictable textural change in growth series of

where this is not explicitly stated in the literature.

Alligator mississippiensis femora, tibiae, and humeri, and hy-

Body or skeletal element size should not be treated as the

pothesized that sensitivity to local and regional environmen-

sole criterion for recognizing somatic age classes. Within a

tal conditions, coupled with long periods of growth before

given taxon, it can be useful to refer to juvenile-, subadult-,

somatic maturity, rendered growth too individually variable

and adult-sized individuals, following the general rule that,

for successful application of textural aging.

on average, subadults will be larger than juveniles, and fully

Surface textural differences that permit categorization as im-

grown adults will be larger than subadults (Ryan et al. 2001).

mature and mature bone have been noted for several groups of

The general pattern, however, may be confounded by intra-

fossil amniotes, including basal synapsids (Brinkman 1988),

specific variation (e.g., Horner and Padian 2004; Sander and

ichthyosaurs ( Johnson 1977), pterosaurs (Bennett 1993), and

Klein 2005; Bybee et al. 2006), and overreliance on size as an

dinosaurs (Callison and Quimby 1984; Horner and Currie

age proxy may inhibit distinction between juveniles of larger

1994; Jacobs et al. 1994; Sampson et al. 1997; Carr 1999; Brill

taxa and adults of smaller, related taxa. Size-independent

and Carpenter 2001; Ryan et al. 2001, Tumarkin-Deratzian

proxies for relative somatic age are therefore necessary.

2003; Ryan and Russell 2005; Brown 2006; Brown et al. 2007).

Size-independent criteria widely used for evaluating skeletal

Textures of immature bone surfaces have been characterized

ontogenetic stages include ossification of limb bone ends

variously as porous, fibrous, or striated depending on the

(e.g., Bennett 1993), fusion of epiphyses and compound skele-

taxon and anatomical area under study, although they have

tal elements (e.g., Carey 1982; Sadler 1991; Brochu 1996;

been differentiated consistently from the denser, more ‘‘fin-

Gotfredsen 1997; Carrano et al. 2005; Irmis 2007), relative

ished’’ surfaces of mature bone. Immature bone textures have

development of cranial ornamentation (e.g., Dodson 1975,

been interpreted as indicative of rapid bone growth in early

1976; Sampson et al. 1997; Ryan et al. 2001; Ryan and Russell

ontogeny, and mature bone textures as indicative of a later

2005; Goodwin et al. 2006; Horner and Goodwin 2006, 2008),

slowing or cessation of growth ( Johnson 1977; Bennett 1993;

and bone microstructure (e.g., Bennett 1993; Chinsamy 1995;

Samson et al. 1997; Ryan et al. 2001).

Curry 1999; Erickson and Tumanova 2000; Horner et al. 2000;

The short-frilled centrosaurine ceratopsids are particularly

Sander 2000; Padian et al. 2001; Steyer et al. 2003; Botha and

well suited for study of ontogenetic trends, as several gen-

Chinsamy 2004, 2005; Erickson et al. 2004; Horner and Pa-

era (Centrosaurus, Styracosaurus, Pachyrhinosaurus, Einiosaurus,

dian 2004; Padian et al. 2004; Ray and Chinsamy 2004; Sander

and possibly Albertaceratops) are known from monodominant

et al. 2004; Carrano et al. 2005; Ray et al. 2005; Sander and

bonebed assemblages yielding disarticulated remains of indi-

Klein 2005; Bybee et al. 2006; Lee 2006, 2007; Reizner and

viduals representing a range of body sizes (Langston 1975;

Horner 2006; Erickson et al. 2007; Lee and Werning 2008).

Currie and Dodson 1984; Tanke 1988; Rogers 1990; Sampson

Another method relies on the existence and recognition of

1995; Ryan et al. 2001; Dodson et al. 2004; Eberth and Getty

252 tumarkin-deratzian

2005; Ryan and Evans 2005; Ryan and Russell 2005; Currie et

texture may be reliably distinguished from mottled texture by

al. 2007; Ryan 2007). Previous studies (Sampson et al. 1997;

the presence of finished surface bone, which is absent from

Ryan et al. 2001) have documented three relative age classes

mottled texture.

(juveniles, subadults, and adults) based on combined evi-

The consistent association, in multiple centrosaurine taxa,

dence from element size, degree of fusion of cranial bones,

of distinctive surface textures with cranial elements of particu-

degree of development of cranial ornamentation, and bone

lar sizes and degrees of fusion and ornamentation (Sampson et

surface textures.

al. 1997; Ryan et al. 2001) strongly suggests that changes in

The term ‘‘subadult’’ has the potential to be ambiguous,

bone surface textures represent an ontogenetic signal. The sig-

even when explicitly used only as a somatic growth stage,

nificance of these changes for centrosaurine paleobiology,

because various authors have applied the term in different

however, remains poorly understood. The current study ex-

ways to different taxa. Among extant archosaurs, for example,

amines the histology of centrosaurine frill elements bearing

subadult status has been assigned to individuals intermediate

examples of each texture type in order to determine the mi-

in size between juveniles and adults [Rice et al. 1999 (Alli-

crostructural basis of each texture and its relationship to the

gator)], individuals with partially fused long bone epiphyses

overall pattern of skeletal maturation in cranial elements.

[Gotfredsen 1997 (several sea bird species)], and individuals whose long bones have reached adult lengths but lack certain

Institutional Abbreviation. TMP: Royal Tyrrell Museum of Palaeontology, Drumheller.

landmarks (e.g., muscle scars) present on bones of adults [Tumarkin-Deratzian et al. 2006 (Branta)]. Among ceratopsids, Horner and Goodwin (2006) recognized four successive onto-

Materials and Methods

genetic stages (baby, juvenile, subadult, adult) in the chas-

EXAMINED MATERIAL

mosaurine Triceratops, but indicated that the names were applied only as a convenience.

Examined bones and their corresponding textures are listed

Sampson et al. (1997) defined subadult centrosaurines as

in Table 17.1. Elements were examined grossly with the un-

individuals of adult or nearly adult-size that had not yet ac-

aided eye and with a 10x hand lens. Textural features of well-

quired all adult characters (e.g., fusion of cranial elements

preserved surfaces were documented through a combination

and/or ornamentations). Ryan et al. (2001) considered bones

of detailed specimen notes, sketches, and photography. Re-

to be subadult-sized if they were between one-half and two-

gions bearing examples of each of the textural types discussed

thirds of adult size (the large, well-ornamented Yale Peabody

by Sampson et al. (1997) and Ryan et al. (2001) were thin-

Museum specimen 2015 was used as a standard for adult size

sectioned for histological analysis (Fig. 17.2).

in Centrosaurus apertus). For purposes of consistency, the definition of Ryan et al. (2001) is followed in this paper. It should be noted, however, that skeletochronological studies of sauris-

THIN-SECTION PREPARATION

chian dinosaur taxa have suggested that individual size at a

Bone segments bearing well-developed characteristic tex-

given age may be quite variable (Horner and Padian 2004;

tures were isolated by exploiting preexisting cracks in the

Sander and Klein 2005; Bybee et al. 2006), and therefore selec-

specimens or by cutting with a diamond saw in cases where

tion of a single adult size standard may not reflect true bio-

the fossil had sufficient structural integrity to withstand un-

logical reality. Whether a similar pattern holds for centrosau-

embedded cutting. The target segments were then embedded

rines, or ornithischian dinosaurs in general, requires further

in resin under vacuum. Wafers with a thickness of 3 mm were

investigation.

cut from the embedded segments at the locations indicated in

Four distinct bone surface textures have been documented

Fig. 17.2; one surface of each wafer was the desired plane of

on centrosaurine cranial elements, particularly the parietal

section for that location. During this process, freshly exposed

and squamosal (Fig. 17.1; Sampson et al. 1997; Ryan et al.

surfaces were coated under vacuum with Paleo-Bond™ Pene-

2001, Tumarkin-Deratzian 2003; Brown 2006; Brown et al.

trant Stabilizer after each cut. Treated surfaces were left un-

2007). Long-grained texture, characterized by regular fine stri-

disturbed for 24 hours to allow the sealing of any opened

ations generally oriented in the direction of longitudinal

pores and to improve the stability of the sample. The surface

growth, is typical of juvenile-sized bone. Mottled texture, char-

of the wafer corresponding to the desired plane of section was

acterized by finely pitted surfaces with an apparent lack of

then ground smooth on a graded sequence of 320–600 grit

finished surface bone, is typical of subadult-sized bone, where

silicon carbide grinding paper, and then mounted on a frosted

it generally co-occurs with long-grained texture. Smooth to

plastic slide using low-viscosity cyanoacrylate adhesive. Ex-

rugose textures (variable depending on the element in question)

cess wafer was ground away on a graduated series of 180, 320,

typify surfaces of adult-sized bones lacking both long-grained

and 600 grit silicon carbide papers. The presence of smectite-

and mottled textures. Irregular surfaces associated with rugose

rich clay matrix within cracks and pore spaces required re-

Histological Evaluation of Ontogenetic Bone Surface Texture Changes in the Frill of Centrosaurus apertus 253

FIGURE 17.1.

Characteristic surface textures of centrosaurine frill elements. (A) Long-grained (TMP 94.12.735); (B) long-grained (top) and mottled (bottom) with mottling overprinting striations of long-grained (TMP 90.57.18); (C) long-grained (top) and mottled (bottom) with mottling completely obscuring striations of long-grained (TMP 94.12.735); (D) smooth (TMP 82.18.79); (E) rugose (TMP 82.18.79). Scale bars are 1 cm in A, B, D, E and 0.5 cm in C.

peated reinforcement of the specimens during the grinding

is marked by a well-defined step-down from higher (long-

process. After grinding at each grit level, slides were air-dried

grained) to lower (mottled) levels (Fig. 17.3B).

and the specimen was coated thinly with Paleo-Bond™ Pene-

In some sections cut parallel to the grain of long-grained

trant Stabilizer and left overnight, before grinding at the next

texture, there is a change in surface contour from uniform to

finer level. Sections were ground until optically translucent

irregular at the boundary between long-grained and mottled

under the petrographic microscope, rather than to a stan-

surfaces (Fig. 17.3B [transition], C [long-grained], D [mot-

dardized thickness. Thin-sections were examined under both

tled]). In other similarly oriented sections, however, this con-

plane- and cross-polarized light.

tour distinction is lacking and surface topography of the two textures is surprisingly similar (Fig. 17.3E [long-grained], F [mottled]). Irrespective of contour, mottled regions are relia-

Histological Description LONG-GRAINED AND MOTTLED TEXTURES (TMP 90.57.18 AND 94.12.735)

bly distinguished by a surface that cuts across and truncates preexisting bony fabric (Fig. 17.3F). In sections cut perpendicular to the grain of long-grained texture, long-grained regions have a ridge and furrow surface

All sections show fibrolamellar bone tissue with well-developed

contour, reflecting incorporation of surface canals (in fibro-

primary osteons. Orientation of primary osteons and vascular

lamellar bone, future primary osteons) into the bone through

canals, especially in more surficial regions, generally paral-

appositional growth along ridges (Fig. 17.3A, G). Diameter of

lels the grain of long-grained surface patterns (Fig. 17.3A).

surface furrows and incorporated canals is in the range of 50–

The transition between long-grained and mottled regions

75 mm. Furrows in varying degrees of closure illustrate se-

254 tumarkin-deratzian

Table 17.1. Examined Specimens and Their Corresponding Textures Specimen

Element*

Texture(s)

TMP 90.57.18

Squamosal

Long-grained, mottled

TMP 94.12.735

Parietal

Long-grained, mottled

TMP 82.18.86

Parietal

Smooth

TMP 91.18.14

Parietal

Smooth

TMP 82.18.66

Parietal

Smooth, rugose

* Elements are shown in Fig. 17.2.

quential stages in canal incorporation (Fig. 17.3A, G [numbered arrows]). Mottled regions contain surface pits that may be superficially similar to the furrows of long-grained texture, but the mottled surface pits are typically larger (150 mm or more in diameter), and resorptive in origin, as shown by the fact that they truncate older primary osteons (Fig. 17.3H). In many cases the erosional character of the mottled surface is evident under plane-polarized light; however, examination under cross-polarized light, where osteonal fiber organization is more readily visible, is recommended for confirmation (Fig. 17.3I). Although the surface of immature bone shows considerable evidence of cortical resorption in mottled areas, evidence of osteonal remodeling in deeper regions of the bone is conspicuously scarce. Large erosion spaces indicative of osteonal remodeling are absent from all sections, and secondary osteons are rare to absent (Fig. 17.3J).

SMOOTH AND RUGOSE TEXTURES (TMP 82.18.86, TMP 91.18.14, and TMP 82.18.66) Sections of elements bearing smooth and rugose textures reveal a three-layered structure. The innermost region contains large erosion cavities, some with the beginnings of secondary deposition (Fig. 17.4A). This innermost layer is surrounded by complexly remodeled regions of dense Haversian bone (Fig. 17.4B, C). Remnants of original primary bone are occasionally visible between secondary osteons (Fig. 17.4B), but in most regions secondary osteons overlap and cross-cut each other, with no intervening primary bone.

FIGURE 17.2. Elements and regions selected for histological analysis. (A, B) long-grained and mottled textures (TMP 90.57.18); (C, D) long-grained and mottled textures (TMP 94.12.735); (E, F) rugose and smooth textures (TMP 82.18.66); (G, H) smooth texture with shallow surface lineations (TMP 91.18.14); (I, J) smooth texture (TMP 82.18.86). Arrows on the left images indicate target regions enlarged on the right, black lines and lowercase letters indicate orientations of individual thin-sections. Scale bars are 5 cm in A, C, E, G, I and 1 cm in B, D, F, H, J.

Histological Evaluation of Ontogenetic Bone Surface Texture Changes in the Frill of Centrosaurus apertus 255

FIGURE 17.3. Histology of long-grained and mottled textures. (A) Section oriented perpendicular to long-grained striations showing preferred orientation of vascular canals and primary osteons (arrows) parallel to grain (TMP 90.57.18, section b, ppl); (B) step-down (arrow) from long-grained (left) to mottled (right) regions (TMP 94.12.735, section b, ppl); (C) uniform surface contour in section oriented parallel to long-grained striations (TMP 90.57.18, section a, ppl); (D) irregular surface contour associated with mottled texture (TMP 90.57.18, section a, ppl); (E) uniform surface contour in section oriented parallel to long-grained striations (TMP 94.12.735, section a, ppl); (F) relatively uniform surface contour associated with mottled texture, distinguished from long-grained region in E by crosscutting of preexisting bony fabric (arrows) by mottled surface (TMP 94.12.735, section a, ppl); (G) ridge and furrow contour in section perpendicular to long-grained striations, arrow indicates furrow nearly closed by appositional growth on adjacent ridges (TMP 94.12.735, section b, ppl); (H) surface pits in mottled region crosscutting preexisting bony fabric (arrows) (TMP 90.57.18, section b, ppl); (I) surface pits in mottled region crosscutting preexisting bony fabric (arrows) (TMP 94.12.735, section b, cpl); ( J) innermost region of element with long-grained and mottled textures, showing lack of osteonal remodeling, dorsal surface toward top of photo (TMP 94.12.735, section b, ppl). Numbered arrows in (A) and (G) indicate increasing degrees of surface canal closure. Black lines covering bone surfaces are preparation artifact. Scale bars are 200 mm. cpl: cross-polarized light; ppl: plane-polarized light.

256 tumarkin-deratzian

FIGURE 17.4. Histology of smooth (A–E) and rugose (F–G) textures. (A) Remodeled innermost region showing erosion cavities (e) and secondary deposition (arrow) (TMP 82.18.86, section a, cpl); (B) highly remodeled region with overlapping secondary osteons (arrowhead) and remnants of primary bone (p) separated by sharp boundary (arrows) from overlying lamellar bone tissue (*) (TMP 82.18.86, section a, ppl); (C) highly remodeled region with overlapping secondary osteons (arrowhead) separated by sharp boundary (arrows) from overlying lamellar bone tissue (*) (TMP 91.18.14, section a, cpl); (D) remodeled region separated by sharp boundary (arrows) from overlying lamellar bone tissue (*), (TMP 91.18.14, section b, ppl); (E) gently undulating contour of smooth surface, associated grossly with shallow surface lineations (see Fig. 17.2H), * indicates outer layers of lamellar bone tissue (TMP 91.18.14, section a, ppl); (F) irregular surface contour associated with rugose texture (TMP 82.18.66, section a, ppl); (G) surficial erosion bay (arrows) with secondary deposition (s) associated with rugose texture (TMP 82.18.66, section a, cpl). Scale bars are 200 mm. Abbreviations as in Figure 17.3.

Histological Evaluation of Ontogenetic Bone Surface Texture Changes in the Frill of Centrosaurus apertus 257

External to these remodeled regions, unweathered surfaces

Long-grained texture is not unique to centrosaurine cra-

are underlain by one or more thin layers of lamellar bone

nial material, and this is not surprising, given the close con-

tissue (Fig. 17.4B–E), which may be sporadically interrupted

nection between striated surface pattern and growth-related

by isolated secondary osteons. Often there is a clear demarca-

incorporation of periosteal vasculature (Tumarkin-Deratzian

tion between this superficial region and the densely remod-

et al. 2006). Long-grained or striated bone textures have also

eled internal tissue; this boundary is best defined under cross-

been noted in centrosaurine postcrania (Sampson et al. 1997;

polarized light, but may be visible under plane-polarized light

Tumarkin-Deratzian 2003), as well as in ornithopods (Samp-

in some cases (Fig. 17.4B–D). In some locations, the superficial

son et al. 1997; Evans et al. 2005), non-avian theropods

region has a gently undulating surface contour, similar to that

(Sampson et al. 1997; Carr 1999), and extant birds (Calli-

associated with the ridge and furrow system in sections per-

son and Quimby 1984; Gotfredsen 1997; Serjeantson 1998;

pendicular to long-grained surfaces (Fig. 17.4E). Grossly, the

Tumarkin-Deratzian et al. 2006). Mottled surface texture, in

undulating pattern is associated with shallow surface linea-

contrast, is at present known only from subadult-sized cen-

tions (Fig. 17.2H), which are similar to, but much less well-

trosaurine skull elements.

defined than, the long-grained striations.

Sampson et al. (1997) described the co-occurrence of long-

Rugose surfaces are associated with irregular surface con-

grained and mottled textures as indicative of bone in which

tours (Fig. 17.4F, G). The superficial lamellar layer is often dis-

certain regions are still experiencing rapid growth (long-

continuous, with the surface marked by complex crenellated

grained texture), whereas in other areas growth has slowed

contours and resorption pits. In at least one location a resorp-

(mottled texture). Ryan et al. (2001) hypothesized that mot-

tion pit is lined by secondary deposition (Fig. 17.4G).

tled texture was transitional between the long-grained and smooth/rugose textures, with the pitting characteristic of

Discussion

mottled texture resulting from disruption of the striated longgrained pattern by newly forming smooth or rugose textures.

Sampson et al. (1997) associated long-grained texture

This latter hypothesis is consistent with gross patterns, and

with rapid bone growth in immature bone, and non-striated

draws support from the fact that two varieties of mottled sur-

textures with a slowing or cessation of growth in mature bone.

faces exist (Ryan et al. 2001; Tumarkin-Deratzian 2003). The

Patterns of bone microstructure in Centrosaurus apertus

first is an overprint of still-visible remnant striations (Fig.

and modern archosaurs support this hypothesis. Tumarkin-

17.1B); the second has an amorphous structure with no re-

Deratzian et al. (2006) documented that the ridge and furrow

maining traces of the long-grained texture (Fig. 17.1C). It is

surface topography of long-grained striations on modern

reasonable to assume that these represent successive stages in

avian bone is histologically associated with incorporation of

the development of mottled surfaces, in which widespread

subperiosteal vasculature into bone matrix during growth, as

resorption of surface bone causes overprint and eventual oblit-

described by Enlow and Brown (1958) and Cormack (1987).

eration of long-grained striations. The histological data re-

Capillaries running within the furrows are gradually incor-

ported here indicate that the smooth and rugose textures form

porated within the bone matrix as a result of appositional

later as a result of a separate phase of renewed surface deposi-

growth along the ridges. Although the bony channels created

tion that follows the mottled resorptive stage. The sharp de-

by this process are referred to as vascular canals, they may also

marcation between highly remodeled inner regions and the

contain lymphatic vessels, nerves, and variable amounts of

comparatively unremodeled superficial layers of smooth and

connective tissue, and therefore the diameter of the canals is

rugose elements marks the former mottled erosional surface.

not a direct indicator of the percent vascularity of the bone.

Widespread resorption of cranial surface bone appears to

In mature bone, where growth has slowed, incorporation of

have been a normal aspect of centrosaurine growth. Mottled

periosteal vasculature is much reduced, accounting for reduc-

texture has been documented in subadult-sized individuals of

tion and loss of striated surfaces. Interestingly, the ridge and

several centrosaurine genera (Sampson et al. 1997; Ryan et al.

furrow pattern recurs, albeit in a more subtle form, during

2001; Ryan and Russell 2005); this argues against individual or

deposition of the new superficial layer that overlies the highly

population-wide pathologies. It is also not confined to bones

remodeled inner regions of smooth-textured centrosaurine

from any single depositional environment. Subadult-sized el-

bone, where it is again associated with muted surface linea-

ements with mottled texture are found in centrosaurine bone-

tions. These feature are presumably linked to renewed incor-

beds associated with paleochannel and overbank facies in wet

poration of periosteal vasculature as the new bony surface is

coastal plain environments (Eberth 2005; Eberth and Getty

deposited; however, the new superficial layer is largely com-

2005), as well as in bonebeds associated with semi-arid sea-

posed of low-vascularity lamellar bone, and the conspicuous

sonal indicators (Rogers 1990). The element-specific nature of

long-grained striations associated with the earlier deposition

mottled texture—it occurs on the parietal, squamosal, jugal,

of fibrolamellar tissue do not recur.

nasal, and postorbital (Sampson et al. 1997; Ryan et al. 2001;

258 tumarkin-deratzian

Brown 2006), but not in the postcrania—and its absence from

2006, 2007) indicates that centrosaurine growth rates were

bones of other taxa preserved in the same deposits further

likely highest in the first 3 years of life, after which growth

argues against a taphonomic origin.

slowed. Lee (2006, 2007) identifies weakly expressed LAGs in

The question remains, however, as to the significance of this

Centrosaurus; these features are assumed to be annual based on

apparently unique mottled texture for deciphering the paleo-

repeated validation of annual periodicity in modern species

biology of subadult centrosaurines. Ryan et al. (2001) attrib-

(e.g., de Buffrenil 1980; Hutton 1986; Castanet et al. 1988,

uted this texture to extreme remodeling associated with the

1993, 2004; Klinger and Musick 1992; Tucker 1997; Coles et al.

development of taxon-specific cranial ornamentation, which

2001). Using skeletal growth curves reconstructed from mod-

occurs in centrosaurines at or near adult size. This could ac-

els of bone circumferential growth, Lee (2006, 2007) estimates

count for its absence in the postcranial skeleton, the bones of

that Centrosaurus individuals would have reached 30% of their

which do not undergo such drastic shape changes. This does

adult body size by the age of 1 year, and 70% of their adult

little to clarify, however, why similar texture has not been

body size by age 2.

reported in chasmosaurine ceratopsids, or in other dinosau-

Considering Lee’s (2006, 2007) results in concert with the

rian groups, such as lambeosaurine hadrosaurs, that also expe-

definition of subadult-sized individuals as one-half to

rienced significant ontogenetic changes in skull morphology

two-thirds adult size, one could suppose that the period of

(Dodson 1975; Evans et al. 2005, 2007). Absence of evidence

subadult growth associated with the co-occurrence of long-

for mottled texture elsewhere in Ceratopsidae may relate to

grained and mottled cranial textures would have occurred

the relative paucity of subadult chasmosaurine material (Dod-

during the second year of life. Moreover, if slowing of growth

son et al. 2004). At present, reasonably complete growth series

in centrosaurines was related to reproductive status, as has

are known only for Triceratops (Horner and Goodwin 2006,

been hypothesized for growth decelerations in other dino-

2008) and Chasmosaurus mariscalensis (Lehman 2007), and

saurian taxa (Rimblot-Baly et al. 1995; Curry 1999; Sander

no detailed studies of cranial bone surface textures have been

2000; Lee and Werning 2008), one could consider 3 years an

undertaken.

approximate age for sexual maturity, with attendant devel-

Recent work by Brown et al. (2007, in press) detailing the

opment of purported secondary sex characteristics such as

topographic expression of textures on a size-series of centro-

taxon-specific cranial ornamentation. Although highly specu-

saurine frills reveals that mottled texture is not confined spe-

lative, such a sequence of cranial textural change during the

cifically to sites where characteristic adult ornamentation de-

second year followed by sexual maturity at age 3 would be

velops. Moreover, there is in many cases a substantial size

consistent with the temporal disconnect between textural

difference (and presumed temporal lag) between replacement

transition and ornamentation development observed by

of long-grained by mottled texture and appearance of taxon-

Brown et al. (in press). Detailed examination of the relation-

specific ornamentation (Brown et al. in press). Although these

ship between cranial surface textures and ornamentation on

findings do not preclude a connection between mottled tex-

the one hand, and somatic and sexual maturity on the other,

ture and development of cranial ornamentation, they strongly

would require combined analyses of cranial morphology and

argue that the link, if present, may be less direct than pre-

long bone histology in associated elements from a size series

viously supposed.

of multiple individuals.

Compared to that of other dinosaurian groups, the histology of ceratopsian bone has received relatively little attention in

Conclusions

the literature, and thus the general growth regime of these dinosaurs is only beginning to be understood. Studies to date

Centrosaurus apertus frill elements with long-grained and mot-

have focused on the postcranial skeleton. Long bone histology

tled bone surface textures are composed of primary fibro-

of Psittacosaurus (Erickson and Tumanova 2000) and Pro-

lamellar bone tissue with little to no osteonal remodeling.

toceratops (Chinsamy 1994; Chinsamy and Dodson 1995;

Long-grained texture is associated with periosteal deposition.

Makovicky et al. 2007) has shown alternating patterns of zones

Active incorporation of vascular canals along periosteal sur-

and annuli and/or lines of arrested growth (LAGs). Chinsamy

faces accounts for the striated pattern of this texture; canal

(1994) described a single femoral section of the centrosaurine

orientation is predominantly longitudinal with respect to the

Pachyrhinosaurus, which showed uninterrupted deposition of

grossly visible grain of the surface striations. Mottled texture is

fibrolamellar bone, and interpreted this as evidence of rela-

associated histologically with widespread resorption of sur-

tively rapid growth to adult or near-adult size. LAGs were

face bone, leading grossly to overprint and gradual oblitera-

noted only late in ontogeny, suggesting a slowing or cessation

tion of striations.

of growth as skeletal maturity and adult size were reached.

Elements with smooth and rugose surfaces are generally

More recent work on long bone histology of growth series of

highly remodeled, with abundant secondary osteons and

Einiosaurus (Reizner and Horner 2006) and Centrosaurus (Lee

large erosion bays in the innermost regions of the bone.

Histological Evaluation of Ontogenetic Bone Surface Texture Changes in the Frill of Centrosaurus apertus 259

Smooth texture is associated with new bony deposition on the formerly resorptive mottled surface. Shallow ridges and furrows may recur during deposition of the new superficial layer; these are associated with muted surface lineations more subtle than the long-grained striations. In areas with rugose texture, effects of remodeling extend to periosteal surface regions. Occurrences of secondary osteons at or near periosteal surfaces are more common, as are erosion bays, some showing the beginnings of secondary deposition within them. The observed histological sequence of deposition followed by resorption followed by renewed deposition lends support to previous hypotheses relating changes from long-grained to mottled to smooth/rugose textures to skeletal maturation of cranial elements during centrosaurine ontogeny. Based on the common occurrence of mottled texture in subadult sized elements, widespread resorption of cranial surface bone, particularly in the frill region, was apparently a normal feature of centrosaurine ontogeny. Previous hypotheses linking mottled texture to extreme bone remodeling are consistent with the histological evidence. However, the link between such remodeling and shape changes associated with development of cranial ornamentation is called into question by recent studies (Brown et al. in press) documenting a spatial and temporal disconnect between mottling and those morphological changes. Further research is necessary in order to assess the importance of resorption and mottled surface texture for skeletal maturation of the centrosaurine skull, as well as to investigate the timing of these somatic growth changes in relation to sexual maturity. Acknowledgments

D. Eberth and J. Gardner (Royal Tyrrell Museum of Palaeontology) granted access to specimens under their care. A. Lee, A. Russell, and M. Ryan reviewed an earlier draft of this manuscript and provided much helpful criticism. C. Brown, A. Chinsamy-Turan, P. Dodson, D. Eberth, B. Grandstaff, S. Sampson, and D. Tanke provided much stimulating and informative discussion. D. Deratzian and D. Terry provided technical assistance with photomicrographs. Portions of this work were supported by the Geological Society of America, the Paleontological Society, the University of Pennsylvania Summer Stipends in Paleontology program, a National Science Foundation Graduate Research Fellowship, and National Science Foundation Grant EAR 95-06694 to P. Dodson. References Cited Andrews, R. M. 1982. Patterns of growth in reptiles. In C. Gans and F. H. Pough, eds., Biology of the Reptilia. Vol. 13: Physiology D, pp. 273–320. New York: Academic Press. Apio, A., M. Plath, R. Tiedemann, and T. Wronski. 2007. Agedependent mating tactics in male bushbuck (Tragelaphus scriptus). Behaviour 144: 585–610.

260 tumarkin-deratzian

Araújo, A., M. F. Arruda, A. I. Alencar, F. Albuquerque, M. C. Nascimento, and M. E. Yamamoto. 2000. Body weight of wild and captive common marmosets (Callithrix jacchus). International Journal of Primatology 21: 317–324. Benecke, N. 1993. On the utilization of the domestic fowl in central Europe from the Iron Age up to the Middle Ages. Archaeofauna 2: 21–31. Bennett, S. C. 1993. The ontogeny of Pteranodon and other pterosaurs. Paleobiology 19: 92–106. Botha, J., and A. Chinsamy. 2004. Growth and life habits of the Triassic cynodont Trirachodon, inferred from bone histology. Acta Palaeontologica Polonica 49: 619–627. ———. 2005. Growth patterns of Thrinaxodon liorhinus, a nonmammalian cynodont from the lower Triassic of South Africa. Palaeontology 48: 385–394. Brill, K., and K. Carpenter. 2001. A baby ornithopod from the Morrison Formation of Garden Park, Colorado. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Palaeontology of Philip J. Currie, pp. 197–205. Bloomington: Indiana University Press. Brinkman, D. 1988. Size-independent criteria for estimating relative age in Ophiacodon and Dimetrodon (Reptilia, Pelycosauria) from the Admiral and Lower Belle Plains Formations of westcentral Texas. Journal of Vertebrate Paleontology 8: 172–180. Brochu, C. A. 1996. Closure of neurocentral sutures during crocodilian ontogeny: Implications for maturity assessment in fossil archosaurs. Journal of Vertebrate Paleontology 16: 49–62. Brown, C. 2006. Surficial bone texture of centrosaurines (Archosauria: Dinosauria: Ceratopsidae): Can patterns be delimited? B.Sc. thesis. University of Calgary, Calgary. Brown, C., A. Russell, and M. J. Ryan. 2007. The developmental patterns of surficial bone texture on the skulls of centrosaurine dinosaurs. Journal of Vertebrate Paleontology 27(3, Suppl.): 53A. ———. In press. Surficial bone texture of the centrosaurine parietal: Ontogenetic patterns of transition and their taxonomic implications. Journal of Vertebrate Paleontology. Bybee, P. J., A. H. Lee, and E.-T. Lamm. 2006. Sizing the Jurassic theropod dinosaur Allosaurus: Assessing growth strategy and the evolution of ontogenetic scaling of limbs. Journal of Morphology 267: 347–359. Callison, G., and H. M. Quimby. 1984. Tiny dinosaurs: Are they fully grown? Journal of Vertebrate Paleontology 3: 200–209. Carey, G. 1982. Ageing and sexing domestic bird bones from some Late Medieval deposits at Baynard’s Castle, City of London. In B. Wilson, C. Grigson, and S. Payne, eds., Ageing and Sexing Animal Bones from Archaeological Sites, pp. 263–268. B.A.R. British Series, Vol. 109. Carr, T. D. 1999. Craniofacial ontogeny in Tyrannosauridae (Dinosauria, Coelurosauria). Journal of Vertebrate Paleontology 19: 497–520. Carrano, M. T., J. R. Hutchinson, and S. D. Sampson. 2005. New information on Segisaurus halli, a small theropod dinosaur from the Early Jurassic of Arizona. Journal of Vertebrate Paleontology 25: 835–849. Castanet, J., S. Croci, F. Aujard, M. Perret, J. Cubo, and E. de Margerie. 2004. Lines of arrested growth in bone and age estima-

tion in a small primate: Microcebus murinus. Journal of Zoology 263: 31–39. Castanet, J., H. Francillon-Vieillot, F. J. Meunier, and A. de Ricqlès. 1993. Bone and individual aging. In B. K. Hall, ed., Bone. Vol. 7: Bone Growth B, pp. 245–283. Boca Raton, Fla.: CRC Press. Castanet, J., D. G. Newman, and H. Saint-Girons. 1988. Skeletochronological data on the growth, age and population structure of the tuatara, Sphenodon punctatus, on Stephens and Lady Alice Islands, New Zealand. Herpetologica 44: 25–37. Chabreck, R. H., and T. Joanen. 1979. Growth rates of American alligators in Louisiana. Herpetologica 35: 51–57. Chaloupka, M., and C. Limpus. 2005. Estimates of sex- and ageclass-specific survival probabilities for a southern Great Barrier Reef green sea turtle population. Marine Biology 146: 1251– 1261. Chaloupka, M., C. Limpus, and J. Miller. 2004. Green turtle somatic growth dynamics in a spatially disjunct Great Barrier Reef metapopulation. Coral Reefs 23: 325–335. Chinsamy, A. 1994. Dinosaur bone histology: Implications and inferences. In G. D. Rosenberg and D. L. Wolberg, eds., Dinofest: Proceedings of a Conference for the General Public, pp. 213– 227. Paleontological Society Special Publication 7. ———. 1995. Ontogenetic changes in the bone histology of the Late Jurassic ornithopod Dryosaurus lettowvorbecki. Journal of Vertebrate Paleontology 15: 96–104. Chinsamy, A., and P. Dodson. 1995. Inside a dinosaur bone. American Scientist 83: 174–180. Cohen, A., and D. Serjeantson. 1996. A Manual for the Identification of Bird Bones from Archaeological Sites, revised ed. London: Archetype Publications. Coles, W. C., J. A. Musick, and L. Williamson. 2001. Skeletochronology: Validation from an adult loggerhead (Caretta caretta). Copeia 2001: 240–242. Cormack, D. H. 1987. Ham’s Histology. Philadelphia: J. B. Lippincott. Craighead, J. J., and D. S. Stockstad. 1964. Breeding age of Canada geese. Journal of Wildlife Management 28: 57–64. Currie, P. J., and P. Dodson. 1984. Mass death of a herd of ceratopsian dinosaurs. In W. E. Reif and F. Westphal, eds., Short Papers of the Third Symposium on Mesozoic Terrestrial Ecosystems, pp. 61–66. Tübingen: Attempto Verlag. Currie, P. J., W. Langston, and D. H. Tanke. 2007. A new pachyrhinosaur from the Wapiti Formation of Grande Prairie, Alberta, Canada. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, p. 22. Drumheller: Royal Tyrrell Museum of Palaeontology. Curry, K. A. 1999. Ontogenetic histology of Apatosaurus (Dinosauria: Sauropoda): New insights on growth rates and longevity. Journal of Vertebrate Paleontology 19: 654–665. de Buffrenil, V. 1980. Données préliminaires sur la structure des marques de croissance squelettiques chez les crocodiliens actuels et fossiles. Bulletin de la Société Zoologique de France 105: 355–361. Dodson, P. 1975. Taxonomic implications of relative growth in lambeosaurine hadrosaurs. Systematic Zoology 24: 37–54.

———. 1976. Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. Journal of Paleontology 50: 929– 940. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A. 2005. The geology. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 54–82. Bloomington: Indiana University Press. Eberth, D. A., and M. A. Getty. 2005. Ceratopsian bonebeds: Occurrence, origins, and significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 501–536. Bloomington: Indiana University Press. Enlow, D. H., and S. O. Brown. 1958. A comparative histological study of fossil and recent bone tissues, Part III. Texas Journal of Science 10: 187–230. Erickson, G. M., K. Curry Rogers, D. J. Varricchio, M. A. Norell, and X. Xu. 2007. Growth patterns in brooding dinosaurs reveals the timing of sexual maturity in non-avian dinosaurs and genesis of the avian condition. Biology Letters 3: 558–561. Erickson, G. M., P. J. Makovicky, P. J. Currie, M. A. Norell, S. A. Yerby, and C. A. Brochu. 2004. Gigantism and comparative life-history parameters of tyrannosaurid dinosaurs. Nature 430: 772–775. Erickson, G. M., and T. A. Tumanova. 2000. Growth curve of Psittacosaurus mongoliensis Osborn (Ceratopsia: Psittacosauridae) inferred from long bone histology. Zoological Journal of the Linnean Society 130: 551–566. Evans, D. C., C. A. Forster, and R. R. Reisz. 2005. The type specimen of Tetragonosaurus erectofrons (Ornithischia: Hadrosauridae) and the identification of juvenile lambeosaurines. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 349–365. Bloomington: Indiana University Press. Evans, D. C., R. Reisz, and K. Dupuis. 2007. A juvenile Parasaurolophus (Ornithischia: Hadrosauridae) braincase from Dinosaur Provincial Park, Alberta, with comments on crest ontogeny in the genus. Journal of Vertebrate Paleontology 27: 642–650. Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian. 2006. The smallest known Triceratops skull: New observations on ceratopsid cranial anatomy and ontogeny. Journal of Vertebrate Paleontology 26: 103–112. Gotfredsen, A. B. 1997. Sea bird exploitation on coastal Inuit sites, West and Southeast Greenland. International Journal of Osteoarchaeology 7: 271–286. Hanson, H. C. 1965. The Giant Canada Goose. Carbondale: Southern Illinois University Press. ———. 1967. Characters of age, sex, and sexual maturity in Canada geese. Natural History Survey Division Biological Notes No. 49. Urbana: State of Illinois Department of Registration and Education. Horner, J. R., and P. J. Currie. 1994. Embryonic and neonatal

Histological Evaluation of Ontogenetic Bone Surface Texture Changes in the Frill of Centrosaurus apertus 261

morphology and ontogeny of a new species of Hypacrosaurus (Ornithischia: Lambeosauridae) from Montana and Alberta. In K. Carpenter, K. F. Hirsch, and J. R. Horner, eds., Dinosaur Eggs and Babies, pp. 312–336. Cambridge: Cambridge University Press. Horner, J. R., A. de Ricqlès, and K. Padian. 2000. Long bone histology of the hadrosaurid dinosaur Maiasaura peeblesorum: Growth dynamics and physiology based on an ontogenetic series of skeletal elements. Journal of Vertebrate Paleontology 20: 115–129. Horner, J. R., and M. B. Goodwin. 2006. Major cranial changes during Triceratops ontogeny. Proceedings of the Royal Society B 273: 2757–2761. ———. 2008. Ontogeny of cranial epi-ossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Horner, J. R., and K. Padian. 2004. Age and growth dynamics of Tyrannosaurus rex. Proceedings of the Royal Society B 271: 1875– 1880. Hoyer, R. F., and G. R. Stewart. 2000. Biology of the rubber boa (Charina bottae), with emphasis on C. b. umbratica. Part I: Capture, size, sexual dimorphism, and reproduction. Journal of Herpetology 34: 348–354. Hutton, J. M. 1986. Age determination of living Nile crocodiles from the cortical stratification of bone. Copeia 1986: 332–341. Irmis, R. B. 2007. Axial skeleton ontogeny in the Parasuchia (Archosauria: Pseudosuchia) and its implications for ontogenetic determination in archosaurs. Journal of Vertebrate Paleontology 27: 350–361. Jacobs, L. L., D. A. Winkler, P. A. Murry, and J. M. Maurice. 1994. A nodosaurid scuteling from the Texas shore of the Western Interior Seaway. In K. Carpenter, K. F. Hirsch, and J. R. Horner, eds., Dinosaur Eggs and Babies, pp. 337–346. Cambridge: Cambridge University Press. Johnson, R. 1977. Size independent criteria for estimating relative age and the relationships among growth parameters in a group of fossil reptiles (Reptilia: Ichthyosauria). Canadian Journal of Earth Sciences 14: 1916–1924. Klinger, R. C., and J. A. Musick. 1992. Annular growth layers in juvenile loggerhead turtles (Caretta caretta). Bulletin of Marine Science 51: 224–230. Langston, W., Jr. 1975. The ceratopsian dinosaurs and associated lower vertebrates from the St. Mary River Formation (Maestrichtian) at Scabby Butte, southern Alberta. Canadian Journal of Earth Sciences 12: 1576–1608. Lee, A. 2006. Evolution of rapid limb growth and vascular canal organization in ceratopsian dinosaurs. Journal of Vertebrate Paleontology 26(3, Suppl.): 89A. ———. 2007. How Centrosaurus (and other ceratopsians) grew to large size. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 105–106. Drumheller: Royal Tyrrell Museum of Palaeontology. Lee, A., and S. Werning. 2008. Sexual maturity in growing dinosaurs does not fit reptilian growth models. Proceedings of the National Academy of Sciences 105: 582–587. Lehman, T. M. 2007. Growth and population age structure in the horned dinosaur Chasmosaurus. In K. Carpenter, ed., Horns and

262 tumarkin-deratzian

Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 259–317. Bloomington: Indiana University Press. Makovicky, P. J., G. M. Erickson, and M. A. Norell. 2007. Life history of Protoceratops andrewsi from Bayn Zag, Mongolia. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 113–114. Drumheller: Royal Tyrrell Museum of Palaeontology. Mannermaa, K. 2002. Bird bones from Jettböle I, a site in the Neolithic Åland archipelago in the northern Baltic. Acta Zoologica Cracoviensia 45: 85–98. Owen, M. 1980. Wild Geese of the World: Their Life History and Ecology. London: B. T. Batsford. Padian, K., J. R. Horner, and A. de Ricqlès. 2004. Growth in small dinosaurs and pterosaurs: The evolution of archosaurian growth strategies. Journal of Vertebrate Paleontology 24: 555– 571. Padian, K., A. de Ricqlès, and J. R. Horner. 2001. Dinosaurian growth rates and bird origins. Nature 412: 405–408. Ray, S., and A. Chinsamy. 2004. Diictodon feliceps (Therapsida, Dicynodontia): Bone histology, growth, and biomechanics. Journal of Vertebrate Paleontology 24: 180–194. Ray, S., A. Chinsamy, and S. Bandyopadhyay. 2005. Lystrosaurus murrayi (Therapsida, Dicynodontia): Bone histology, growth and lifestyle adaptations. Palaeontology 48: 1169–1185. Reizner, J., and J. Horner. 2006. An ontogenetic series of the ceratopsid dinosaur Einiosaurus procurvicornis as determined by long bone histology. Journal of Vertebrate Paleontology 26(3, Suppl.): 114A. Rice, K. G., H. F. Percival, A.R. Woodward, and M. L. Jennings. 1999. Effects of egg and hatchling harvest on American alligators in Florida. Journal of Wildlife Management 63: 1193–1200. Ricklefs, R. E. 1968. Patterns of growth in birds. Ibis 110: 419– 451. Rimblot-Baly, F., A. de Ricqlès, and L. Zylberberg. 1995. Analyse paléohistologique d’une série de croissance partielle chez Lapparentosaurus madagascarensis ( Jurassique Moyen): Essai sur la dynamique de croissance d’un dinosaure sauropode. Annales de Paléontologie 81: 49–86. Rogers, R. R. 1990. Taphonomy of three dinosaur bonebeds in the Upper Cretaceous Two Medicine Formation of northwestern Montana: Evidence for drought-related mortality. Palaios 5: 394–413. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., and D. C. Evans. 2005. Ornithischian dinosaurs. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 312–348. Bloomington: Indiana University Press. Ryan, M. J., and Russell, A. P. 2005. A new centrosaurine ceratopsid from the Oldman Formation of Alberta and its implications for centrosaurine taxonomy and systematics. Canadian Journal of Earth Sciences 42: 1369–1387. Ryan, M. J., Russell, A. P., D. A. Eberth, and P. J. Currie. 2001. The taphonomy of a Centrosaurus (Ornithischia: Ceratopsidae) bonebed from the Dinosaur Park Formation (Upper Campa-

nian), Alberta, Canada, with comments on cranial ontogeny. Palaios 16: 482–506. Sadler, P. 1991. The use of tarsometatarsi in sexing and ageing domestic fowl (Gallus gallus L.), and recognising five toed breeds in archaeological material. Circaea 8: 41–48. Sampson, S. D. 1995. Horns, herds, and hierarchies. Natural History 104: 36–40. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Sander, P. M. 2000. Longbone histology of the Tendaguru sauropods: Implications for growth and biology. Paleobiology 26: 466–488. Sander, P. M., and N. Klein. 2005. Developmental plasticity in the life history of a prosauropod dinosaur. Science 310: 1800–1802. Sander, P. M., N. Klein, E. Buffetaut, G. Cuny, V. Suteethorn, and J. Le Loeuff. 2004. Adaptive radiation in sauropod dinosaurs: Bone histology indicates rapid evolution of giant body size through acceleration. Organisms, Diversity, and Evolution 4: 165–173. Sanz, J. L., L. M. Chiappe, B. P. Pérez-Moreno, J. J. Moratalla, F. Hernández-Carrasquilla, A. D. Buscalioni, F. Ortega, F. J. Poyato-Ariza, D. Rasskin-Gutman, and X. Martinez-Delclòs. 1997. A nestling bird from the Lower Cretaceous of Spain: Implications for avian skull and neck evolution. Science 276: 1543–1546. Serjeantson, D. 1998. Birds: A seasonal resource. Environmental Archaeology 3: 23–33. ———. 2002. Goose husbandry in Medieval England, and the problem of ageing goose bones. Acta Zoologica Cracoviensia 45: 39– 54. Shine, R., and E. L. Charnov. 1992. Patterns of survival, growth, and maturation in snakes and lizards. American Naturalist 139: 1257–1269.

Starck, J. M., and A. Chinsamy. 2002. Bone microstructure and developmental plasticity in birds and other dinosaurs. Journal of Morphology 254: 232–246. Steyer, J. S., M. Laurin, J. Castanet, and A. de Ricqlès. 2003. First histological and skeletochronological data on temnospondyl growth: Palaeoecological and palaeoclimatological implications. Palaeogeography, Palaeoclimatology, Palaeoecology 206: 193–201. Tanke, D. 1988. Ontogeny and dimorphism in Pachyrhinosaurus (Reptilia, Ceratopsidae), Pipestone Creek, N.W. Alberta, Canada. Journal of Vertebrate Paleontology 8(3, Suppl.): 27A. Tucker, A. D. 1997. Validation of skeletochronology to determine age of freshwater crocodiles (Crocodylus johnstoni ). Marine and Freshwater Research 48: 343–351. Tumarkin-Deratzian, A. R. 2003. Bone surface textures as ontogenetic indicators in extant and fossil archosaurs: Macroscopic and histological evaluations. Ph.D. diss., University of Pennsylvania, Philadelphia. Tumarkin-Deratzian, A. R., D. R. Vann, and P. Dodson. 2006. Bone surface texture as an ontogenetic indicator in long bones of the Canada goose Branta canadensis (Anseriformes: Anatidae). Zoological Journal of the Linnean Society 148: 133– 168. ———. 2007. Growth and textural ageing in long bones of the American alligator Alligator mississippiensis (Crocodylia: Alligatoridae). Zoological Journal of the Linnean Society 150: 1–39. Voelker, G. 1997. The molt cycle of the arctic tern, with comments on aging criteria. Journal of Field Ornithology 68: 400–412. Wilkinson, P. M., and W. E. Rhodes. 1997. Growth rates of American alligators in coastal South Carolina. Journal of Wildlife Management 61: 397–402.

Histological Evaluation of Ontogenetic Bone Surface Texture Changes in the Frill of Centrosaurus apertus 263

18 Modeling Structural Properties of the Frill of Triceratops ANDREW A. FARKE, RALPH E. CHAPMAN, AND ART ANDERSEN

Introduction

Triceratops has an unusual parietosquamosal frill among ceratopsids, with a strongly arched profile both in lateral and rostral views, as well as complete absence of parietal

The parietosquamosal frill of the chasmosaurine ceratopsid

fenestrae. Previous workers have suggested that the frill

Triceratops displays unusual morphology relative to the condi-

may have had a defensive role, protecting the neck from

tion in many other neoceratopsians. Whereas most derived

the horns of other Triceratops or from attacks by preda-

neoceratopsians (e.g., Protoceratops, Centrosaurus, and Chasmo-

tors. Yet these functions have not been assessed within a

saurus) possess prominent parietal fenestrae, the frill of Tri-

biomechanical context. In order to evaluate the struc-

ceratops has no such openings. Furthermore, Triceratops has a

tural properties of the frill of Triceratops, a three-

relatively shorter frill (as compared to basal skull length) than

dimensional digital model was constructed from surface

seen in most other chasmosaurines ceratopsids, and the frill

scan data of an original fossil specimen. This model was

displays prominent mediolateral and rostrocaudal arching

analyzed using finite element analysis, in which the ef-

(Fig. 18.1; Forster 1996). This unique combination of charac-

fects of various loads to the frill were simulated. Greatest

ters has led to speculation that the frill of Triceratops was an

overall stresses within the frill occur under loads applied

effective protective shield against predators or the horns of

to its most distal portions, as would be expected if the

other Triceratops (e.g., Hatcher et al. 1907; Lull 1908; Lull

frill behaves at least in part as a cantilevered beam. Strain

1933).

energy density (reflecting where the forces of loads are

Significantly, only one previous study has considered the

absorbed) generally is confined to the element experienc-

mechanical properties of the ceratopsian frill (Tyson 1977).

ing the loading. In particular, the arched profile of the

Tyson focused in particular on the properties of the frill for

squamosal seems to be particularly effective at prevent-

counterbalancing the weight of the front half of the skull. She

ing loads to the squamosal from affecting the parietal.

also noted that the frill in some ceratopsids was constructed as

This suggests that some structural features of ceratopsian

a ‘‘perfect frame,’’ and hypothesized that the parietal fenestrae

frills, such as the arched profile or thickened medial bars

probably formed in areas of relatively low stress under typical

on the squamosals of certain ceratopsians, may have

loads. Here, we investigate the structural properties of the un-

played a role in maintaining the structural integrity of

usual frill of Triceratops. The present study is not intended as a

the frill under externally applied loads (such as those

direct test of hypotheses of frill function, but instead an explo-

from the horns of other Triceratops).

ration of how the frill would have behaved under certain me-

264

FIGURE 18.1.

Cross sections of the frill (parietal + squamosals) of USNM 2100, Triceratops horridus, in (AB) transverse (taken just caudal to the end of the dorsal temporal fenestrae) and (B) mid-sagittal sections. (C) Right oblique view of the finite element mesh of a Triceratops frill used in this analysis, with load points and directions indicated by arrows. The numbers next to the arrows indicate the load case as discussed in the table and text. The grey unfilled circles bordering the rostral end of the frill indicate the point of constraint. Along the coordinate axes, X indicates dorsal, Y indicates mediolateral, and Z indicates caudal. Scale bar is 500 mm.

chanical loads. For instance, what would happen to the frill if

chanical response of a complexly shaped structure than does

the horn of an opponent struck? Were certain parts of the frill

beam theory (e.g., Weishampel 1993) or other simplifications;

better adapted to withstand these (hypothetical) forces, and

(2) FEM allows ‘‘experimentation’’ with extinct animals that

would the frill be more likely to fracture under certain condi-

cannot be analyzed with conventional laboratory techniques

tions? Thus, the structural behavior of the frill itself may help

such as strain gages; and (3) FEM offers relatively quick

illuminate its evolution and function.

and easy comparison of a variety of loading conditions and morphologies.

Methods

Within dinosaur paleontology, FEM has been used to study theropod and ornithopod cranial and postcranial adaptations

Finite element modeling (FEM) is an engineering method that

(e.g., Moreno et al. 2007; Rayfield et al. 2001; Rayfield 2005;

models the physical behavior of complexly shaped objects by

Snively and Russell 2002), but the present study is the first to

breaking them into a number of simpler objects (elements).

apply the method to ceratopsians. Finite element modeling

Equations calculating stress, strain, or other parameters of in-

was used to investigate the frill of Triceratops because the cera-

terest are applied to each of the elements that make up the

topsian frill is an irregularly shaped structure that does not

whole object. Solutions for the equations are iterated across

approximate a beam or other simple structure that could be

the entire model until the solutions converge, and the results

modeled with less complex engineering equations. Addition-

can then be viewed graphically or extracted for statistical anal-

ally, FEM allows a more thorough characterization of a struc-

ysis (Richmond et al. 2005; Rayfield 2007 present technical

ture than possible with other methods (Snively et al. 2006).

and practical overviews of FEM in vertebrate biomechanics

In the present study, a three dimensional digital model of

and evolution). Finite element modeling has received increas-

the frill was constructed for a Triceratops horridus skull, USNM

ingly broad application in vertebrate paleontology, because

2100 (National Museum of Natural History, Washington,

(1) FEM has the potential to more accurately describe the me-

D.C.). The original specimen, in which the frill is nearly com-

Modeling Structural Properties of the Frill of Triceratops 265

FIGURE 18.2.

Patterns of strain energy density in the frill of Triceratops horridus under various loading conditions. (AB) Load case (lc) 1; (B) lc 2; (C) lc 3; (D) lc 4; (E) lc 5; (F) lc 6; (G) lc 7; (H) lc 8; (I) lc 9; ( J) lc 10; (K) lc 11; (L) lc 12; (M) lc 13; (N) lc 14. Note that warm colors (reds, oranges, and yellows) indicate relatively high strain energy density, whereas cool colors (blues and greens) indicate relatively low strain energy density. Scale bar is 500 mm. Reproduced in color on the insert.

plete and relatively undistorted, was surficially scanned as

a point cloud starting with over twenty-million points. This

part of the ‘‘Virtual Triceratops Project’’ (Chapman et al. 1999,

was edited down to approximately one million points and

in press; Moltenbrey 2001). Approximately 90% of the skull

then polygonized to the basic model used in this study using

was scanned using a Steinbichler Optotechnik Optotrak opti-

Imageware’s Surfacer software. The resulting STL file was used

cal scanner. An NVision Modelmaker laser scanner mounted

to create a model in SolidWorks. Only the frill was considered

on a Faro articulating arm was used for those parts on the back

here, in order to allow more efficient modeling. The Solid-

of the frill that were unreachable by the optical scanner. The

Works model was then imported into finite element analysis

most difficult areas were captured using an Immersion Micro-

software, Algor FEMPRO 20.

scribe point digitizer. The virtual model was constructed using

266 farke, chapman, & andersen

The resulting finite element model (Fig. 18.1C) had 19,287

nodes and 48,036 brick elements, and nodes at the rostral margin were constrained from translation (but not rotation). Material properties were derived from averages for primate

Table 18.1. Percentile Values and Mean Values for von Mises Stress (in Units of MPa) within Frill Models for Various Loading Conditions

cranial bone (Wang et al. 2006), with an elastic modulus of

Quantile

14.55 GPa and Poisson’s ratio of 0.28. Isotropic, linear elastic properties were assumed, as no data on the anisotropy of cera-

Load case

50th

95th

topsian bone are available. Arbitrary point loads of 1,000 N

1

0.0427

0.1762

(equivalent to a 102 kg weight being placed on the frill) were

2

0.0687

0.3390

applied at various locations on the frill (Fig. 18.1C), respec-

3

0.0278

0.1112

tively parallel and perpendicular to the sagittal plane, in order

4

0.0547

0.1735

to determine the effect of various loads upon the structural

5

0.0449

0.3116

behavior of the frill. These loads could, for example, be inter-

6

0.0523

0.4744

preted to be analogs for blows from the horns of opponents

7

0.0281

0.1315

8

0.0432

0.2315

9

0.0061

0.0607

10

0.0146

0.1132

11

0.0145

0.2554

12

0.0110

0.1876

von Mises stress were extracted from the entire frill, with the

13

0.0085

0.1049

exception of the loaded nodes and the rostralmost portion of

14

0.0106

0.1408

(under hypothetical intraspecific combat). Graphical results and nodal values for strain energy density and von Mises stress were extracted for each model. In order to determine which loads the frill best resisted, nodal values for

the frill. Von Mises stress provides a measurement of overall stress values, incorporating both tensile and compressive stresses. Strain energy density is a measure of where energy is stored as an object deforms (Hibbeler 1997); thus, it provides

of load direction. For loads applied to the parietal lateral of the

an indication of what portions of the frill most effectively help

midline (load cases 1, 2, 7, and 8; Fig. 18.2A, B, G, and H),

the frill maintain mechanical integrity. In order to examine

strain energy density is concentrated on the side of the frill

overall deformation in the frill, parasagittal and coronal cross

ipsilateral to the load, with concentrations both in the pari-

sections were taken of the deformed model, with deformation

etal and within the arched medial portion of the squamosal.

exaggerated by an arbitrary amount for ease of visualization.

When loads are applied to the rostral end of the squamosal

For each loading condition, the median and 95% maximum

(load cases 11–14; Fig. 18.2K–N), strain energy density is al-

values were calculated from the von Mises stress values for

most exclusively concentrated in the squamosal itself. Under

each loading condition, using the statistical analysis package

loads to the distal end of the squamosal (load cases 5 and 6;

R. Ninety-five percent maximum values (not to be confused

Fig. 18.2E, F), some concentrations appear in the parietal, but

with 95% confidence intervals) were used instead of absolute

these do not exceed the area or magnitude seen in the squa-

maximum values, in order to avoid potential modeling arti-

mosal. Much of the concentration in the squamosal is along

facts caused by poor-quality elements.

its arched medial border. Under a medially directed load to the proximal squamosal

Results

(e.g., load case 11; Fig. 18.3C), most of the deformation of the frill occurs lateral to the point of application, and very

Bulk von Mises stress values for the frill are summarized in

little deformation occurs more medially on the frill. Under

Table 18.1. The highest absolute stresses (as measured by the

loads applied to the midline of the parietal, overall defor-

ninety-fifth percentile) are seen in loads applied to the dis-

mation of the frill, even when the load is distally applied, is

tal squamosal (load cases 5 and 6), ventrally to the distolat-

quite minimal compared to other loading conditions, and

eral portion of the parietal (load case 2), or medially to the

most deformation is in the dorsoventral plane (e.g., load case

proximo-lateral corner of the squamosal (load case 11). In

4; Fig. 18.3D). The strongly arched shape of the frill (both

terms of median stresses, the highest bulk stress values for the

rostrocaudally and mediolaterally) likely contributes to this

frill are seen in a ventrally directed load applied to the distal

effect, by virtue of its shape confining the deformation to

midline of the parietal (load case 4), or medially or ventrally

a relatively small area. Under loads applied more laterally

directed loads to the distal squamosal or disto-lateral portion

on the parietal or medially directed loads to the distal squa-

of the parietal (load cases 1, 2, 5, and 6).

mosal (e.g., load cases 2 and 5; Fig. 18.3A, B), a significant

For loads applied to the midline of the parietal (load cases 3,

amount of deformation occurs along the entire ipsilateral

4, 9, and 10; Fig. 18.2C, D, I, and J), the highest concentrations

portion of the frill (more so than under any other loading

of strain energy are at the point of load application, regardless

condition).

Modeling Structural Properties of the Frill of Triceratops 267

FIGURE 18.3. Patterns of deformation in the frill of Triceratops horridus under various loading conditions. The undeformed cross section is shown in grey with a dotted outline; the deformed cross section is shown with a solid outline and no fill. Deformation is exaggerated, showing ‘‘relative’’ rather than ‘‘absolute’’ patterns. (AB) Load case (lc) 2, transverse cross section (taken just caudal to the end of the dorsal temporal fenestrae); (B) lc 5, transverse cross section (taken just caudal to the end of the dorsal temporal fenestrae); (C) lc 11, transverse cross section (taken just caudal to the end of the dorsal temporal fenestrae); (D) lc 4, mid-sagittal cross section. Scale bar is 500 mm.

Discussion

were generally absorbed within the squamosal, and similarly for the parietal. This has implications for an understanding of

Under some loading cases modeled here, the frill behaves sim-

frill morphology across ceratopsians. The apparently delicate

ilar to a cantilevered beam. Ventrally directed loads applied to

nature of the parietal in many taxa—such as Torosaurus or

the most distal portions of the frill produce the greatest me-

Chasmosaurus—is frequently cited as an argument against the

dian stress (especially load cases 2, 4, and 6; Table 18.1), which

use of the frill as a defensive structure (e.g., Dodson et al.

is expected because these loads are furthest from the fixed

2004). But, assuming that the squamosal had the potential for

rostral end of the frill. Conversely, loads closest to the base

more frequent impacts due to its peripheral location on the

of the frill (cases 9–14) have the lowest median stress. Based

frill, a strong squamosal would be more important than a

on this information, it is predicted that loads to these caudal

strong parietal. Following this line of reasoning, the thickened

points on the frill would be most likely to cause failure (frac-

medial margin of the squamosal in many taxa (e.g., Torosau-

ture) of the bone that might affect a large area or multiple

rus; Farke 2007) may be an adaptation to strengthen the frill.

elements. Interestingly, load case 11 has a relatively low bulk

Thus, the potential role of the frill as a defensive structure (in

median stress value (tenth out of 14) but a relatively high

addition to roles such as species recognition or jaw muscle

ninety-fifth percentile stress value (fourth out of 14). This

attachment) cannot be completely eliminated.

case, in which a medially directed force is applied to the

A second important, and often overlooked, characteristic of

proximo-lateral portion of the squamosal, suggests that the

the ceratopsid frill is the squamosal-parietal suture. In nearly

bulk of the stress is confined to the immediate loading point,

all specimens of ceratopsids, these sutures appear to remain

and the arched central portion of the blade of the squamosal

unfused (Farke pers. obs.). The strain allowed at such sutures

keeps the stress low in the adjacent parietal. Under this condi-

may have prevented forces on the squamosal from being

tion, failure (fracture) of the bone would be restricted to a very

transmitted into the relatively thin parietal, and vice versa. A

limited area of the squamosal.

second highly unusual characteristic of the parieto-squamosal

Concomitantly with the von Mises stress values, loads

suture is its relatively flat, smooth profile. This contrasts

placed more distally on the squamosal produce the greatest

greatly with the highly interdigitated sutures in the frontal

amount of deformation within the frill. Loads placed on the

region, for instance, and also indicates that the bones of the

proximal part of the squamosal or the midline of the squam-

frill may have been comparatively loosely connected to each

osal produce the least deformation. Based on this infor-

other. Further research may help to clarify these observations

mation, one would expect more fractures in the distal squa-

and their functional significance, if any.

mosal than in the proximal squamosal or midline parietal.

The benefits and limitations of FEM for this study must be

This could be investigated with analysis of paleopathologies

recognized. For instance, patterns of muscle activation can

amongst ceratopsid squamosals.

have important effects upon the patterns of stress and strain in

Interestingly, most strain energy density remained concen-

bone (Ross et al., 2005). No muscle forces were modeled for

trated within the bony element (parietal or squamosal) to

Triceratops, although there are clear muscle insertion scars on

which the load was applied. Loads applied to the squamosal

the ventral surface of the frill at the rostral end (for cervical

268 farke, chapman, & andersen

musculature) and on the dorsal surface of the frill within the

is far from complete, and FEM will be an important part of

supratemporal fenestrae (for adductor musculature). These

unraveling this puzzle.

muscle scars are confined to the rostral margin of the frill (and their relative areas can be constrained by the extent of neuro-

Acknowledgments

vascular impressions demarcating where tight-fitting skin cov-

We thank M. Carrano (USNM) for facilitating access to the

ered the bone more directly), and any muscle forces likely had a

digital Triceratops data, and C. Forster for helpful discussion.

minimal effect upon stress distribution for loads applied to the

Reviews by M. Ryan and E. Snively were helpful in improving

caudal end of the frill. Thus, the effects of muscle force proba-

the manuscript. This work was funded in part by a grant from

bly can be safely ignored under the load cases we examine.

the Jurassic Foundation and a National Science Foundation

A second limitation of this study concerns the assumption

Graduate Research Fellowship (both to AAF).

of uniform, isotropic material properties. Virtually all bone is anisotropic, and a casual examination of the orientations of

References Cited

Haversian systems and trabeculae in fossil specimens clearly

Chapman, R. E., A. F. Andersen, B. H. Breithaupt, and N. A. Matthews. In press. Technology and dinosaurs. In M. K. BrettSurman, J. O. Farlow, and T. H. Holtz, Jr., eds., The Complete Dinosaur, 2nd ed. Bloomington: Indiana University Press. Chapman, R. E., A. F. Andersen, and S. J. Jabo. 1999. Construction of the virtual Triceratops: Procedures, results, and potentials. Journal of Vertebrate Paleontology 19(3, Suppl.): 37A. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Farke, A. A. 2007. Cranial osteology and phylogenetic relationships of the chasmosaurine ceratopsid Torosaurus latus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 235–257. Bloomington: Indiana University Press. Forster, C. A. 1996. New information on the skull of Triceratops. Journal of Vertebrate Paleontology 16: 246–258. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. Monographs of the U.S. Geological Survey 49: 1–300. Hibbeler, R. C. 1997. Mechanics of Materials. Upper Saddle River: Prentice Hall, Inc. Lull, R. S. 1908. The cranial musculature and the origin of the frill in the ceratopsian dinosaurs. American Journal of Science 25: 387–399. ———. 1933. A revision of the Ceratopsia or horned dinosaurs. Memoirs of the Peabody Museum of Natural History 3: 1–175. Moltenbrey, K. 2001. No bones about it. Computer Graphics World 24: 24–30. Moreno, K., M. T. Carrano, and R. Snyder. 2007. Morphological changes in pedal phalanges through ornithopod dinosaur evolution: A biomechanical approach. Journal of Morphology 268: 50–63. Rayfield, E. J. 2005. Aspects of comparative cranial mechanics in the theropod dinosaurs Coelophysis, Allosaurus and Tyrannosaurus. Zoological Journal of the Linnean Society 144: 309–316. ———. 2007. Finite element analysis and understanding the biomechanics and evolution of living and fossil organisms. Annual Review of Earth and Planetary Sciences 35: 541–576. Rayfield, E. J., D. B. Norman, C. C. Horner, J. R. Horner, P. M. Smith, J. J. Thomason, and P. Upchurch. 2001. Cranial design and function in a large theropod dinosaur. Nature 409: 1033– 1037.

indicates that this was likely the case for Triceratops too. Unfortunately, no presently developed methodology can correct for this. However, most experimental validations of FEM demonstrate that even simplified models with isotropic material properties do approximate real-world results, although even if specific details of the pattern may vary from reality (e.g., Strait et al. 2005). Thus, for the purposes of this study (examining overall patterns in a large cranial element), the assumption of isotropy probably does not bias our general results for the frill of Triceratops, or their implications for ceratopsid biology.

Conclusions The finite element models presented here indicate that the squamosal in the parietosquamosal frill of Triceratops could confine deformation to the squamosal itself, by virtue of its arched profile. This suggests that similar arched or thickened portions of the squamosal in other taxa may assist in the structural support of the frill. What, then, does this mean in ‘‘real life’’ conditions? The results are particularly relevant to the hypothesis that the frill functioned as a defensive structure, against the horns of opponents or the jaws of predators. Certainly, some of the structural features described here would be useful in this context. However, it must be emphasized that the results are only consistent with this function, not irrefutable support of it. Additionally, this study does not negate the potential role that the frill had as a display structure or platform for jaw muscle attachment. The future potential for FEM in understanding ceratopsian frills is quite strong. As stated above, the frill of Triceratops modeled here is relatively unusual among ceratopsids. An important next step is to model the frills of more ‘‘typical’’ ceratopsids, such as Chasmosaurus and Centrosaurus. In particular, it will be interesting to see how structures such as the thickened medial bar on the squamosals may have functioned within the frill. Furthermore, FEM may also prove illuminating in testing Tyson’s hypothesis that fenestrae developed in unstressed areas of the frill (Tyson 1977). Our understanding of the ceratopsian frill, in terms of its evolution and function,

Modeling Structural Properties of the Frill of Triceratops 269

Richmond, B. G., B. W. Wright, I. Grosse, P. C. Dechow, C. F. Ross, M. A. Spencer, and D. S. Strait. 2005. Finite element analysis in functional morphology. Anatomical Record 283A: 259–274. Ross, C. F., B. A. Patel, D. E. Slice, D. S. Strait, P. C. Dechow, B. G. Richmond, and M. A. Spencer. 2005. Modeling masticatory muscle force in finite element analysis: Sensitivity analysis using principal coordinates analysis. Anatomical Record 283A: 288–299. Snively, E., D. M. Henderson, and D. S. Phillips. 2006. Fused and vaulted nasals of tyrannosaurid dinosaurs: Implications for cranial strength and feeding mechanics. Acta Palaeontologica Polonica 51: 435–454. Snively, E., and A. P. Russell. 2002. The tyrannosaurid metatarsus: Bone strain and inferred ligament function. Senckenbergiana Lethaea 82: 35–42.

270 farke, chapman, & andersen

Strait, D. S., Q. Wang, P. C. Dechow, C. F. Ross, B. G. Richmond, M. A. Spencer, and B. A. Patel. 2005. Modeling elastic properties in finite-element analysis: How much precision is needed to produce an accurate model? Anatomical Record 283A: 275– 287. Tyson, H. 1977. Functional craniology of the Ceratopsia (Reptilia: Ornithischia). M.Sc. thesis, University of Alberta, Calgary. Wang, Q., D. S. Strait, and P. C. Dechow. 2006. A comparison of cortical elastic properties in the craniofacial skeletons of three primate species and its relevance to the study of human evolution. Journal of Human Evolution 51: 375–382. Weishampel, D. B. 1993. Beams and machines: Modeling approaches to the analysis of skull form and function. In J. Hanken and B. K. Hall, eds., The Skull. Vol. 3: Functional and Evolutionary Mechanisms, pp. 303–344. Chicago: University of Chicago Press.

Plate 1. (Figure 2.2) Psittacosaurus sp. (LHPV2). Skull in (A) lateral-

view; (B) enlarged view of postorbital-jugal horn in lateral view; (C) enlarged view of maxillary crowns in lateral view. Scale bars are (A) 3 cm; (B) 2 cm; (C) 1 cm.

2 . (Figure 2.4) Psittacosaurus major (LH PV1). (A) Stereopairs of anterior palate in ventral view; (B) stereopairs of posterior palate and braincase in posterioventral view. See text for abbreviations. Scale bar is 3 cm.

P l at e

P l at e 3 . (Figure 2.5) Psittacosaurus major (LH PV1) stereopairs of occiput in posterioventral view. See text for abbreviations. Scale bar is 3 cm.

4 . (Figure 2.6) Psittacosaurus major (LH PV1). Maxillary crown in (A) lateral and (B) medial views; (C) dentary crown in medial view. Scale bar is 5 mm.

P l at e

P l at e 5. (Figure 2.18) Psittacosaurus mongoliensis (AMNH 6535) stereopairs of hatchling cranium in (A) right lateral view; (B) ventral view. Scale bar is 1 cm.

P l at e 6 . (Figure 2.21) Psittacosaurus mongoliensis (AMNH 6536) stereopairs of hatchling skull and anterior cervical vertebrae in (A) left lateral view; (B) right lateral view. Scale bar is 1 cm.

P l at e 7. (Figure 18.2) Patterns of strain energy density in the frill of Triceratops horridus under various loading conditions. (A) Load case (lc) 1; (B) lc 2; (C) lc 3; (D) lc 4; (E) lc 5; (F) lc 6; (G) lc 7; (H) lc 8; (I) lc 9; (J) lc 10; (K) lc 11; (L) lc 12; (M) lc 13; (N) lc 14. Note that warm colors (reds, oranges, and yellows) indicate relatively high strain energy density, whereas cool colors (blues and greens) indicate relatively low strain energy density. Scale bar is 500 mm.

8 . (Figure 8.4) Left lateral view of Diabloceratops eatoni holotype (UMNH VP 16699). (A) Diagrammatic representation of skull; (B) skull.

P l at e

19 New Evidence Regarding the Structure and Function of the Horns in Triceratops (Dinosauria: Ceratopsidae) J O H N W. H A P P

Introduction

analysis of a partial skull of Triceratops provides additional information concerning the structure and function of the taxon’s horns. Skull SUP 9713.0 (Shenandoah

Ever since the fossilized remains of Triceratops were first de-

University collections) was recovered from the Hell

scribed by Marsh (1889a), the structure and function of its

Creek Formation (Upper Cretaceous) near Jordan, Mon-

horns has been a focus of interest (Dodson et al. 2004). Al-

tana. The size and degree of co-ossification of skull ele-

though many horncores have been described, very few data

ments indicate that the specimen was an adult at time of

are available regarding the outer covering of the horns. When

death. A claystone layer, 7–33 mm thick, covers the

the type of Triceratops flabellatus (Yale Peabody Museum of

horns and does not appear elsewhere on the skull. It ap-

Natural History 1821) was first discovered by J. B. Hatcher in

pears to be a replacement of the horn sheath. Underlying

the Lance Formation (Marsh 1889b), a black powdery layer,

this claystone layer is an outer bone layer that has not

one-half inch thick, surrounded the base of the left horncore

been previously described in any specimen. It is 1.2–5.3

(Gilmore 1920). Hatcher et al. (1907) attributed the black layer

mm thick and is composed of compact Haversian bone.

to decomposed horny sheath material. Unfortunately, the

Embedded in this outer bone layer is a network of fer-

black layer was not preserved, so its composition remains un-

ruginous vascular casts that also have not been reported

known. In this report I describe new data regarding the outer

previously in this or any other dinosaur taxa or speci-

covering of a Triceratops horn that may relate to Hatcher’s orig-

mens. Primary casts range in diameter from 2.3 to 5.8

inal discovery.

mm. Branches from these primary casts have diameters

Researchers have long puzzled over the function of the

from 0.16 to 1.3 mm. Running through the supraorbital

horns in Triceratops and other ceratopsians. Ideas include:

horncore and through the clay in the cornual sinus at

(1) inter- and intraspecies recognition, (2) competition for

the base of the horn is a network of vascular traces. An-

mates and, possibly, territory, (3) defensive weaponry, and

other vascular trace connects the base of the supraorbital

(4) thermoregulation. For complete reviews see Farlow and

horn to the brain cavity. These new observations add

Dodson (1975), Barrick et al. (1998), Dodson et al. (2004) and

support for the hypothesis that the orbital horns in Tri-

Farke (2004). It is the possibility of a thermoregulatory func-

ceratops had a thermoregulatory function.

tion for the horns in Triceratops that is explored here.

271

FIGURE 19.1. SUP 9713.0 in the quarry site before removal. Skull is upside down and articulated except for supraorbital horns and one epijugal. Left supraorbital horn removed and not shown. Scale bar is 10 cm.

Modern terrestrial vertebrates have evolved protective mechanisms and behaviors to selectively cool the brain because their central nervous systems are very sensitive to elevated temperatures. Wheeler (1978) first pointed out that ceratopsians and other larger dinosaurs would have experienced thermal challenges similar to those of modern large vertebrates, and he proposed that dinosaurs possessed physiological mechanisms to protect the brain. Because the horns of Triceratops were highly vascularized, they were interpreted as thermoregulatory structures by Rigby (1989). Studies by Barrick et al. (1998) measuring oxygen isotopic composition of bone phosphate suggested that the horns of Triceratops were used as thermoregulatory structures in the stabilization of brain temperatures at extreme ambient temperatures. The discovery of neurovascular impressions lining the cornual sinus under the caudomedial portion of the Triceratops supraorbital horncores lends additional support to a thermoregulatory function (Farke 2006). Anatomical Abbreviations. cb: compact bone; cs: cornual

Geological Setting A partial skull of Triceratops sp. (SUP 9713.0) from the Shenandoah University Paleontology (SUP) Collection was retrieved from the Hell Creek Formation (upper Maastrichtian, Upper Cretaceous) near the town of Jordan, Garfield County, Montana (Sec2, T20N, R37E; 47\31%40&N, 106\58%29&W; 838 m above sea level [masl]). The nearest exposure of the CretaceousTertiary boundary, separating the Hell Creek Formation from the overlying Tullock Formation, occurs 384 m southeast of this locality at 845 masl. This places SUP 9713.0 in the uppermost part of the Hell Creek Formation. SUP 9713.0 was collected from fine-grained sediments that preserved fine details of the skull. Because Fastovsky (1987) interpreted similar sediments (siltstones) at nearby Brownie Butte as flood-plain deposits, it is likely that SUP 9713.0 was buried on a flood-plain during an alluvial flood.

Specimen Description

sinus matrix; nh: nasal horn; r: rostral; sh: right supraorbital

The skull is approximately 70% complete and was buried up-

horn; t: trabecular bone.

side down (Fig. 19.1). Both supraorbital horncores were de-

272 happ

tached during burial, but were preserved in vertical orienta-

The internal morphology of the skull was analyzed by com-

tion against the skull with their bases pointing up. Modern

puterized tomography (CT) using a CTi-GE scanner with a 3

erosion and weathering has exposed all but the tip of the left

mm thickness per scan. Densities were estimated from weights

supraorbital horn, and has removed all evidence of the origi-

of samples and their volumes that were measured by displace-

nal tissues and structures that covered the horncore. In con-

ment of water.

trast, the right supraorbital and nasal horns were not exposed prior to excavation and provide new and important information about some of the original structures and tissues associ-

Results

ated with these horncores. Other preserved cranial elements

SILTSTONE MATRIX LAYER

include: rostral, premaxilla, fragmented maxilla with teeth, isolated teeth scattered in the matrix surrounding the nares,

Surrounding the skull is a matrix of unstratified, medium to

nasal horn, nasal, left jugal, right squamosal, fragmented par-

light grey siltstone that weathers like ‘‘popcorn’’ when ex-

ietal, fragmented left squamosal, and braincase. No post-

posed at the surface. Orange, iron-oxide streaks as well as frag-

cranial elements were found.

mentary plant remains occur throughout the matrix. This

The non-striated bone surface texture, the degree of co-

layer is referred to here as the ‘‘siltstone matrix layer.’’

ossification, and the large relative size of skull elements indi-

The results of EDX chemical analysis (Fig. 19.2A) indicate

cate that the specimen was an adult at time of death (Forster

that the siltstone is rich in silica and aluminum; but poor in

1996; Sampson et al. 1997; Horner and Goodwin 2008). For

iron, manganese and calcium. The average density of the ma-

example, the nasal horncore measures 300 mm from the ven-

trix is 2.6 g/cm3, which falls within the normal range of 2.3–

tral surface of the nasal to apex of the horncore and represents

2.8 g/cm3 for illite and smectite clay minerals (Totten et al.

the largest of 41 Triceratops nasal horncores measured by the

2002). The properties of the matrix are consistent with Fastov-

author. The greatest length measured in a line from the base to

sky’s (1987) analysis of siltstone deposits 3.1 km to the west at

the apex of the disarticulated right supraorbital horn is 718

Brownie Butte in the Hell Creek Formation. There, the domi-

mm and the distance from the domed roof of the cornual

nant minerals are mixed-layer smectite/illite species with a

sinus to the apex of the same horn is 399 mm. There are also

minor amount of kaolinite.

healed bite marks on the skull attributed to Tyrannosaurus rex (Happ 2008).

SHEATH MATRIX LAYER

Methods of Analyses

When the siltstone matrix layer was removed during preparation, a black to dark grey mineralized layer was found cover-

Scanning electron microscopy was conducted on samples

ing the right supraorbital and nasal horns (Fig. 19.3) but not

from the horns of SUP 9713.0 and the matrix immediately

covering other areas of the skull that had remained buried.

surrounding the horns. Samples were coated with carbon

Whereas the siltstone matrix layer was reasonably easy to re-

and examined using a JEOL JSM-840, which provided sec-

move from skull material, this underlying layer was not. The

ondary electron and back-scattered electron images. Semi-

layer ranges in thickness from 7 to 33 mm, and is different in

quantitative analysis of these samples was carried out using a

appearance and composition compared to either the compact

Princeton Gamma Tech Analyzer using an energy dispersive

bone of the horncores or the siltstone matrix surrounding it

X-ray detector (EDX) coupled to the SEM. Analyses were made

and the remainder of the skull. The layer thins toward the

at 15 kV accelerating voltage and a probe current of 20 nA.

apex of either supraorbital or nasal horns and is absent at their

Count times were 120 s, and the data were corrected using the

tips. Because the layer appears where a keratinous sheath

Phi-Rho-Z program as supplied by Princeton Gamma Tech.

might be expected to have occurred in the living animal (For-

Analyses are reported as oxide weight percent. The relative

ster 1996), this layer is referred to here as the ‘‘sheath matrix

accuracy of the analyses is estimated to be 5–10 weight per-

layer.’’

cent. An analysis was performed as a standardless assay (i.e.,

EDX chemical analysis indicates that the sheath matrix

element peaks are compared against an internal single ele-

layer is composed of iron-rich aluminum silicates (Fig. 19.2B).

ment standard rather than a direct comparison with a known

The iron and manganese content contributes to its dark ap-

sample as a standard (Goldstein et al. 1992). Because it was

pearance. The elemental composition reflects the presence of

used as a coating and is a component of the stabilizing preser-

calcium carbonate. A much higher content of iron contributes

vative used in the field, carbon was excluded in the numerical

to the higher density of 2.9 g/cm3. The siltstone matrix layer is

analyses. Other lightweight elements such as oxygen and fluo-

easily separated from the sheath matrix layer beneath it. The

rine yield artificially high numerical percentages and are not

change in color and chemical composition between the silt-

reported.

stone and sheath matrix layers is sharp. Chemical analysis of

New Evidence Regarding the Structure and Function of the Horns in Triceratops 273

FIGURE 19.2.

EDX chemical analyses of bulk samples of SUP 9713.0 expressed as percentage by weight of oxide. Total Fe calculated as FeO. All analyses normalized to 100% (anhydrous). (A) Siltstone matrix layer; (B) sheath matrix layer surrounding nasal horn; (C) outer bone layer of right supraorbital horn.

274 happ

FIGURE 19.3.

Lateral view of nasal horn. (A) Horncore surface with sheath matrix and outer bone layers removed; (B) sheath matrix layer. Scale bar is 10 mm.

the siltstone matrix layer surrounding the sheath matrix layer

an elevated iron content (Fig. 19.2C). A calcium to phos-

and the skull shows a significantly lower iron, manganese,

phorous elemental weight ratio of approximately 2:1 is ob-

calcium, carbon, and oxygen content. The sheath matrix layer

served for the outer bone layer and is comparable to ratios

has a ‘‘popcorn’’ microstructure as seen in the SEM. The parti-

reported for modern bone (Schweitzer et al. 1997).

cles range in diameter from 20 to 30 mm, which is within the normal range of 3–50 mm for silt particles.

VASCULAR CASTS At the contact between the outer bone layer and compact

OUTER BONE LAYER

bone of the supraorbital and nasal horncores is a complex

Closely appressed to the underside of the sheath matrix layer

network of hollow ferruginous tubular structures running per-

and in contact with the horncore is a 1.2–5.3 mm thick layer

pendicular to the Haversian canals (Fig. 19.5). They have not

of compact Haversian bone that has not been described pre-

been described previously in Triceratops and are interpreted as

viously in ceratopsian horns (Figs. 19.4, 19.5A, C, D). Here,

vascular casts. Because some occupy grooves on the surface of

this layer is referred to as the ‘‘outer bone layer.’’ Its secondary

the inner horncore, they are unlikely to be postmortem plant

osteons have hollow central Haversian canals surrounded by

casts or burrow structures.

concentric lamellae (Fig. 19.4). The Haversian systems are lon-

On average there are 12 vascular casts per 100 mm of bone

gitudinal with external diameters of the encircling cement

surface traced mediolaterally. Primary casts range in diame-

lines averaging 0.42 mm for the nasal horn and 0.36 mm for

ter from 2.3 to 5.8 mm. Branches from these primary casts

the supraorbital horn. Their diameters are comparable with

vary in diameter from 0.16 to 1.3 mm. The branch shown in

the range of those observed for Triceratops rib (Enlow 1969)

Fig. 19.5C is less than 1 mm in maximum diameter. In cross

and other dinosaur bone (Chinsamy-Turan 2005). The high

section, each cast exhibits thick walls surrounding a sickle-

density (3.4 g/cm3) of the outer bone layer reflects the pres-

shaped opening that is partially obstructed by globular masses

ence of nearly pure iron that occurs as rounded microstruc-

(Fig. 19.5D). The larger vascular casts fill grooves interpreted

tures (0.1–0.2 mm in diameter) scattered over its surface. In

as vascular traces on the surface of the compact bone of the

addition to its high density, the dark brown color of this bone

horncores whereas smaller casts leave no trace on the horn-

indicates that permineralization or replacement of bone tissue

cores underneath the casts.

has taken place. EDX chemical analysis of the outer bone layer

EDX microelemental analysis of the vascular casts indi-

shows that its composition is mostly calcium phosphate with

cates that they are composed of goethite, a-FeO(OH). The

New Evidence Regarding the Structure and Function of the Horns in Triceratops 275

FIGURE 19.4. Photomicrographs of outer bone layer. Arrows point to secondary osteons in compact Haversian bone. (A) Near base of nasal horncore; (B) right supraorbital horncore. Scale bars are 0.2 mm.

high calcium content observed in the analysis may be due

CORNUAL SINUS MATRIX

to calcium ion migration from adjacent bone. The globular masses that occur inside the hollow structures are composed

The cornual sinus under the caudomedial portion of the su-

of siderite, FeCO3.

praorbital horncore is infilled with clay matrix. Additional vascular traces (Fig. 19.6D) run rostrocaudally through the INNER HORNCORE

interior of the cornual sinus matrix. The three trace structures shown in Fig. 19.6D are well away from the walls of the bony

The bone below the vascular casts is referred to as the ‘‘inner

horncore and continue through the matrix to its caudal end

horncore’’ in this report. Exterior surface grooves on inner

where they become visible with the unaided eye. There they

horncores of Triceratops have been described in detail (Hatcher

have hollow centers with outer ferruginous rims of approxi-

et al. 1907; Dodson 1996; Forster 1996). In general, larger vas-

mately 0.2 mm in diameter.

cular grooves on these surfaces run longitudinally from the

Three types of vascular structures lining the right cornual

base to the apex of the horn. Grooves in SUP 9713.0 vary in

sinus of SUP 9713.0 are observed: (1) solid casts running more

diameter from 1.2 to 5.4 mm and correspond to the diameters

or less rostrocaudally and varying in diameter from 1.0 to 4.3

of larger vascular grooves described in the references listed

mm, (2) unfilled grooves in the surface of the cornual sinus

above. CT images in Figs. 19.6B, C, and D (right supraorbital

matrix about 1 mm in diameter that are comparable though

horn) show an outer layer of compact bone, 15–37 mm thick,

smaller in size to those on the surface of the inner horncore,

encircling a core of trabecular bone. Starting near the apex

and (3) hollow casts with ferruginous rims with diameters be-

of the horn, vascular traces pass rostrocaudally through both

tween 1.1 and 2.1 mm running medially into the cornual

trabecular and compact bone. Proximal to the domed roof of

sinus matrix at 90\ to the solid casts in (1). Farke (2006) notes

the right cornual sinus, CT images show traces exiting trabec-

that vascular impressions lining the cornual sinus are com-

ular bone and proceeding rostrocaudally through compact

mon among chasmosaurines. In fact, the cornual sinus matrix

bone (Fig. 19.6C, D).

reported here is directly analogous to a natural cast of the

276 happ

FIGURE 19.5. Network of vascular casts of ferruginous mineralization at contact between outer bone layer and compact bone of right supraorbital horn. (A) Photomicrograph of ventral surface of outer bone layer; (B) drawing of the detail of vascular casts in (A); (C) photomicrograph of vascular casts. Hollow tubular structures exhibit fine detail of branching (arrow); (D) cross section of thick walls surrounding sickle-shaped opening partially obstructed by globular siderite. Scale bars are 10 mm (A, B); 1 mm (C, D).

New Evidence Regarding the Structure and Function of the Horns in Triceratops 277

FIGURE 19.6. Right supraorbital horncore of SUP 9713.0. (A) Lateral schematic of horn; (B–D) CT images in cross section. Darker areas represent higher radiodensity; (B, C) hollow ferruginous vascular traces passing through trabecular and outer compact bone (examples marked by wedges); (D) hollow ferruginous traces passing through cornual sinus matrix and compact bone (wedges). Dense ferruginous coating encircles outer wall of cornual sinus matrix and borders compact bone. Image quality limited by the CT scanner. Scale bars are 10 mm.

cornual sinus of a chasmosaurine (likely Triceratops) possess-

removed than the sheath matrix. (3) The sheath matrix layer

ing similar vascular surface impressions as described by Farke

differs in chemical composition from adjoining layers, espe-

(2006). Because of this precedence, the traces associated with

cially in proportions of iron, manganese, and calcium. (4) The

the cornual sinus matrix described here are also unlikely to be

layer differs in color and density from adjoining layers. (5) The

plant casts or burrow structures. In addition to vascular traces,

contact of the layer with the siltstone matrix surrounding the

a high density ferruginous layer was observed on the mar-

skull is sharp. Because there is no continuum of properties

gin of the cornual sinus and can be seen circling the matrix in

that gradually grade from the sheath matrix layer to the silt-

Fig. 19.6D.

stone matrix layer, there was minimal to no diffusion or mi-

Additional CT images that include the braincase reveal a

gration of ions between the two layers.

large ferruginous tubular trace extending through bone from

Based on these features, I propose that the two layers as

the base of the right supraorbital horn to the brain cavity

preserved in the fossil originated from different starting mate-

(Fig. 19.7).

rials. Whereas the siltstone matrix layer is interpreted as the result of floodplain sedimentation that buried SUP 9713.0, the sheath matrix layer is interpreted here as a clay replacement

Discussion

of the outer covering of the horns. Although SUP 9713.0 does

HORN STRUCTURE AND PRESERVATION

not preserve evidence of the original composition of the covering, it probably consisted of keratin (Forster 1996). Because

The sheath matrix layer exhibits properties that distinguish

the sheath matrix layer is absent from the apex of either supra-

it from the adjoining siltstone matrix layer. (1) The layer is

orbital or nasal horns, the length of the sheath beyond the

unique to the horns and does not appear elsewhere on the

horncores cannot be determined.

skull. (2) The siltstone matrix layer is softer and more easily

278 happ

Why is preservation of the outer bone layer rare in cera-

FIGURE 19.7. CT image of cross section of braincase of SUP 9713.0. Ferruginous trace (arrows) extends from brain cavity between foramen magnum and exit for olfactory nerves to base of right supraorbital horn (upper left). Darker areas represent higher radiodensity. Image quality limited by CT scanner. Scale bar is 10 mm.

topsians? The outer bone layer of SUP 9713.0 was easily re-

easily separated and lost before burial. In this context, SUP

moved from the inner horncore during preparation because

9713.0 may have been buried before the vascular tissue hold-

the dense network of vascular casts lie in a plane that forms an

ing the two bone layers together had completely decayed. Al-

interface between the two layers (Fig. 19.5A). I propose that

ternatively, permineralization of the vascular canals in SUP

this vascular interface significantly reduced the bone-to-bone

9713.0 after burial may have contributed to bonding the outer

contact area between the outer bone layer and inner horncore

bone layer to the inner horncore.

in most individuals, and that after death (but before burial),

The observation of ferruginous vascular traces running

rapid decay of the vascular tissue may have resulted in a fragile

through the interior of the cornual sinus matrix in SUP 9713.0

connection between the two bone layers in most ceratopsians.

(Fig. 19.6D) supports the hypothesis that the sinuses con-

Such a condition could allow for the outer bone layer to be

tained tissue rather than pneumatic structures (Sampson et al.

New Evidence Regarding the Structure and Function of the Horns in Triceratops 279

1997; Farke 2006). These sinuses lacked a bone structure to

a keratinous sheath that originally surrounded the bone.

serve as a mold for formation of vascular casts in SUP 9713.0.

A dense network of vascular casts is preserved in fine detail

Tissue structures in extant animals that are preserved by min-

and surrounds the inner horncore. The vascular casts are im-

eralization are much more resistant to degradation. However,

bedded in a previously undescribed outer bone layer. Major

the process by which soft vascular tissue in the cornual sinus

vascular traces at the base of the supraorbital horn can be

of SUP 9713.0 mineralized to form a mold for formation of

traced into the brain cavity. The presence of these tissues,

ferruginous casts is not understood (Schweitzer et al. 2005).

structures, and their arrangements, in combination with anal-

The following is a possible sequence of taphonomic events

ogous structures in goats, lends support to the hypothesis that

that led to formation of the sheath matrix layer in SUP 9713.0.

the orbital horns in Triceratops served a thermoregulatory

After the animal died, its skin rotted away, but the keratin on

function of selectively stabilizing brain temperature.

the horns was retained due to a slower rate of decomposition. During exposure the specimen was scavenged. During a subse-

Acknowledgments

quent flood event the skull was buried upside down with the

I thank J. R. Happ, G. Bennett, C. Morrow, J. Kelley, S. Snyder,

sheath of the horns intact. After burial the sheath completely

C. Gorman, R. Egli, and Shenandoah University students for

decayed leaving a space that was infilled by a clay matrix (the

their help in fieldwork and preparation. I appreciate the help

sheath matrix layer).

of H. Belkin, USGS at Reston, Virginia, for scanning electron microscope analysis. I thank M. Monroe and the staff at

THERMOREGULATORY FUNCTION

Winchester Medical Center for computerized tomography. I am grateful to John and Sylvia Trumbo and the Bureau of Land

Experiments on the thermal physiology of goat horns by Tay-

Management (Permit no. M-79223) for permitting access to

lor (1966) provide a reasonable basis for understanding the

the land and its fossils. Reviews by D. Eberth and A. Farke

horncore data of Triceratops. Like Triceratops, the Toggenburg

improved the content of this paper. Financial support for this

goat horns as well as those of other bovids are highly vascu-

research was provided by grants from the Ohrstrom Founda-

larized. Taylor observed dramatic increases in horncore tem-

tion, the Little River Foundation, and Shenandoah University.

perature when the central nervous system caused the blood vessels to swell with blood. The blood vessels dilated dur-

References Cited

ing heat stress, resulting in increased blood flow through the

Barrick, R. E., M. K. Stoskopf, J. D. Marcot, D. A. Russell, and W. J. Showers. 1998. The thermoregulatory functions of the Triceratops frill and horns: Heat flow measured with oxygen isotopes. Journal of Vertebrate Paleontology 18: 746–750. Chinsamy-Turan, A. 2005. The Microstructure of Dinosaur Bone: Deciphering Biology with Fine-Scale Techniques. Baltimore: Johns Hopkins University Press. Dodson, P. 1996. The Horned Dinosaurs. Princeton: Princeton University Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Enlow, D. H. 1969. The bones of reptiles. In A. C. Gans, ed., Biology of the Reptilia. Vol. 1: Morphology, pp. 45–80. New York: Academic Press. Farke, A. A. 2004. Horn use in Triceratops (Dinosauria: Ceratopsidae): Testing behavioral hypotheses using scale models. Palaeontologia Electronica 7: 1–10. ———. 2006. Morphology and ontogeny of the cornual sinuses in chasmosaurine dinosaurs (Ornithischia: Ceratopsidae). Journal of Paleontology 80: 780–785. Farlow, J. O., and P. Dodson. 1975. The behavioral significance of frill and horn morphology in ceratopsian dinosaurs. Evolution 29: 353–361. Fastovsky, D. 1987. Paleoenvironments of vertebrate-bearing strata during the Cretaceous-Paleogene transition, Eastern Montana and Western North Dakota. Palaios 2: 282–295.

horns. The blood was then cooled by convection and radiation from the surface of the horns. It then drained into the cavernous sinus at the base of the horn where it came into contact with a dense network of blood vessels called the ‘‘carotid rete.’’ The rete allowed a counter-current heat transfer to occur. There, the cooled venous blood returning from the goat horn cooled the arterial blood headed toward the brain. In the context of Taylor’s study, the presence of (1) a highly vascularized outer bone layer, (2) a vascular network on the exterior and interior of the cornual sinus matrix, and (3) a prominent vascular trace connecting the base of the supraorbital horn to the brain cavity in SUP 9713.0 (and supposedly all members of the Triceratops clade) provides additional analogous support for the hypothesis that the horns in Triceratops served the thermoregulatory function of selectively stabilizing brain temperatures. Finally, given that goats use their horns for multiple purposes, it is important to note that a brain-cooling function for the horns of Triceratops does not preclude other functions for these horns.

Conclusions SUP 9713.0 preserves several features not previously seen in ceratopsian horncores. A clay layer surrounding both nasal and supraorbital horncores appears to be a replacement of

280 happ

Forster, C. A. 1996. New information on the skull of Triceratops. Journal of Vertebrate Paleontology 16: 246–258. Gilmore, C. W. 1920. The horned dinosaurs. Annual Report Smithsonian Institution 1920: 381–397. Goldstein, J., D. E. Newbury, P. Echlin, D. C. Joy, A. D. Romig Jr., C. E. Lyman, C. Fiori, and E. Lifshin. 1992. Scanning Electron Microscopy and X-Ray Microanalysis, 2nd ed. New York: Plenum Press. Happ, J. W. 2008. Analysis of predator-prey behavior in a headto-head encounter between Tyrannosaurus rex and Triceratops. In P. Larson and K. Carpenter, eds., Tyrannosaurus rex: The Tyrant King, 344–368. Bloomington: Indiana University Press. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49. Horner, J. R., and M. B. Goodwin. 2008. Ontogeny of cranial epiossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Marsh, O. C. 1889a. Notice of new American Dinosauria. The American Journal of Science 37: 331–336. ———. 1889b. Notice of gigantic horned Dinosauria from the Cretaceous. The American Journal of Science 38: 173–175.

Rigby, J. K., Jr. 1989. Thermoregulation in latest dinosaurs. Journal of Vertebrate Paleontology 9(3, Suppl.): 36A. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Schweitzer, M. H., C. Johnson, T. G. Zocco, J. R. Horner, and J. R. Starkey. 1997. Preservation of biomolecules in cancellous bone of Tyrannosaurus rex. Journal of Vertebrate Paleontology 17: 349– 359. Schweitzer, M. H., J. L. Wittmeyer, J. R. Horner, and J. K. Toporski. 2005. Soft-tissue vessels and cellular preservation in Tyrannosaurus rex. Science 307: 1952–1955. Taylor, C. R. 1966. The vascularity and possible thermoregulatory function of the horns in goats. Physiological Zoology 39: 127– 139. Totten, M. W., M. A. Hanan, D. Knight, and J. Borges. 2002. Characteristics of mixed-layer smectite/illite density separates during burial diagenesis. American Mineralogist 87: 1571–1579. Wheeler, P. E. 1978. Elaborate CNS cooling structures in large dinosaurs. Nature 275: 441–443.

New Evidence Regarding the Structure and Function of the Horns in Triceratops 281

20 Evolutionary Interactions between Horn and Frill Morphology in Chasmosaurine Ceratopsians D AV I D A . K R A U S S , A N T O I N E P E Z O N , P E T E R N G U Y E N , ISSA SALAME, AND SHANTI B. RYWKIN

chasmosaurine dinosaurs exhibit a wide variation in

and horns seen in neoceratopsians, but do, however, show

cranial ornamentation. Their extravagant postorbital

significant modification of some skull elements (i.e., jugal,

horncores and parieto-squamosal frills were almost cer-

parietal, and squamosal) beyond what seems to be functional

tainly central to sexual displays and species recognition.

for the purely mechanical requirements of chewing or other

We hypothesized that males sparred with their horns

uses (Dodson et al. 2004). Expansion of the modified psit-

during sexual competition for mates and that the use of

tacosaur cranial structures led to the development of the more

the postorbital horns in such sparring matches set limits

elaborate postcranial frill seen in more derived basal ceratop-

on the morphology of parietal fenestrae in this group. Al-

sians, such as Leptoceratops (Russell 1970) and Protoceratops

though there were no statistically significant correla-

(Granger and Gregory 1923). The function(s) of these struc-

tions among the orientations of horns and frills and the

tures may have been important in mating behavior; for exam-

ranges of impact points on the chasmosaurine frills ex-

ple, the modified elements may have provided an advantage

amined, our analysis using two-dimensional animation

in head pushing competitions like those seen in some modern

did indicate that a close qualitative correlation between

reptiles, or been used as a display structure to attract a mate.

horn morphology, range of impact points on the frill,

Alternatively (or in addition), basal ceratopsians may have en-

and location and size of the parietal fenestrae could be

gaged in mating displays involving head bobbing like those

demonstrated.

seen in many modern reptiles and birds. If either hypothesis is correct, one could predict a scenario in which directional

Introduction Ceratopsian dinosaurs possess some of the most elaborate cra-

selection for these traits lead to the evolution of the elaborate cranial structures that reached a pinnacle with the largebodied ceratopsids.

nial ornamentation in the animal kingdom. Their frills and

In the early days of ceratopsian research Tait and Brown

long horns have presented an interesting evolutionary co-

(1928) suggested that the large, solid frill of Triceratops could

nundra for as long as the group has been known (Dodson

have protected its vulnerable neck from predatory bites from

1996; Dodson et al. 2004; Farlow and Dodson 1975; Molnar

Tyrannosaurus. There is now also ample evidence from healed

1977; Sampson et al. 1997), raising questions as to how and

injuries indicating that at least some ceratopsians sparred

why such ornaments evolved.

with rivals (assumed to be conspecifics) and parried horn

Basal ceratopsians such as Psittacosaurus lack the large frills

282

thrusts with their frills (Marshall and Barreto 2001; Tanke and

Farke 2007). However, frills with large parietal fenestrae would

brow horn contact is limited to the non-fenestrated portion of

probably have provided little protection against predatory

the frill. We also suggest that the mechanism of sexual selec-

attacks or have been useful in sparring matches given that

tion could have acted over time to shape the large and distinc-

the comprising bones (parietal + sqaumosals) would have had

tive cranial ornamentation of the ceratopsids involved in such

very little resistance to torsional stresses caused by a predatory

interactions.

bites (Erickson et al. 1996) or horn-on-frill contact (but see

Previous investigations of ceratopsids have used modern analogs such as the bovids (Farlow and Dodson 1974; Molnar

Farke et al. this volume). The discovery of additional large-bodied ceratopsids with

1977; Sampson 1997) to aid in the understanding of the evo-

thinner, highly fenestrated frill compared to Triceratops has

lution and use of these structures. We agree that the bovid

shifted the interpretation of the use of their horn and frills

model (Geist 1966; Lundrigan 1996; Caro et al. 2003) is a logi-

from a purely defensive function to one that incorporates sex-

cal one to use for future investigations and therefore apply it

ual selection (Dodson 1993; Farlow and Dodson 1975; Hopson

in this investigation. We have restricted our study to the chas-

1975; Padian et al. 2004; Sampson 1997). Ceratopsids other

mosaurine ceratopsians since the clade in dominated by taxa

than Diceratops and Triceratops have the frills with parietal

with long brow horns which we believe are a better analog to

fenestrae ranging from large (e.g., Chasmosaurus russelli, and

the horns of bovids than the dominant nasal horns of cen-

Agujaceratops) to small (Torosaurus). These openings would

trosaurines. Although the nasal horns of chasmosaurs may

have been covered with integument and filled with connec-

have been particularly useful in sparring matches, they would

tive tissue in life. They also would have been highly vulnerable

have posed minimal threat to the frill of a rival as their relative

to penetrating injuries from the horns of a rival leaving the

position would have made it nearly impossible for one to im-

animal susceptible to infection, and potential destruction of

pact on a rival’s frill. For this reason nasal ornamentation was

the frill altogether. It is arguable that a penetrating injury to

excluded from our analysis. It is also important to note that

the bone may have had a greater potential for lethality than an

recent discoveries of the basal ceratopsian Zuniceratops and

injury to soft tissue; however, the presence of numerous such

the new centrosaurine genera Albertaceratops (Ryan 2007) and

injuries in ceratopsian frills (Tanke and Farke 2007) that have

the Wahweap taxon (Kirkland and DeBlieux this volume)

healed indicate that ceratopsids had the ability to recover from

show that long postorbital horns are a synapomorphy for Cer-

such wounds.

atopsidae. However, as this feature is most widespread in chas-

Although sexual mechanism is now suggested to be the pri-

mosaurines we feel that this group presents better examples

mary driving force in the evolution of ceratopsian cranial

for investigating the relationship between postorbital horn-

ornamentation (Farke 2004; Sampson 1997; Sampson et al.

core and frill shape.

1997), questions remain as to how these structures were used and what forces shaped their development and diversity.

HYPOTHESIS

Institutional Abbreviations. AMNH: American Museum of Natural History, New York; TMP: Royal Tyrrell Museum of Pa-

A fundamental premise of our analysis is that chasmosaurs

laeontology, Drumheller.

would have sparred in a manner like that of some modern bovids (e.g., Cape Buffalo [Syncerus caffer], and most antelope,

Investigating Interspecific Horn and Frill Interactions in Chasmosaurinae

notably including the genera Alcelaphus, Gazella, and Oryx among others) in which the horns of opposing males are brought parallel to each other during confrontations (Molnar

In this study we investigate the hypothesis that postorbital

1977). The analyses were designed to test the relationship be-

horncore shape is linked to the size and position of parietal

tween horn morphology and the size and position of parietal

fenestrae in the frills of chasmosaurine ceratopsians. Samp-

fenestrae in the frill. Our null hypothesis (Ho) is that the di-

son (1997) proposed that some ceratopsids engaged in in-

mensions of the horns of each taxon and their likelihood of

traspecific behavior that involved locking their postorbital

penetrating the fenestrae during a potential sparring match is

horns together and engaging in shoving matches. These inter-

unrelated to the area of the frill taken up by the fenestrae.

actions between individuals could be related to acquiring food

The alternative hypothesis (H∞) is that the dimensions of the

resources or mates, among other reasons. Success in such

horns of each taxon and their likelihood of penetrating the

matches would be dependant, in part, on the size and shape of

fenestrae during sparring limited the area of the frill taken up

both the brow horns and frill. However, because the soft tissue

by the fenestrae. In our experiment Ho is rejected if the ani-

covering the fenestrae could be punctured, or otherwise dam-

mations demonstrate that the postorbital horns could reach

aged, by horn-on-frill contact we propose that brow horn

and potentially penetrate the integument covering the pari-

shape (a function of length and curvature) and parietal fe-

etal fenestrae.

nestrae size are linked such that, in ‘‘typical’’ interactions,

We can also predict that:

Evolutionary Interactions between Horn and Frill Morphology in Chasmosaurine Ceratopsians 283

1. The percentage of the frill taken up by the parietal

as this was considered to be the optimal position for spar-

fenestrae will be inversely proportional to the length

ring. Then the diagrams were brought together until the skulls

of the horn, and, following this,

came into direct contact with each other to represent the max-

2. If a chasmosaur lacks parietal fenestrae (e.g.,

imum possible reach of the horns. At each position the dia-

Triceratops), it will have the highest percentage of frill

grams were pivoted and rotated to show the range of impact

surface area that can be reached by a conspecific’s

points the horncores could reach on the frill of an opponent.

brow horns.

In our analyses both brow horncores impacted on an opposing animal’s frill simultaneously. If the animals rotated their

Methods

heads so that only one horncore at a time impacted a slight increase in range was seen, as Farke (2004) demonstrated. In

In order to understand the interplay of horns and frills during

order to account for horn sheaths that would have been pres-

intrasexual combat among conspecifics we made simple pro-

ent in life, the horns were extended by a conservative estimate

file line diagrams of selected taxa and animated them to deter-

of 10% the total horncore length. Therefore, assuming that we

mine the range of impact points possible during a fight in a

tested sheath-covered horncores, we refer to ‘‘horns’’ through-

manner similar to that of Farke’s (2004) three-dimensional

out the remainder of this paper.

computer models. Outline drawings were copied from technical illustrations of skulls in the literature (Farke 2004; Dodson

Results

1996; Dodson et al. 2004), traced from our own photographs of museum specimens and from photographs of specimens (Farke 2007) printed in the literature. The taxa included in our study were Agujaceratops, Anchiceratops, Chasmosaurus belli, C. russelli, Diceratops, Pentaceratops, Torosaurus, and Triceratops. In order to search for morphological correlations between horncore and frill orientation, proximately anchored lines were laid along the long axis of the frill and the axis of the skull parallel to the maxilla. The principle axis of the postorbital

STATISTICAL ANALYSES Our chi-square analysis showed that there was no significant difference in the replicates of our diagrams. Measurements of angles between methods were within a few degrees of each other and were similar to differences in replicates using the same method. No comparison generated a p-value less than 0.7 so we concluded that our methods were reliable and valid.

horncores is defined as the longest line that could be drawn entirely within the horncore. This process resulted in three intersecting lines for each specimen (five in Pentaceratops where there were two different horn orientations in a single specimen) representing a simple structural representation of the cranium/rostrum, postorbital horncores and frill. The angles

HYPOTHESIS TESTING There were no significant correlations among the orientations of horns and frills and the ranges of impact points on the frills, therefore Ho could not be rejected (Table 20.1).

at which these lines intersected were measured with a protractor. Measurements were made measuring the angle with the rostrum designated as 0\ and the tail as 180\. All of these mea-

Table 20.1. Brow Horn and Frill Orientations in Chasmosaurine Skulls

surements were performed five times (once per author) and a Frill/brow

chi-square analysis was conducted to compare the results of each replicate and ensure that our method was consistent. Line diagrams for each species were duplicated and flipped

Genus

Frill/skull

Brow horn/

horn

angle

skull angle

angle

180\ horizontally using Adobe Photoshop Elements 4.0 to use

Agujaceratops

71\ †3.4

78\ †1.8

34\ †3.7

in an animation of face-to-face intraspecific sparring. Multiple

Anchiceratops

167\ †6.4

122\ †7.1

39\ †4.3

diagrams for highly variable genera were used so that individ-

Arrhinoceratops

135\ †4.2

37\ †2.4

109\ †3

Chasmosaurus belli

140\ †4.1

60\ †6.2

74\ †8.5

Chasmosaurus russelli

132\ †4.2

83\ †3.9

65\ †4.6

Diceratops

130\ †3.5

79\ †2

57\ †3.2

Pentaceratops

137\ †2.7

45\ †8.1

86\ †5.3

Torosaurus

142\ †8.7

47\ †5.8

96\ †5.2

Triceratops

131\ †8.4

51\ †11.3

77\ †11.5

uals with differing morphologies could be compared. From an initial horizontal position of the skull axis, the diagrams were rotated until the horn axes of both animals were parallel to each other. The range of motion required to achieve this goal was actually smaller that that exhibited in modern bovids as recorded in our own photographic documentation of various sparring bovids, so specific anatomical constraints were not placed on the head movement of ceratopsians in our study. Once the postorbital horncores were parallel, the diagrams were brought together until the horns completely overlapped

284 krauss, pezon, nguyen, salame, & rywkin

Note: All angle measurements were made with respect to the frontal plane of the animal in which the rostrum is 0\ and the tail is 180\. The variation presented is one standard deviation encompassing multiple measurements of each specimen made by different observers. Measurements for Pentaceratops are an average of both horns. No correlations were found between any variables.

Discussion HYPOTHESIS TESTING The failure to reject our null hypothesis was not unexpected. Variation in the curvature of the horns and the distances between the horns and frills on various chasmosaurine skulls are critical confounding factors. Ultimately, these complications make it impossible to tease out simple mathematical relationships between particular structural elements and their functions. The lack of a mathematical relationship does not diminish from the evidence derived from our animation analysis or from our ability to make inferences from our findings.

ANIMATION ANALYSES Fig. 20.1 illustrates the maximum variation in the ranges of possible impact points on a chasmosaurine frill, comparing a relatively short-horned Chasmosaurus belli to a long-horned Triceratops. The results of our animations were clear. In each taxa we investigated it was found that the uppermost impact point possible for the horns of one animal on the frill

FIGURE 20.1. Dorsal view of the simplified skulls of Chasmosaurus (A) and Triceratops (B). The light grey area shows the possible range of impact for the brow horns of an opposing animal of the same taxon during sparring. Horns could also impact anywhere on the face, thus defending the eyes would have been of paramount importance. However, in this and all subsequent figures, we only show the range of impact on the frill because it is the area critical to our hypotheses.

of another during sparring was just below the lower boundary of the parietal fenestrae, confirming our first prediction— chasmosaurs with the shortest brow horns (many specimens

the significant development of postorbital resorption pitting

of Chasmosaurus) have the largest parietal fenestrae. The ex-

(M. J. Ryan pers. com. 2008) indicative of an advanced age.

ception to this pattern was Triceratops (that lacks parietal fe-

Based on the relatively small size of its parietal fenestrae we

nestrae) and confirmed our second prediction—that the horns

predict that when less mature specimens of this taxon are

of Triceratops could have reached nearly any point (]90%) on

found they will have unmodified postorbital horncores signif-

the frill of an opponent, except the very edge of the frill, dur-

icantly longer than either C. belli or C. russelli.

ing interaction (Fig. 20.1B).

Agujaceratops. In contrast to Chasmosaurus, the long horns

Chasmosaurus. Among Chasmosaurinae, Chasmosaurus taxa

of Agujaceratops mariscalensis (Fig. 20.3) impacted a much

with brow horns (C. russelli and some C. belli examined here)

larger area of a conspecific’s frill while sparring. The change

have the shortest postorbital horncores and large parietal

from a frill oriented at a nearly 180 angle to the main axis of

fenestrae (Fig. 20.2A, B). The area covered by the latter is ap-

the head to one with an approximately 90\ orientation to the

proximately equivalent to that seen or inferred for Agujacera-

long axis of the skull may also reflect a greater risk of horn

tops, but that taxon has broader squamosals, reducing the area

penetration of the frill during sparring. Although it may seem

of the frill taken up by the fenestrae. Some individuals of C.

counter intuitive, this orientation of the frill actually reduces

belli had longer horns (possibly males) than others, but all

the risk of impact to it from the horns of a rival when the head

specimens display relatively small horns compared to other

is oriented to bring the horns of two sparring animals parallel

species. While they would have posed a risk of significant in-

with each other. The fact that the horns are slightly recurved

jury to the head of an opponent, our analysis shows that the

and directed upwards at an approximately 79\ angle to the

entire risk would have been to the facial and cranial regions of

plane of the skull meant that in order to spar with the horns

the skull, and that only the squamosal portion of the frill

parallel Agujaceratops would have had to dip it’s head signifi-

could have been impacted by the horns. It seems possible that

cantly downwards increasing the stress on the neck muscles

the extended lateral flaring of the lower portion of the frill

during sparring and thus reducing the force one animal could

may have been an adaptation to parry thrusts from postorbi-

exert on another. Granting our premise that horns would

tal horns of a rival that might otherwise have impacted the

have been brought together in approximately parallel and

vulnerable neck area.

horizontal orientations during sparring, Agujaceratops would

The only known specimens of Chasmosaurus irvinensis (not

then have had to engaged in a large degree of ventroflexion of

examined here) lack postorbital horncores and have rela-

the neck to created this situation. Such ventroflexion would

tively small parietal fenestrae (Holmes et al. 2001). The lack

have resulted in maximal stretching of the dorsal muscles of

of postorbital horncores is probably due to their loss through

the neck thus limiting the force they could generate.

Evolutionary Interactions between Horn and Frill Morphology in Chasmosaurine Ceratopsians 285

Pentaceratops. Pentaceratops poses an interesting dilemma. The best-known specimen has highly asymmetrical postorbital horns with each horn exhibiting a significantly different orientation. Although horned animals typically exhibit a limited amount of asymmetry in horn growth it is extremely rare for an animal to have such drastically different horn orientations. No modern horned or antlered animal exhibits this degree of asymmetry in non-pathologic specimens. We therefore feel that it is unlikely that the asymmetry of AMNH 6325 is normal variation. Whether the asymmetry exhibited in this specimen is pathologic or due to diagenetic effects is unknown, but either way the difference between the right and left postorbital horns of this animal provides a useful range of orientation into which Pentaceratops horns probably fell. We assumed that typical Pentaceratops individuals had approximately symmetrical postorbital horns and therefore analyzed Representative illustrations (from our animation analyses) showing position of skulls during sparring in two opposing individuals of the same size in (A) Chasmosaurus russelli and (B) Chasmosaurus belli. The brow horns would not have posed a risk of penetrating the frills, but would have been advantageous in a head-pushing contest. In this and all subsequent figures the light grey area represents the range of possible impact points of the brow horns of the white animal on the dark grey animal’s frill. FIGURE 20.2.

the animal twice, once using the right horn as a model (Fig. 20.4A) and once using the left horn as a model (Fig. 20.4B). In both cases we found that the postorbital horns would have been unlikely to impact the frill in the area of the fenestrae. They could have impacted the lower solid portion of the frill, but the risk of a penetrating wound was small. We found that the change in angle almost exactly offset the change in curvature so that both horns with significantly different shapes had nearly identical impact ranges. This particular analysis emphasized the importance of overall horn structure and underscored why no simple statistical correlations were found. Anchiceratops. Anchiceratops exhibits a postorbital horn orientation and size similar to that of the right horn of Pentaceratops, but has a much flatter frill (Fig. 20.5). The upright orientation of Pentaceratops frill (46\ from the axis of the skull) significantly reduced the range of impact points possible when the postorbital horns were held parallel for sparring. The frill of Anchiceratops was oriented at a 167\ angle to the axis of the skull opening a wider range of impact points. Anchiceratops has correspondingly smaller fenestrae that are also outside the range of impact points of the postorbital horns of a sparring partner. The high curvature of the frill in Arrhinoceratops places it effectively between the range of the two Pentaceratops horns (Fig. 20.4) and the risk of penetration is low. Diceratops and Torosaurus. Diceratops (Fig. 20.6A) and Torosaurus (Fig. 20.6B) both have upright, dorsorostrally recurved horns. Their orientation makes a wide range of impact points possible and we see reduced parietal fenestrae in both genera. Of the two taxa, Diceratops has proportionally longer horns and a more upright frill. The curvature of the horns in Toro-

Representative illustration of Agujaceratops sparring (from our animation analyses). Agujaceratops had significantly longer horns than Chasmosaurus, posing a significant risk of a penetrating injury to the frill of a rival. However, even with these long horns, the smaller and distally positioned opening would have been out of impact range. FIGURE 20.3.

286 krauss, pezon, nguyen, salame, & rywkin

saurus is also greater than that of Diceratops. This overall structure gives Diceratops a much greater range of frill impact points than Torosaurus and we see an extreme reduction in the size of the parietal fenestrae similar to that seen in Agujaceratops mariscalensis. Arrhinoceratops. Arrhinoceratops has strongly recurved horns.

FIGURE 20.5. Representative illustration of Anchiceratops sparring (from our animation analyses). The orientation of the horns and frill of Anchiceratops created a situation where a penetrating wound to the frill is highly unlikely. Even though the horns of this animal were rostrally oriented, the low orientation of the frill made its penetration improbable.

straight, anteriorly directed orientations. Animals with upwardly angled horns (horn-skull long axis angle approaching 90\) have a narrower range of frill impact points (Fig. 20.8A) than other Triceratops, but a range that is still extensive compared to other chasmosaurines. Triceratops specimens with more upward-facing horns also seem to have smaller horns than those with forward facing horns, although this is not statistically demonstrable. Animation analyses in which an animal with forward facing horns sparred with one with upward facing horns showed that the animal with forward facing horns had a much greater range of impact points than its Representative illustration of Pentaceratops sparring (from our animation analyses). The asymmetry in the skull of Pentaceratops that was analyzed here presents two different potential configurations for the horns (A and B). In each position, the combination of horn orientation with respect to the rest of the skull, and horn curvature limit the range of possible impact on the frill to just below the parietal fenestrae. Thus, risk of a penetrating injury would have been minimal. FIGURE 20.4.

opponent (Fig. 20.8B). Triceratops specimens with horns that are directed straight forward are capable of reaching nearly every point on the frill of an opponent (Fig. 20.8C). Our analysis indicates that almost the entire frill surface of Triceratops was available for a piercing injury from a conspecifics long brow horn. This may account, in part, for the thickness of the parietal and squamosals in this taxon. Triceratops appears to share with the geologically older Eotriceratops (Wu et al. 2007; not included in this analysis) the lack of parietal fenestrae, although the complete frill is not pre-

This curvature creates a forward directed horn axis that re-

served on the only known specimen (TMP 2002.57.7) of the

duces the degree of neck tilt necessary to bring horns parallel

latter taxon. The long postorbital horncores of Eotriceratops

during sparring (Fig. 20.7). The forward orientation of the

with their slight forward curve give them a potential impact

postorbital horns produces a range of horn impact points that

range of nearly the entire frill of a rival. Our model predicts

extends far up the frill. As a result we see a posteriorly directed

that that when a complete Eotriceratops frill is recovered it will

small parietal fenestrae.

lack fenestrae.

Triceratops. Triceratops has by far the most variable orienta-

The large number of Triceratops specimens known allow us

tion of postorbital horncores of the chasmosaurines, so we

to form hypotheses about the use and structure of horns by

examined all named species. The postorbital horns of Tricera-

different demographics of a hypothetical population. Allo-

tops range from upwardly directed and slightly recurved to

metric growth has also been documented in chasmosaurines

Evolutionary Interactions between Horn and Frill Morphology in Chasmosaurine Ceratopsians 287

FIGURE 20.7. Representative illustration of sparring in Arrhinoceratops (from our animation analyses). The curvature of the horns of Arrhinoceratops reduces the range of impact on the frill of an opposing individual of similar size. Although the parietal fenestrae are outside the range of impact, the squamosal fenestrae border the lateral area where impacts might occur.

entation would reduce the risk of serious injury in youthful sparring as animals learned the ‘‘rules of engagement’’ while they grew. Ontogenetic change in the postorbital horncores of Triceratops individuals resulted in a significant change that produced long, rostrally oriented horns in adults. This fits our model of the horns being used, in part, by reproductively mature individuals for intraspecific engagements over resources, including access to females. Representative illustrations of sparring in (A) Diceratops, and (B) Torosaurus (from our animation analyses). The combination of horn and frill orientations in chasmosaurines make penetrating wounds in the frill improbable. In Diceratops (A) the long vertical horns have an extended reach on the frill of an opposing individual, but the parietal fenestrae are highly reduced in size, and distally located, thus minimizing the possibility of their penetration. Torosaurus (B) shows a larger opening, but its shorter horns present less risk. Here, two different specimens of Torosaurus are shown sparring to demonstrate the relative reach of different horn configurations. FIGURE 20.6.

Adult Triceratops postorbital horn morphology generally falls into two broad categories—those animals with dorsally directed (forming a nearly 90\ angle with the axis of the skull) horns and others with rostrally oriented horns. It is certainly possible that this variation is due to specific differences within the genus, or individual variation as discussed above, but we propose that it may also be due to sexual dimorphism. It has long been hypothesized that ceratopsians used their horns in defense against predators. This is a plausible hypothesis that in no way negates any other hypotheses about the function(s) horns may have had. In many modern horned mammals both sexes have horns. In such species, exemplified by the musk ox (Ovibos moschatus) and numerous other bovids, the horns of females are smaller than those of males and can be used

including Triceratops (Carpenter 1984; Goodwin and Horner

against predators. In contrast, males additionally use their

2001; Goodwin et al. 2001; Lehman 2007; Tokaryk 1997) for

horns in competition for mates. We believe that this system

which we have the best growth sequence among chasmosau-

is the most appropriate model for ceratopsian horn use, es-

rines. Younger animals have small postorbital horncores that

pecially for Triceratops and other long-horned species. Males

are directed upwards, not forwards as in adults (Goodwin et al.

would be at an advantage in procuring mates with rostrally

2006; Horner and Goodwin 2008). This morphology and ori-

oriented horns while dorsally directed horns in females might

288 krauss, pezon, nguyen, salame, & rywkin

have been advantageous for defending offspring against large predators like tyrannosaurs.

Conclusion The extreme diversity of cranial ornamentation in the chasmosaurines defied our efforts at statistical analysis. However, we found that there is a consistent pattern in chasmosaurines in which the parietal fenestrae are positioned just outside the range of possible impact points from the brow horns of a rival, thereby limiting injuries to the soft tissue covering and filling the space bounded by the openings. Although longer horns increase the range of possible impact points on the frill, we found that rostrally oriented horns have a greater range of impact points than horns that project from the skull closer to a 90\ angle to the skull or those that are caudally curved. The results of our analyses suggest that the size and positioning of parietal fenestrae in chasmosaurines were largely determined by the potential range of impact points by brow horns on the frills. One test of this conclusion would be that if and when cranial integument is recovered for any of the chasmosaurs, the integument covering the portions of the skull most venerable to brow horn impact would have structures (e.g., expanded, thickened scales) to help offset possible damage. We concentrated our analysis on the chasmosaurines because of the predominance of long brow horns in this taxon. Centrosaurines were omitted from our study because the predominance of robust, elongate nasal horns in this taxon was unlikely to present a major risk of injury to the frill during sparring matches. A brief animation analysis of several genera confirmed this impression. At the time of our writing, however, several new centrosaurines (e.g., Ryan 2007; Currie et al. 2008; Kirkland and DeBlieux this volume; McDonald and Horner this volume) and chasmosaurines (Ott and Larson this volume; Loewen et al. this volume; Ryan et al. this volume) have been discovered that could have a bearing on our findings and will be included in an expanded future version of this analysis. We agree with the prevailing wisdom that the evolutionary forces driving the morphology of the frills of ceratopsians were sexual display and species recognition. We further suggest that the primary force determining the specific internal structure (excluding fringe ornamentation) of the

FIGURE 20.8. Representative illustration of Triceratops sparring with opposing individuals having the same morphology and size (from our animation analyses). The structure of Triceratops horns influences the range of impact on an opponent’s frill. Individuals with upright horns (A) have a narrower range of impact and would be at a disadvantage sparring with animals with rostrally oriented horns (B and C), which have a wider range of potential impact.

Evolutionary Interactions between Horn and Frill Morphology in Chasmosaurine Ceratopsians 289

frills in chasmosaurines is the structure of the postorbital

by providing greater leverage, or perhaps by creating pressure

horns and their use in sparring as part of intraspecific compe-

points for the origin of horns. Once such postcranial ridges

tition for mates.

had evolved they could have developed into display structures. If head-bobbing displays like those of modern iguanas

Comments on the Origin of Horns and Frills

were part of the mating rituals of primitive basal ceratopsians then these cranial ridges could have played an important role

It seems likely that head pushing contests like those seen in

in display as well as pure physical contests. Perhaps by simply

some modern lizards, notably the iguanids (Carpenter 1982;

making an animal look larger, or by displaying bright colors

Rand and Rand 1978) and chameleons (Van Mater 1971)

when excited or during mating seasons, they could have been

could form the model for ceratopsid intraspecific interactions

used to intimidate a sexual rival, or attract a mate. If any of

as suggested by Sampson (1997), may have been a plesiomor-

these hypotheses are correct, then, as with horns, a significant

phic behavior for all ceratopsids. If this were so then increas-

directional pressure towards the development of large, elabo-

ing large cranial ornamentation that would have offered an

rate frills would have occurred in the ceratopsians.

advantage in defeating an opponent in a head pushing match would also given an individual an advantage in procuring

Acknowledgments

mates. Irregularities on the skull such as those seen in liz-

We thank C. Mehling for his help with access to specimens

ards like the rhinoceros iguana (Cyclura cornuta) would have

at the American Museum of Natural History. We also thank

created pressure points when the heads of rival males were

the reviewers of this paper as well as the editors of this book

pushed against each other. There would have been significant

and the organizers of the 2007 Ceratopsian symposium at the

evolutionary pressure towards increasing the surface area

Royal Tyrrell Museum. Finally, we thank P. Dodson for his

and thickness of such points. If such points expanded and

input and helpful conversation at the onset of this project and

morphed to the point of becoming horns, an animal would

for his reviews on this paper, and M. Ryan for his very detailed

have a distinct advantage over a rival by being able to inflict a

review of our initial and subsequent drafts.

significant amount of discomfort, or injury, during a head pushing confrontation without suffering an equal effect from

References Cited

its rival. Such a situation could have created directional sexual

Burghardt, G. M. 1977. Of iguanas and dinosaurs: Social behavior and communication in neonate reptiles. American Zoologist 17: 177–190. Caro, T. M., C. M. Graham, C. J. Stoner, and M. M. Flores. 2003. Correlates of horn and antler shape in bovids and cervids. Behavioral Ecology and Sociobiology 55: 32–41. Carpenter, C. C. 1982. The aggressive displays of iguanine lizards. In G. M. Burghardt and A. S. Rand, eds., Iguanas of the World: Their Behavior, Ecology and Conservation, pp. 213–232. Norwich: William Andrew Inc. Carpenter, K. 1984. Baby dinosaurs from the Late Cretaceous Lance and Hell Creek Formations and a description of a new species of theropod. Contributions to Geology, University of Wyoming 20: 123–134. Currie, P. J., W. Langston, Jr., and D. H. Tanke. 2008. A new species of Pachyrhinosaurus (Dinosauria: Ceratopsidae) from the Upper Cretaceous of Alberta. In P. J. Currie, W. Langston, Jr. and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. 1–108. Ottawa: NRC Research Press. Dodson, P. 1993. Comparative craniology of the Ceratopsia. American Journal of Science 293A: 200–234. ———. 1996. The Horned Dinosaurs: A Natural History. Princeton: Princeton University Press. Dodson, P., C. A. Forster, and S. D. Sampson 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed. pp. 494–513. Berkeley: University of California Press. Erickson, G. M., S. D. Van Kirk, J. Su, M. E. Levenston, W. E.

selective forces towards animals with larger horns. This hypothesis is consistent with what we know about the evolution of the ceratopsians (Dodson et al. 2004; Molnar 1977), as well as the behaviors seen in modern horned animals (Geist 1966; Lundrigan 1996; Molnar 1977). Iguanas have been used as models for dinosaur behavior in the past (e.g., Burghardt 1977) and we feel that they may be appropriate models for ceratopsians in the context of the evolution of mating behavior. Before resorting to physical combat iguanas typically engage in head bobbing displays or flash brightly colored appendages (Carpenter 1982). Such behaviors are consistent with our model of the development of frills and horns in ceratopsians. If frills were brightly colored, then they may well have evolved as display structures similar to the dewlaps seen in many species of iguanids; however, this is purely conjecture. The evolution of frills in the ceratopsians is a more complicated issue and less easily resolved than the evolution of horns. The fossil record demonstrates that frills evolved before horns in basal ceratopsians. It does seem likely that frills originated through an expansion of the jaw musculature; however, this hypothesis cannot explain the subsequent extreme size increases of these structures. Sexual selection could also account for the origin and expansion of a postcranial frill. The ridges seen along the parietal in primitive species may have created an advantage in head pushing competitions, possibly

290 krauss, pezon, nguyen, salame, & rywkin

Caler, and D. R. Carter. 1996. Bite-force estimation for Tyrannosaurus rex from tooth-marked bones. Nature 382: 706–708. Farke, A. A. 2004. Horn use in Triceratops (Dinosauria: Ceratopsidae): Testing behavioral hypotheses using scale models. Palaeontologica Electronica 7: 1–10. ———. 2007. Cranial osteology and phylogenetic relationships of the chasmosaurine ceratopsid Torosaurus latus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 235–257. Bloomington: Indiana University Press. Farke, A. A., R. E. Chapman, and A. Andersen. 2010. Modeling structural properties of the frill of Triceratops. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 264–270. Bloomington: Indiana University Press. Farlow, J. O., and P. Dodson 1975. The behavioral significance of frill and horn morphology in ceratopsian dinosaurs. Evolution 29: 353–361. Forster, C. A. 1996. Species resolution in Triceratops: Cladistic and morphometric approaches. Journal of Vertebrate Paleontology 16: 259–270. Geist, V. 1966. The evolution of horn-like organs. Behavior 27: 175–214. Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian. 2006. The smallest known Triceratops skull: New observations on ceratopsid cranial anatomy and ontogeny. Journal of Vertebrate Paleontology 26: 103–112. Goodwin, M. B., and J. R. Horner. 2001. How Triceratops got its horns: New information from a growth series on cranial morphology and ontogeny. Journal of Vertebrate Paleontology 21(3, Suppl.): 56A. Goodwin, M. B., J. R. Horner, and W. A. Clemens. 1997. Morphological variation and ontogeny in the skull of Triceratops. Journal of Vertebrate Paleontology 17(3, Suppl.): 49A. Granger, W., and W. K. Gregory. 1923. Protoceratops andrewsi: A pre-ceratopsian dinosaur from Mongolia. American Museum Novitates 72: 1–9. Holmes, R. B., C. A. Forster, M. J. Ryan, and K. M. Shepherd. 2001. A new species of Chasmosaurus (Dinosauria: Ceratopsia) from the Dinosaur Park Formation of southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Hopson, J. A. 1975. The evolution of cranial display structures in hadrosaurian dinosaurs. Paleobiology 1: 21–43. Horner, J. R., and M. B. Goodwin. 2008. Ontogeny of cranial epiossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Kirkland, J. I., and D. D. DeBlieux. 2010. New basal centrosaurine ceratopsian skulls from the Wahweap Formation, Grand Staircase–Escalante National Monument, southern Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 117–140. Bloomington: Indiana University Press. Lehman, T. M. 2007. Growth and population age structure in the horned dinosaur Chasmosaurus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 259–317. Bloomington: Indiana University Press.

Loewen, M.A., S. D. Sampson, E. K. Lund, A. A. Farke, M. C. Aguillón-Martínez, C. A. de Leon, R. A. Rodríguez-de la Rosa, M. A. Getty, and D. A. Eberth. 2010. Horned dinosaurs (Ornithischia: Ceratopsidae) from the Upper Cretaceous (Campanian) Cerro del Pueblo Formation, Coahuila, Mexico. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 99–116. Bloomington: Indiana University Press. Lundrgian, B. 1996. Morphology of horns and fighting behavior in the family Bovidae. Journal of Mammalogy 77: 462–475. Marshall, C., and C. Barreto. 2001. A healed Torosaurus skull injury and the implications for developmental morphology of the ceratopsian frill. Journal of Morphology 48: 259. McDonald, A. T., and J. R. Horner. 2010. New material of ‘‘Styracosaurus’’ ovatus from the Two Medicine Formation of Montana. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 156–168. Bloomington: Indiana University Press. Molnar, R. E. 1977. Analogies in the evolution of combat and display structures in ornithopods and ungulates. Evolutionary Theory 3: 165–189. Ott, C. J., and P. L. Larson. 2010. A new, small ceratopsian dinosaur from the latest Cretaceous Hell Creek Formation, northwest South Dakota, United States: A preliminary description. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 203–218. Bloomington: Indiana University Press. Padian, K., J. R. Horner, and J. Dhaliwal 2004. Species recognition as the principal cause of bizarre structures in dinosaurs. Journal of Vertebrate Paleontology 24(3, Suppl.): 100A. Rand, A. S., and W. M. Rand. 1978. Display and dispute settlement in nesting iguanas. In N. Greenberg and P. D. MacLean, eds., Behavior and Neurology of Lizards: An Interdisciplinary Colloquium, pp. 245–251. Rockville: National Institute of Mental Health. Russell, D. 1970. A Skeletal reconstruction of Leptoceratops gracilis from the Upper Edmonton Formation (Cretaceous) of Alta. Canadian Journal of Earth Sciences 7: 181–184 Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., A. P. Russell, and S. Hartman. 2010. A new chasmosaurine ceratopsid from the Judith River Formation, Montana. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 181–188. Bloomington: Indiana University Press. Sampson, S. D. 1997. Dinosaur combat and courtship. In J. O. Farlow and M. K. Brett-Surman, eds., The Complete Dinosaur, pp 383–393. Bloomington: Indiana University Press. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnaean Society 121: 293–337.

Evolutionary Interactions between Horn and Frill Morphology in Chasmosaurine Ceratopsians 291

Tait, J., and B. Brown. 1928. How the Ceratopsia carried and used their head. Transactions of the Royal Society of Canada, Series 3, 22: 13–23. Tanke, D. H., and A. A. Farke. 2007. Bone resorption, bone lesions, and extracranial fenestra in ceratopsid dinosaurs: A preliminary assessment. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 319–347. Bloomington: Indiana University Press. Tokaryk, T. T. 1997. The first evidence of juvenile ceratopsians

292 krauss, pezon, nguyen, salame, & rywkin

(Reptilia: Ornithischia) from the Frenchman Formation (late Maastrichtian) of Saskatchewan. Canadian Journal of Earth Sciences 34: 1401–1404. Van Mater, J., Jr. 1971. The natural history of two generations of Chameleo jacksoni in captivity. Herpetology 5: 1–23. Wu, X.-C., D. B. Brinkman, D. A. Eberth, and D. R. Braman. 2007. A new ceratopsid dinosaur (Ornithischia) from the uppermost Horseshoe Canyon Formation (upper Maastrichtian), Alberta, Canada. Canadian Journal of Earth Sciences 44: 1243–1265.

21 Skull Shapes as Indicators of Niche Partitioning by Sympatric Chasmosaurine and Centrosaurine Dinosaurs DONALD M. HENDERSON

the fossil record of ceratopsid dinosaurs demon-

or where similarly sized species of Anolis lizards in Cuba dem-

strates that postcranially these animals were all quite

onstrate habitat partitioning with some species choosing

similar, that centrosaurines and chasmosaurines were of

more arboreal locations while others choose more terrestrial

roughly equal body size, and that frequently, several spe-

ones (Ruibal 1961). Alternatively, bodily attributes can differ

cies of these latter two taxa are found preserved in similar

between species, such as the differences in body size and denti-

stratigraphic positions. A biomechanical analysis of the

tions of sympatric carnivores, which are associated with differ-

skulls of ceratopsids using beam theory shows that there

ences in the size of prey taken; for example, felids in Israel

are distinct structural differences between centrosaurines

(Dayan et al. 1990), or the different body sizes and muzzle

and chasmosaurines. For cases of sympatric ceratopsids,

shapes in sympatric grazers; for example, antelope (Bovidae)

it was found that centrosaurines tended to have stronger

on the African savannahs (Estes 1991; Clutton-Brock and

skulls, and are inferred to have exerted higher bite forces

Harvey 1983). Similar sorts of physical character differences

than chasmosaurines. These differences are interpreted

have also been observed in the fossil record of vertebrates that

as reflecting different foraging behavior and/or choice of

have been interpreted as living sympatrically: for example,

foodstuffs. In cases where only centrosaurines or only

Miocene hyaenas (Werdelin 1996) and theropod dinosaurs of

chasmosaurines were coexisting, marked differences in

the Late Jurassic (Henderson 1998) and Late Cretaceous (Far-

skull strengths are still apparent. This suggests evolution-

low and Pianka 2003).

ary plasticity in all ceratopsians to reduce interspecific

The remains of species from the two principal clades of ceratopsian dinosaurs, chasmosaurines and centrosaurines (Dod-

competition.

son et al. 2004), are commonly found preserved together in

Introduction

the same strata, and often in great abundance; for example, Dinosaur Provincial Park (Ryan et al. 2001; Ryan and Evans

Studies of sympatric living carnivores and herbivores show

2005). Similar stratigraphic positions for sets of fossils imply

that animals differ from each other in distinct ways so as to

that populations of different species occupied the same habi-

avoid direct competition (Brown and Wilson 1956). These dif-

tat at the same time. A notable feature of ceratopsids is that

ferences can take a behavioral form such as the nocturnal ver-

postcranially they are all quite similar (Dodson et al. 2004),

sus diurnal feeding times of birds of prey such as owls (Strigi-

and most are of roughly equivalent body length and mass (4–

formes) and hawks (Falconiformes), respectively ( Jaksic 1985),

5 m and 1–2 t; Paul 1997). Exceptions include Triceratops with

293

a body length of 6.22 m and an estimated body mass of 3.9 t

duce some uncertainty into the finished result. However, the

(Henderson 1999). So many dinosaurs of similar size, mor-

use of published illustrations that have passed peer review

phology, and presumed diet living in sympatry raise the pos-

implies that the restorations are at least plausible (see below

sibility of interspecific competition for food resources.

for the estimates of uncertainty in the final results). The char-

If interspecific competition occurred between sympatric

acteristic shapes of the skulls were captured using a technique

ceratopsian dinosaurs then the fossil record has the potential

known as three-dimensional slicing of digitized contours of

to provide evidence for this. Given that ceratopsians are very

the skulls (Henderson 2002; Snively and Henderson 2006). In

conservative postcranially (Forster and Sereno 1997), the skull

addition to the basic shapes of the skulls, the horns, fossae, fe-

would be the primary place to look for distinguishing features

nestrae, epoccipitals, etc. were also digitized in order to make

that relate to possible feeding differences. Species differentia-

the species recognizable and facilitate interpretation of the

tion based on skulls is already well established for ceratopsians

results.

given their wide variety of horns and frill shapes (Dodson et

Although details of entire skulls were digitized and used to

al. 2004), but trophic differentiation has not been considered

produce the figures, only the fundamental shape of the ‘‘fa-

in detail. A skull can be viewed as a cantilevered beam that

cial’’ region of a skull from the quadrates forward to the snout

projects from the end of the neck (Henderson 2002; Snively

tip was analyzed for bending and torsion properties. The

and Henderson 2006), and two features of interest for beams

quadrate-articular jaw joint marks an important site where

are their resistances to bending and torsion (Biewener 1992;

jaw reaction forces impinge upon the skull, and the dorsal

Gordon 1978). The bending and torsional strengths of a skull

region of the skull between the orbits and the quadrate experi-

become important during feeding as bilateral biting produces

ences the forces of the jaw adductor muscles during biting.

reaction forces that act to bend the skull upwards, while uni-

The maxillary tooth row transmits bite forces into the max-

lateral chewing or biting induces twisting of the sides of the

illa, its surrounding bones, and the skull table, while the

face about the long axis of the skull. A skull that deformed

premaxillary-rostral region receives bite forces related to food

substantially during these activities would impair the ability

acquisition and transmits those forces to the dorsolateral

of an animal to orally process food as the forces that should be

nasal region. The facial models exclude any horns, crests or

detaching and breaking down foodstuffs are instead warping

openings, and treat the skulls as simple solid beams. This

and possibly weakening the skull.

viewing of the skull as a solid is justified in part by the fact that

Sympatric animals can reduce competition by not eating

all the animals being modeled are closely related and have the

the same foods, and differences in the mechanical nature of

same basic structure. An added complication when dealing

foodstuffs consumed by two different species can often be

with fossil skulls is that many of them become somewhat flat-

mirrored in the structure of their skulls. A prime example of

tened and warped during burial, and often have sections miss-

this has been documented over many years in the various

ing. Given the gross scale of the analysis presented here, and

species of finches that inhabit the Galapagos Islands (Grant

the built-in level of uncertainty in many of the restorations,

1999). More robust beaks belong to those birds that eat larger

any attempt to include variations in the thickness of skull

more robust seeds, while more gracile beaks are associated

walls would give a false and misleading sense of precision. The

with birds that eat small insects or small seeds (Grant 1999).

mandibles were excluded from the analysis as not all the cera-

This paper applies simple beam theory (Biewener 1992; Gor-

topsian skulls studied here are associated with mandibles.

don 1978) to a series of ceratopsian skulls, with the aim of

Fig. 21.2 presents an example of the slicing of the facial

detecting potential mechanical differences. These differences

region of Chasmosaurus irvinensis (after Godfrey and Holmes

will then be discussed as a possible result of natural selection

1995) that is typical of that used for all the taxa in this study.

for differences in diet in an attempt to reduce interspecific

The large black triangles highlight the locations of maxillary

competition among co-habiting ceratopsians.

and premaxillary-rostal biting in both dorsal and lateral views that were used to assess skull strengths. The central portion of

Methods

the figure shows the quadrilateral slices derived from the dimensions of the digitized left and right ventrolateral and dor-

Published restorations of ceratopsian skulls showing both lat-

solateral edges of the skull image, and 16 quadrilateral slices

eral and dorsal views were used as the primary data (Fig. 21.1),

were used for all skulls. The planar geometry of each quadri-

and the sources of these illustrations are listed in Table 21.1.

lateral is used in a series of computations that provide param-

Eight centrosaurines and ten chasmosaurines were selected,

eters characterizing its mechanical properties. Full details of

and Leptoceratops gracilis was used as a comparative outgroup

the skull slice calculations and strength estimation methods

taxon based on a recent cladogram (You and Dodson 2004).

can be found in Henderson (2002) and Snively and Henderson

Admittedly, some of the skull material used in the restorations

(2006), but a summary of the basic steps is presented here. The

is crushed and/or warped in some fashion, and this will intro-

first set of calculations determines the Y-axis centroid, Iy, of a

294 henderson

FIGURE 21.1. Cladogram (after Sampson and Loewen 2007) showing dorsal and lateral views of the three-dimensional skull models and phylogenetic relationships of the taxa included in this study. (A) Centrosaurines with the basal ceratopsian Leptoceratops gracilis as outgroup; (B) chasmosaurines. See Table 21.1 for sources used to generate the skull models.

slice. The distances to the top and bottom of the slice from this

vertical integration. z is the mediolateral distance from the

centroid, YT and YB, are then used to compute a slice’s vertical

horizontal center, and zL and zR are the lateral limits of hori-

second moment of area in the following expression: 1 2 I= yB4 – yT4 + bR yT3 – yB3 (1) 2mr 3mR y – yB , with zT and zB, yB where bR = yY – mRzR and mR = T zT – zB being the coordinate pairs for the top and bottom right hand

zontal integration, and are simple linear functions of the ver-

冉

冊

冉

冊

冉 冊

冉

冊

corners of the prism.

tical coordinate. zL is just the negative of zR, and this latter term is given by: ZR(y) =

y–bR mR

(3)

As the cross-sectional slices of the skulls are asymmetric

The horizontal distance between a given slice and a bite

about their horizontal and vertical centroids, the use of J in

force application position is treated as the lever arm of the

formulae for estimating stress will give incorrect results (Daeg-

applied force and used to calculate the bending moment.

ling 2002). However, the focus of the present study is not to

The resistance of a skull at the position of a slice to torsional

predict stress levels, but to give a general characterization of

forces was estimated by computing the polar moment of iner-

the potential for torsional resistance by a skull. The use of J as

tia (Gordon 1978) for the slice with the following expression: yT zR(y) 2 2 J = y=y (2) z=zL(y) z + y dzdy B

an estimator is reasonable as it takes into account how skull

Where y is the dorsoventral height from the Y-axis centroid,

The resistance of a structure to torsion is inversely propor-

and YT and YB are the top and bottom limits, respectively, of

tional to the distance of the applied force from the base of

冕

冕

再

冎

material is distributed across a slice and moment arms that are perpendicular to forces tangential to the surface of the skull.

Skull Shapes as Indicators of Niche Partitioning 295

Table 21.1. Taxa Examined in This Study, the Sources of the Skull Illustrations Used, and the Labels Used to Identify Taxa on the Graphs Taxon

Illustration source

Figure label

You and Dodson (2004)

L.g

Forster et al. (1993)

A.m

Basal Leptoceratops gracilis Chasmosaurine Agujaceratops mariscalensis Anchiceratops ornatus

Sampson and Loewen (2007)

A.o

Chasmosaurus belli

Godfrey and Holmes (1995)

C.b

Chasmosaurus irvinensis Chasmosaurus russelli

Holmes et al. (2001)

C.i

Godfrey and Holmes (1995)

C.r

Lull (1905)

D.h

Eotriceratops xerinsularis

Diceratops hatcheri

Sampson and Loewen (2007)

E.x

Pentaceratops sternbergi

Lehman (1998)

P.s

Torosaurus latus

Dodson et al. (2004) and Farke (2007)

T.l

Dodson et al. (2004)

T.h

Achelousaurus horneri

Sampson (1995)

A.h

Albertaceratops nesmoi

Sampson and Loewen (2007)

A.n

Dodson et al. (2004) and Lull (1933)

Cn.a

Sampson and Loewen (2007)

Cn.b

Triceratops horridus Centrosaurine

Centrosaurus apertus Centrosaurus brinkmani Einiosaurus procurvicornis Pachyrhinosaurus n.sp

Sampson (1995)

E.p

Currie et al. (2008)

P.ns

Styracosaurus albertensis

Ryan et al. (2007)

S.a

Wahweap new taxon A

Sampson and Loewen (2007)

W.nt

the structure (Gordon 1978). The overall torsional resistance

the same as shown in Fig. 21.1 and is included for visual ref-

of a skull was determined by summing the polar moments of

erence. A test slice was used to estimate the effects of 10%

each slice divided by the slice’s longitudinal distance from the

changes in the restored skull height on the computed me-

quadrate-articular joint.

chanical properties. This slice was of unit height and had dor-

In Fig. 21.2 those slices between the quadrate and the max-

sal and ventral widths that would make a prism shape similar

illary bite location are distinguished from the remaining slices

to that found with the actual digitized skulls. The tensile and

with a darker color, and the average strength of these slices

compressional strengths, and the polar moment of inertia,

was used for estimating the maxillary bite bending strength.

were computed for this slice (top left of Fig. 21.3B). The shape

The average of the strength indicators for slices between the

of the test slice was then systematically altered in six inde-

quadrate and the premaxillary-rostral bite point was used for

pendent ways, and new values computed for the mechanical

the anterior bite. As bone is weakest under tension (Currey

properties. These alterations included lowering and raising

1984), the defining strength parameter was chosen to be the

the top edge of the prism, narrowing and widening the top

tensile strength. During bilateral biting the dorsal surface of

edge, and narrowing and widening the bottom edge (center

the skull will be under compression, while the ventral edges

and right-hand side of Fig. 21.3B). The greatest source of error

will be under tension, so the ventral width of a slice was used

comes from overestimating the height, where a 10% increase

in the final calculation of bending resistance.

in height raises the values of computed properties by at least

Burial, compaction, and warping introduce uncertainty as

21%. 10% changes in the top width result in changes in com-

to the exact shape of many of the skull restorations used in

puted values of a similar magnitude. Changes in the basal

this analysis. An attempt was made to estimate the magnitude

width have the least effect, on the order of 5%.

of the uncertainty, and to see how it would affect the results and conclusions of this study. Fig. 21.3A shows three versions of the Centrosaurus apertus skull in lateral view. The top and

Results

bottom illustrations have the overall skull height increased

Fig. 21.4 graphically displays skull strengths calculated for the

and decreased, respectively, by 10%, while the transverse di-

posterior (maxillary, Fig. 21.4A) and anterior (premaxillary-

mensions were unchanged. The middle skull of Fig. 21.3A is

rostral, Fig. 21.4B) bite points with centrosaurines represented

296 henderson

FIGURE 21.2.

Example of the 3D slicing of the facial region of a ceratopsid skull (Chasmosaurus irvinensis after Holmes et al. 2001). The black triangles mark, in lateral and dorsal views, the positions of the two points of bite force application that were used to estimate skull resistances to forces associated with biting. The darker-colored posterior slices are those that were used for strength calculations associated with the maxillary bite point. The small figure on the left-rear panel is a grey-scale representation of all the facial slices stacked upon one another, with the lighter-colored slices more anterior than the darker ones. This subfigure is intended to give a sense of the shape of the skull if it were viewed from the front. Dimensions are in meters.

by squares, chasmosaurines by triangles, and the outgroup

ing on slice height, small changes in height give rise to large

taxon Leptoceratops gracilis marked with a diamond. The data

differences in strengths. Lastly, for a given skull length, cen-

used to construct these graphs are summarized in Table 21.2,

trosaurines have stronger posterior bite strengths than do

and three main features are apparent from the data. Overlain

chasmosaurines. This clade difference generally holds for an-

on the computed strength estimates of Fig. 21.4 (and subse-

terior bite strengths as well, but two exceptions are seen where

quent plots) are error bars of +/- 20%, and these assume the

the values associated with the chasmosaurines Diceratops

worst case effects of 10% over- and underestimates of skull

hatcheri and Triceratops horridus exceed that associated with

height based on Fig. 21.3B. Firstly, anterior bite strengths, and

the centrosaurine Albertaceratops nesmoi.

by implication anterior bite forces, are approximately one-

Three examples of the skull slicings used for strength de-

third to one-fifth those of the posterior strengths. Secondly, to

terminations are shown in Fig. 21.5, and these demonstrate

a good first approximation, the size of a skull determines its

how skull shape plays an important part in determining bite

resistance to bending—with larger skulls in both clades hav-

strength. These skulls are all from within Centrosaurinae

ing greater resistances. Length normalization of the estimated

(Sampson and Loewen 2007), and represent the extremes of

bending strengths of the skulls was not performed since body

bite strengths for all the ceratopsid skulls examined. Einiosau-

and skull size would be correlated with feeding ecology (Peters

rus procurvicornis (Fig. 21.5A) is the smallest in the sample and

1983). Removing size differences would remove ecological in-

has the lowest strength values. The other two skulls, Alberta-

formation. Within each clade there are differences, and these

ceratops (Fig. 21.5B) and Pachyrhinosaurus n.sp (Fig. 21.5C) rep-

arise from difference in cross-sectional shapes of the slices,

resent the largest examples of centrosaurines, and signifi-

especially their heights as the strength parameter is partly a

cantly the former is one of the earliest known members for the

function of the height of a slice raised to the fourth power (see

group. Although the Albertaceratops skull length is 91% that

the first equation in Methods). With this large exponent act-

of Pachyrhinosaurus, the posterior (maxillary) bite strength is

Skull Shapes as Indicators of Niche Partitioning 297

Effects of changes in skull height on computed strength parameters. (A) Visual demonstration of subtle differences in restored skull shapes with Centrosaurus apertus as an example. Top skull is 10% higher than the middle one. Bottom skull is 10% lower then the middle one. (B) Relative changes in computed mechanical properties of a typical skull slice when the linear dimensions are altered by 10%. Top left-hand image shows the original shape and its computed tensile strength (T) along the base, compression strength (C) along the top, and the polar moment of inertia ( J) as a proxy for torsional resistance. FIGURE 21.3.

more than twice that of Pachyrhinosaurus. This arises from a

posterior and anterior torsional resistances. Within the genus

relatively broader and deeper region of the skull between the

Chasmosaurus, C. irvinensis, C. belli, and C. russelli all show the

quadrates and the maxillary bite point in Albertaceratops, and

same relative rankings among each other for both torsional

highlights the importance of the transverse dimensions and

resistance measures, and these rankings are consistent with

the strongly non-linear nature of the skull strength calcula-

the pattern seen in their respective bending strengths. The

tions. In contrast, the rapid anterior reduction in width of the

consistency of rankings in all measures of resistances also

Albertaceratops skull reduces its anterior bite strength relative

holds for the two species of Centrosaurus (Table 21.3).

to that of Pachyrhinosaurus with its wider anterior facial region. The torsional resistances of the ceratopsian skulls are pre-

Discussion

sented graphically in Fig. 21.6 and the data are summarized in

The data tabulated above are useful for broad generalizations

Table 21.3. The same relative rankings of taxa with respect to

of the differences between chasmosaurines and centrosau-

strengths (resistances) seen for bending are seen in the tor-

rines, but the true value of the computed strength parameters

sional strength rankings. The observation that larger skulls

becomes apparent when comparisons are made between sym-

have greater bending strengths is also repeated with torsional

patric taxa. Comparing the inferred mechanical properties of

strengths. Diceratops exceeds all other chasmosaurines in both

the skulls of animals that lived side-by-side has the potential

298 henderson

FIGURE 21.4.

Plots of computed tensile skull resistances to (A) posterior (maxillary) and (B) anterior (premaxillary-rostral) biting. See Table 21.1 for the key to taxa labeling, and Table 21.2 for the data used to construct these graphs. The vertical bars show the +/–20% uncertainty estimated for each skull based on Fig. 21.3B.

Skull Shapes as Indicators of Niche Partitioning 299

Table 21.2. Skull Lengths and the Tensile Bending Strengths Computed for Maxillary and Premaxillary-Rostral Bite Points PremaxillaryTaxon

Skull

Maxillary bite

rostral bite

length (m)

strength (m2)

strength (m2)

Basal Leptoceratops gracilis

0.380

16.9

3.87

Agujaceratops mariscalensis

1.51

30.4

14.8

Anchiceratops ornatus

1.50

24.3

12.3

Chasmosaurus belli

1.72

34.7

12.6

Chasmosaurus irvinensis

1.72

44.6

17.1

Chasmosaurus russelli

1.40

15.7

Diceratops hatcheri

1.80

97.7

43.2

Eotriceratops xerinsularis

2.00

43.2

21.7

Pentaceratops sternbergi

2.26

49.3

27.7

Torosaurus latus

2.13

46.9

27.5

Triceratops horridus

2.03

Chasmosaurine

120

9.68

31.8

Centrosaurine Achelousaurus horneri

1.18

30.5

18.0

Albertaceratops nesmoi

1.83

1.61—103

31.9

Centrosaurus apertus

1.18

32.9

19.8

Centrosaurus brinkmani

1.25

38.6

24.6

Einiosaurus procurvicornis

0.837

10.3

Pachyrhinosaurus n.sp

2.00

638

6.40 119

Styracosaurus albertensis

1.46

30.6

21.7

Wahweap new taxon A

1.08

34.7

16.4

to reveal evidence of character displacements that relate to

the selected intervals. The first time slice where three taxa were

differences in trophic habits.

present is at 20 m with the addition of Styracosaurus, and the

Dinosaur Provincial Park (DPP) in Alberta, Canada, has pro-

centrosaurine taxa here are closely spaced with respect to max-

duced the remains of at least eight different kinds of cera-

illary strengths, but differ more in torsional strengths. The

topsians (five centrosaurines, three chasmosaurines), and

following time slice with three taxa at 30 m shows less dis-

most are known from multiple specimens (Ryan and Evans

parity in strengths with the replacement of C. russelli by C.

2005; Sampson and Loewen 2007). An added benefit of the

belli. This situation did not persist into the 40 m time slice

DPP ceratopsian sample is that precise stratigraphic data for

where the ceratopsid diversity has returned to the situation

the occurrences of the fossils are available (Currie and Russell

where there is just one chasmosaurine and one centrosaurine.

2005). Fig. 21.7A shows the known stratigraphic distribution

The interpretation here is that at the 20 m slice, with distinct

of ceratopsids through the Dinosaur Park Formation (DPF)

skull shapes and strengths, three species of ceratopsid could

within DPP, and their vertical positions are measured relative

comfortably coexist. With increased bite strength similar-

to the contact between the DPF and the underlying Oldman

ity among the taxa at 30 m, there was an unstable situation

Formation. Apart from a brief period from 20 to 30 m with

of trophic overlap that could not persist, leading to the lo-

three species, there are only two currently known species of

cal extinction of a centrosaurine. At the very top of the DPF,

ceratopsids present at any time. This suggests the possibility

the ‘‘standard’’ pattern of one centrosaurine and one chasmo-

that there is some form of ecological process operating that

saurine has resumed. It may just be coincidence, but the cen-

normally only allows for two sympatric species of ceratopsid at

trosaurine lost to local extinction between the 30 and 40 m

any time. Fig. 21.7B is a plot comparing the maxillary tensile

levels was C. apertus, which has a weaker skull with respect to

bending strengths of ceratopsids present at seven 10 m hori-

torsion than Styracosaurus. A stronger skulled animal could

zontal slices through the DPF in DPP. Fig. 21.7C is similar, but

cope with the same foodstuffs as a weaker skulled animal, but

shows maxillary torsional strengths instead. These plots show

the latter could not overlap with the former. The greater range

differences between the potential bite strengths of the taxa at

of food resistances that could be handled may have favored

300 henderson

Plots of computed torsional resistances to (A) maxillary and (B) premaxillary-rostral biting. See Table 21.1 for the key to taxa labeling. See Table 21.3 for the data used to construct these graphs. The vertical bars show the +/–20% uncertainty estimated for each skull based on Fig. 21.3B. FIGURE 21.6.

Illustrations of the facial slice geometries. (A) Einiosaurus procurvicornis (after Sampson 1995) shows a low strength value; (B) Albertaceratops nesmoi (after Sampson and Loewen 2007); and (C) Pachyrhinosaurus n. sp. (after Currie et al. 2008) show high strength values. The small size and narrow face of Einiosaurus contributes to its low strength values, whereas the combination of a large size and a broad skull leads to high strength values for Albertaceratops and Pachyrhinosaurus. See Table 21.2 for strength values. Dimensions in meters. FIGURE 21.5.

the survival of the stronger forms. A greater dietary range has been observed for Galapagos finches with larger and stronger beaks than for those with small beaks (Abbott et al. 1977). At 50 m, only the centrosaurine Styracosaurus albertensis is found, but this animal has a skull that was intermediate in strength between the preceding Chasmosaurus belli and subsequent Chasmosaurus irvinensis. Speculatively, Styracosaurus could be viewed as a generalist feeder that spanned a broader niche, which would have enabled it to fill an ecospace that

Skull Shapes as Indicators of Niche Partitioning 301

Table 21.3. Skull Lengths and the Torsional Resistances Computed for Maxillary and Premaxillary-Rostral Bite Points

Taxon

Maxillary torsional

Premaxillary-rostral

Skull

resistance

torsional resistance

length (m)

(m3—10-2)

(m3—10-2)

Basal Leptoceratops gracilis

0.380

4.80

0.850

Agujaceratops mariscalensis

1.51

11.27

4.33

Anchiceratops ornatus

1.50

7.22

3.05

Chasmosaurus belli

1.72

12.54

3.94

Chasmosaurus irvinensis

1.72

28.9

7.07

Chasmosaurus russelli

1.40

Diceratops hatcheri

1.80

78.5

Eotriceratops xerinsularis

2.00

21.7

Pentaceratops sternbergi

2.26

25.4

Torosaurus latus

2.13

15.0

Triceratops horridus

2.03

66.4

Chasmosaurine

6.55

2.71 24.1 7.25 10.4 7.75 16.2

Centrosaurine Achelousaurus horneri

1.18

Albertaceratops nesmoi

1.83

15.9

Centrosaurus apertus

1.18

14.4

6.80

Centrosaurus brinkmani

1.25

17.9

8.74

Einiosaurus procurvicornis

0.837

Pachyrhinosaurus n.sp

2.00

Styracosaurus albertensis

1.46

16.9

9.36

Wahweap new taxon A

1.08

13.9

5.17

896

2.82 333

6.95 17.46

1.18 48.6

normally would have been shared with a chasmosaurine. The

the latter. In this case it is not chasmosaurine/centrosaurine

replacing centrosaurine-chasmosaurine pair that appears at

ecological sharing, but two centrosaurines showing modifica-

60 m is C. irvinensis and Pachyrhinosaurus n. sp. (Ryan et al.

tions of their skulls. These differences, based on what is ob-

2006). Similar to the exit of Centrosaurus apertus, it is the

served in Dinosaur Provincial Park, are interpreted to possibly

weaker skulled Styracosaurus that is replaced by stronger

reflect dietary differences that mitigated direct competition.

skulled C. irvinensis and Pachyrhinosaurus n. sp.

The Lance Formation of the western United States has pro-

Unfortunately, the sequence of events described above is

duced three type of large chasmosaurine—Torosaurus latus, Tri-

based on the sampling of just one environment. Ideally, simi-

ceratops horridus, and Diceratops hatcheri. (Dodson et al. 2004).

lar sets of spatio-temporal data involving sympatric ceratop-

When maxillary bite strengths are compared for these three,

sians from other sites would be needed to support or refute the

Torosaurus (46.9 m2) and Diceratops (97.7m2) are distinct, but

inferred sequence of events in what was to become DPP. How-

Triceratops (120 m2) is just 19% stronger than Diceratops. How-

ever, three other ceratopsian-bearing formations show sim-

ever, when maxillary torsional strengths are compared, Tri-

ilarities to what has been observed in DPP, but unfortunately

ceratops (66 m3) is weaker than Diceratops (78.5 m3). These

they either lack demonstrable evidence of sympatry for their

contrasting torsional resistances make sense when the skulls

taxa, or detailed stratigraphic data are absent. The first of

are viewed from above (Fig. 21.1B), as Diceratops has an ex-

these is the Two Medicine Formation in Montana, which has

tremely wide posterior facial region. The possibility exists that

produced Achelousaurus horneri and Einiosaurus procurvicornis

the differences seen in the skull shapes of Diceratops and Tri-

(Sampson 1995). However, the specimens are stratigraphi-

ceratops could also be interpreted as species recognition traits,

cally separated by about 25 m (Sampson 1995), which argues

and not just the mechanics of biting different foods. Similar to

against sympatry, but the possibility exists that ancestors or

what was seen with character displacement among sympatric

descendents of these genera were overlapping in space and

centrosaurines in the Two Medicine Formation, here we have

time, leading to potential competition. In all measures of skull

a similar phenomenon among sympatric chasmosaurines. Ad-

bending strength and torsion resistance Achelosaurus and

ditionally, the wide gaps in computed strengths for these taxa

Einiosaurus differ, with the former at least twice as strong as

may have been the factor that allowed the three genera to

302 henderson

FIGURE 21.7. (A) Stratigraphic ranges of the three centrosaurine species (Centrosaurus apertus, Styracosaurus albertensis, and Pachyrhinosaurus n. sp.) and the three chasmosaurines species (Chasmosaurus russelli, C. belli, and C. irvinensis) that have been recovered from the Dinosaur Park Formation within the boundaries of Dinosaur Provincial Park. For most of the time represented by the Dinosaur Park Formation, there was just one type each of centrosaurine and chasmosaurine, suggesting that there is some sort of niche partitioning and/or competitive exclusion among sympatric ceratopsians. Stratigraphic data are from Currie and Russell (2005). (B) Maxillary tensile bending strengths of the taxa occurring at each of seven 10 m intervals in (A). (C) Torsional resistances of the taxa occurring at each of seven 10 m intervals in (A). The distinctive skull strength values for coexisting taxa suggest some sort of character displacement that mitigated interspecific competition for resources. The strength value for Pachyrhinosaurus exceeded the plot range used for all other taxa. The vertical bars about each symbol represent the estimated +/–20% uncertainty in the calculations assuming that skull restorations may differ by 10% in the vertical direction.

exist in sympatry, similar to what was seen for a short spell

natus. Arrhinoceratops, although not modeled for the present

in DPP.

study, appears to have skull proportions similar to that of Di-

The Horseshoe Canyon Formation of southern Alberta,

ceratops. This would place it at an intermediate strength posi-

Canada hosts the remains of Anchiceratops ornatus, Arrhino-

tion between Anchiceratops and Pachyrhinosaurus. The large

ceratops brachyops, and Pachyrhinosaurus canadensis (Dodson et

differences in the strengths computed, or predicted, for these

al. 2004). The latter centrosaurine has the strongest skull of

ceratopsians of the Horseshoe Canyon Formation would

any of the ceratopsids in this study, and here it shares its hab-

again allow for all three to coexist.

itat with a low bite strength chasmosaurine Anchiceratops or-

A relatively simple metric of skull shape and inferred bite

Skull Shapes as Indicators of Niche Partitioning 303

FIGURE 21.8.

Plot showing how the ratio of the span between the maxillary (MX) and premaxillary-rostral (PMX) bite points changes relative to facial length with increasing skull size. Centrosaurines show a strong trend to shorten their faces with increasing size (heavy dashed line), and this short face is correlated with the high anterior bite forces computed for this group. In contrast, chasmosaurines only show a very weak trend to elongate their faces with increasing skull size (dotted line), and show a concomitantly weaker anterior bite.

force was investigated by computing the ratio of the distance

members of this clade based on the three-dimensional analy-

between the two bite force application points (termed here as

ses of entire facial regions. In contrast, there is a weak trend

the span) and the length of the facial region for all the skulls.

for chasmosaurines to elongate their spans, which would

These ratios are plotted in Fig. 21.8 and the data are sum-

reduce anterior bite force. This relative elongation would

marized in Table 21.4. A decreasing relative span indicates a

partially counteract some of the potential increases in skull

more shortened face, and is correlated with an increasing bite

strength associated with an overall larger skull size. Appropri-

force. This correlation arises from the simple mechanical prin-

ately for a primitive outgroup, the ratio of span length to facial

ciple that relates the force supplied with an in-lever to the

length for Leptoceratops gracilis is intermediate between the

force applied with the associated out-lever (Hildebrand 1982).

extreme values shown for the derived centrosaurines and

The in-lever in the present situation is represented by the

chasmosaurines.

short distance between the jaw joint and the adductor muscle

The preceding analysis and interpretations of skull shapes

insertion point on the mandible just in front of the coronoid

has been done in terms of feeding ecology and character dis-

process, while the out-lever is represented by the longer dis-

placement. However, there does exist the equally reasonable

tance between the jaw joint and the premaxillary-rostral bite

possibility that any computed mechanical properties of a skull

point. This configuration describes a third-class lever, and the

based on its shape are merely selectively neutral effects associ-

longer the out-lever, the less force that can be applied at the

ated with skull changes driven by other factors. The variety of

front of the skull. With increasing skull size, the relative span

crests, frills, bumps and horns on the skulls of ceratopsians

in centrosaurines consistently gets shorter, indicating an in-

have been reasonably interpreted as a means of interspecific

creased potential to apply high force during premaxillary-

and intraspecific recognition (Sampson 1997). The raised

rostral biting, and this correlates well with the exceptionally

bosses on the snouts and above the eyes of Pachyrhinosaurus or

high anterior skull bending strengths computed for the largest

Achelousaurus, or the just above the eyes in Einiosaurus, and

304 henderson

Table 21.4. Distances between the Maxillary Tooth Row Bite Points and the Premaxillary-Rostral Bite Points (Diastema D) Relative to Facial Length (FL) Expressed as a Ratio (D/FL) Skull

Diastema

Face length

length (m)

(m)

(m)

D/FL

0.380

0.167

0.324

0.514

Anchiceratops ornatus

1.50

0.279

0.635

0.440

Agujaceratops mariscalensis

1.52

0.477

0.952

0.501

Chasmosaurus belli

1.72

0.399

0.782

0.509

Chasmosaurus irvinensis

1.72

0.507

0.895

0.566

Chasmosaurus russelli

1.40

0.293

0.612

0.479

Diceratops hatcheri

1.80

0.548

0.896

0.612

Eotriceratops xerinsularis

2.00

0.594

1.09

0.543

Pentaceratops sternbergi

2.26

0.469

0.975

0.481

Torosaurus latus

2.13

0.379

0.723

0.525

Triceratops horridus

2.03

0.491

1.03

0.477

Achelousaurus horneri

1.18

0.247

0.531

0.464

Albertaceratops nesmoi

1.83

0.369

1.02

0.363

Centrosaurus apertus

1.18

0.329

0.645

0.510

Centrosaurus brinkmani

1.25

0.353

0.822

0.429

Einiosaurus procurvicornis

0.837

0.182

0.382

0.477

Pachyrhinosaurus sp.

2.00

0.351

0.916

0.383

Styracosaurus albertensis

1.46

0.292

0.785

0.372

Wahweap new taxon A

1.08

0.304

0.683

0.445

Taxon Basal Leptoceratops gracilis Chasmosaurine

Centrosaurine

any increase in the breadth of a skull could have been associ-

mosaurines. These skull strength differences are interpreted

ated with head shoving matches between individuals, similar

to possibly represent examples of character displacement and

to that proposed for some hadrosaurs (Hopson 1975). Sexual

niche partitioning that reduced interspecific competition

selection might also have driven the development of differing

between morphologically similar animals. However, there

skull proportions (Sampson 1997), independent of any mas-

exists the possibility that the differences in skull strength pa-

ticatory needs. Despite these caveats, it is felt that feeding

rameters may be an unexpected side effect of changes in skull

ecology is a plausible interpretation, and future studies can

shapes driven by other factors such as sexual selection or the

either refute or support the current interpretations of the data

requirements for species recognition.

presented here. Acknowledgments

Conclusions An analysis of the bending and torsional strengths of a range of centrosaurine and chasmosaurine skulls demonstrates that there are distinct differences between the two clades. In gen-

The comments and suggestions for improvement from the two reviewers, Brenda J. Chinnery-Allgeier and Andrew A. Farke, made for a much-improved final version. References Cited

eral, centrosaurines have stronger posterior and anterior biting strengths than do chasmosaurines. However, considerable evolutionary plasticity is evident, as some chasmosaurines can approach or exceed centrosaurine skull shapes with respect to skull strengths. Comparisons of sympatric, or inferred to be sympatric, sets of ceratopsian dinosaurs show marked differences in skull strength parameters. This is observed for cases of coexisting centrosaurines and chasmosaurines, as well as for environments with only centrosaurines or only chas-

Abbott, I., L. K. Abbott, and P. R. Grant. 1977. Comparative ecology of Galapagos ground finches (Geospiza Gould ): Evaluation of the importance of floristic diversity and interspecific competition. Ecological Monographs 47: 151–184. Biewener, A. A. 1992. Overview of structural mechanics. In D. Rickwood and B. D. Hames, eds., Biomechanics—Structures and Systems, pp. 1–20. Oxford: Oxford University Press. Brown, W. L., and E. O. Wilson. 1956. Character displacement. Systematic Zoology 5: 49–64.

Skull Shapes as Indicators of Niche Partitioning 305

Clutton-Brock, T. H., and P. H. Harvey. 1983. The functional significance of variation in body size among mammals. In J. F. Eisenberg and D. G. Kleiman, eds., Advances in the Study of Mammalian Behaviour, pp. 632–663. Special Publication 7. Lawrence: American Society of Mammalogists. Currey, J. D. 1984. The Mechanical Adaptations of Bones. Princeton: Princeton University Press. Currie, P. J., W. Langston, Jr., and D. H. Tanke. 2008. A new species of Pachyrhinosaurus (Dinosauria: Ceratopsidae) from the Upper Cretaceous of Alberta. In P. J. Currie, W. Langston, Jr. and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. 1–108. Ottawa: NRC Research Press. Currie, P. J., and D. A. Russell. 2005. The geographic and stratigraphic distribution of articulated and associated dinosaur remains. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 537–569. Bloomington: Indiana University Press. Daegling, D. J. 2002. Estimation of torsional rigidity in primate long bones. Journal of Human Evolution 43: 229–239. Dayan, T., D. Simberloff, E. Tchernov, and Y. Yom-Tov. 1990. Feline canines: Community-wide character displacement among the small cats of Israel. American Naturalist 36: 39–60. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Estes, R. D. 1991. The Behaviour Guide to African Mammals. Berkeley: University of California Press. Farke, A. A. 2007. Cranial osteology and phylogenetic relationships of the chasmosaurine ceratopsid Torosaurus latus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 235–257. Bloomington: Indiana University Press. Farlow, J. O., and E. R. Pianka. 2003. Body size overlap, habitat partitioning and living space requirements of terrestrial vertebrate predators: Implications for the paleoecology of large theropod dinosaurs. Historical Biology 16: 21–40. Forster, C. A., and P. C. Sereno. 1997. Marginocephalians. In J. O. Farlow and M. K. Brett-Surman, eds., The Complete Dinosaur, pp. 317–329. Bloomington: Indiana University Press. Forster, C. A., and P. C. Sereno. T. W. Evans, and T. Rowe. 1993. A complete skull of Chasmosaurus mariscalensis (Dinosauria: Ceratopsidae) from the Aguja Formation (late Campanian) of west Texas. Journal of Vertebrate Paleontology 13: 161–170. Godfrey, S. J., and R. Holmes. 1995. Cranial morphology and systematics of Chasmosaurus (Dinosauria:Ceratopsidae) from the Upper Cretaceous of western Canada. Journal of Vertebrate Paleontology 15: 726–742. Gordon, J. E. 1978. Structures: Or Why Things Don’t Fall Down. New York: Plenum Publishing. Grant, P. R. 1999. Ecology and Evolution of Darwin’s Finches, 2nd ed. Princeton: Princeton University Press. Henderson, D. M. 1998. Skull and tooth morphology as indicators of niche partitioning in sympatric Morrison Formation theropods. Gaia 15: 219–226.

306 henderson

———. 1999. Estimating the masses and centers of mass of extinct animals by 3-D mathematical slicing. Paleobiology 25: 88–106. ———. 2002. The eyes have it: The sizes, shapes, and orientations of theropod orbits as indicators of skull strength and bite force. Journal of Vertebrate Paleontology 22: 766–778. Hildebrand, M. 1982. Analysis of Vertebrate Structure. 2nd ed. New York: John Wiley and Sons. Holmes, R. B., C. A. Forster, M. J. Ryan, and K. M. Shepherd. 2001. A new species of Chasmosaurus from the Dinosaur Park Formation of southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Hopson, J. A. 1975. The evolution of cranial display structures in hadrosaurian dinosaurs. Paleobiology 1: 21–43. Jaksic, F. M. 1985. Toward raptor community ecology: Behaviour bases of assemblage structure. Raptor Research 19: 107–112. Lehman, T. M. 1998. A gigantic skull and skeleton of the horned dinosaur Pentaceratops sternbergi from New Mexico. Journal of Paleontology 72: 894–906. Lull, R. S. 1905. Restoration of the horned dinosaur Diceratops. American Journal of Science, Series 3, 20: 420–422. ———. 1933. A revision of the Ceratopsia or horned dinosaurs. Memoirs of the Peabody Museum of Natural History 3: 1–175. Paul, G. S. 1997. Dinosaur models: The good, the bad and using them to estimate the mass of dinosaurs. In D. L. Wolberg, E. Stump, and G. D. Rosenberg, eds., Dinofest International: Proceedings of a Symposium held at Arizona State University, pp. 129– 154. Philadelphia: Academy of Natural Sciences. Peters, R. H. 1983. The ecological implications of body size. Cambridge: Cambridge University Press. Ruibal, R. 1961. Thermal relations of five species of tropical lizards. Evolution 15: 98–111. Ryan, M. J., D. B. Brinkman, D. A. Eberth, P. J. Currie, and D. H. Tanke. 2006. A new Pachyrhinosaurus-like ceratopsian from the upper Dinosaur Park Formation (Late Campanian) of southern Alberta, Journal of Vertebrate Paleontology 26(3, Suppl.): 117A. Ryan, M. J., and D. C. Evans. 2005. Late Cretaceous ceratopsids from the Judith River Group, Alberta, and the recognition of dinosaurian faunal zones within Dinosaur Provincial Park. In D. R. Braman, F. Therrien, E. B. Koppelhus, and W. Taylor, eds., Dinosaur Park Symposium: Short Papers, Abstracts, and Program, pp. 92–93. Drumheller: Royal Tyrrell Museum of Palaeontology. Ryan, M. J., R. Holmes., and A. P. Russell. 2007. A revision of the centrosaurine ceratopsid Styracosaurus (Dinosauria: Ornithischia) Journal of Vertebrate Paleontology 27: 944–962. Ryan, M. J., A. P. Russell, D. A. Eberth, and P. J. Currie. 2001. The taphonomy of a Centrosaurus bone bed from the Dinosaur Park Formation (Upper Campanian), Alberta, Canada, with comments on cranial ontogeny. Palaios 16: 482–506. Sampson, S. D. 1995. Two horned dinosaurs from the Upper Cretaceous Two Medicine Formation of Montana; with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15: 743–760. ———. 1997. Dinosaur combat and courtship. In J. O. Farlow and M. K. Brett-Surman, eds.,The Complete Dinosaur, pp. 383–393. Bloomington: Indiana University Press.

Sampson, S. D., and M. A. Loewen. 2007. New information on the diversity, stratigraphic distribution, biogeography, and evolution of ceratopsid dinosaurs. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 125–133. Drumheller: Royal Tyrrell Museum of Palaeontology. Snively, E., and D. M. Henderson. 2006. Fused and vaulted nasals of tyrannosaurid dinosaurs: Implications for cranial

strength and feeding mechanics. Acta Palaeontologica Polonica 51: 435–454. Werdelin, L. 1996. Community-wide character displacement in Miocene hyaenas. Lethaia 29: 97–106. You, H., and P. Dodson. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press.

Skull Shapes as Indicators of Niche Partitioning 307

22 The Function of Large Eyes in Protoceratops: A Nocturnal Ceratopsian? NICK LONGRICH

among herbivorous dinosaurs, Protoceratops andrewsi

aperture is consistent with this interpretation, as is the

is characterized by relatively large eyes. A study of the

presence of binocular vision. Nocturnality may have al-

size of sclerotic rings in extant birds shows that relative

lowed Protoceratops to exploit the hot desert environ-

eye size correlates with ecological habits: for a given body

ment of the Djadokhta while minimizing the need for

mass, herbivores, omnivores, and scavengers tend to

evaporative cooling. In this scenario, Protoceratops for-

have modest-sized eyes; diurnal, visually hunting preda-

aged at night when the air temperature was relatively

tors have much larger eyes; and nocturnal birds using vi-

low, and sheltered under vegetation or in burrows during

sion to forage have the largest eyes. Compared to birds,

the day to prevent overheating.

dinosaurs have small eyes relative to their body mass, and so direct comparisons with birds are not informative. However, the dinosaur data show an internally con-

Introduction

sistent pattern that is similar to birds: predatory

Few vertebrates, either fossil or extant, are as well-represented

theropods tend to have larger eyes than the plant-eating

in museum collections as is the ceratopsian Protoceratops. In

ornithischians and sauropodomorphs.

1922, the first fossils of Protoceratops andrewsi were recovered

In Protoceratops, however, the sclerotic ring diameter is

from the Upper Cretaceous outcrops of the Djadokhta Forma-

larger than predicted by the ornithischian regression,

tion, at Bayn Dzak (Flaming Cliffs). The next year, the Central

and the ring aperture is larger than predicted for either

Asiatic Expedition collected a dozen skeletons of Protoceratops,

the theropod or ornithischian regression. Whereas large

along with more than sixty skulls, jaws, and partial skeletons.

eyes in Protoceratops could have been an adaptation for

At the end of that fieldwork, in 1925, more than 100 speci-

predatory habits—a hypothesis that is consistent with

mens had been collected (Brown and Schlaikjer 1940). Since

the presence of binocular vision in that taxon—that hy-

then, many more fossils of Protoceratops have been recovered

pothesis is inconsistent with other observations, includ-

from Bayn Dzak (Kurzanov 1972), as well as exposures of the

ing the morphology of the jaws and dentition, tooth

Djadokhta at Tugriken Shireh ( Jerzykiewicz et al. 1993; Fas-

wear, and the relative abundance of Protoceratops in the

tovsky et al. 1997), Ukhaa Tolgod (Gao and Norell 2000), and

Djadokhta Formation. Accordingly, a more likely pos-

Bayan Mandahu, in Inner Mongolia (Dong and Currie 1993;

sibility is that large eye size in Protoceratops was an adap-

Jerzykiewicz et al. 1993; Lambert et al. 2001). This remarkable

tation to a nocturnal lifestyle. The presence of a large

assemblage is the basis of important insights into the sys-

308

FIGURE 22.1. Skull of Protoceratops andrewsi (AMNH 6466) and comparisons with Centrosaurus apertus. (A) A large ?female showing the sclerotic ring; (B) Centrosaurus apertus (AMNH 5351) after Brown (1917); (C) Protoceratops (AMNH 6466) after Brown and Schlaikjer (1940). Note that (B) and (C) are drawn to the same scale and, thus, show the relative size of the sclerotic ring in both ceratopsians. Despite its smaller size, the eye of Protoceratops (inferred from the size of the sclerotic ring) is roughly the same size as that of Centrosaurus. Scale bar is 1 m.

tematics and evolution of the Ceratopsia, dinosaur ontogeny,

argue that we still have much to learn about the biology of

and sexual dimorphism (Brown and Schlaikjer 1940; Kurza-

Protoceratops (Fig. 22.1).

nov 1972; Dodson 1976). In terms of morphology, the genus is now known about as well as it is possible to know an ex-

The Importance of Eye Size

tinct animal. But to what extent do we actually understand Protoceratops?

This study of eye size in Protoceratops grew out a survey of

The paleobiology of Protoceratops has been the subject of

relative eye size in non-avian Dinosauria. The original ques-

much speculation. Gregory and Mook (1925) suggested that

tion posed was: can relative eye size be used to infer aspects of

Protoceratops was aquatic, because of its large feet and deep tail.

a dinosaur’s paleoecology? There are good reasons for think-

However, the arid paleoenvironments of the Djadokhta For-

ing that eye size could be informative about the biology of

mation pose obvious problems for this hypothesis. Ostrom

extinct animals, because, assuming that the density of photo-

(1966) briefly discussed the functional morphology of the

receptors in the retina is constant, eye size has a strong in-

jaws and teeth of Protoceratops in his study of ceratopsian evo-

fluence on both the maximum resolution and sensitivity of

lution. Kurzanov (1972), Farlow and Dodson (1975), and Dod-

the eye.

son (1976) considered the function of the nasal ornamentation and frill, and what the existence of sexual dimorphism in the animal (Kurzanov 1972; Dodson 1976) might say about

RESOLUTION

its biology. More recently, a clutch of hatchlings found at

A large eye typically has improved resolution relative to a

Tugriken Shireh suggests that the animal was gregarious, at

small eye. As the eye increases in size, the image is projected

least at some point in its life (You and Dodson 2004).

over a larger area of the retina. So long as the aperture is in-

Yet many questions remain unanswered. How did the jaws

creased by the same factor, image brightness remains con-

and teeth function, and what did Protoceratops eat? What was

stant. The result of large eye size under these conditions is that

the function of the frill? How did the animal move? How

each photosensitive cell gets the same number of photons per

could this dinosaur survive in the deserts of ancient Mongolia,

second, but the image is now projected over a larger number

which are unlikely to have been more hospitable in the Cre-

of photoreceptors (Fig 22.2A, B). If each of the photoreceptors

taceous than they are today? This paper will not provide

acts as a photoreceptive unit, this increases resolution (Ross et

hard answers to these questions; but by addressing another

al. 2007), as is generally the case in birds (Hall and Ross 2007).

question—why did Protoceratops have such large eyes?—I will

The effect is like upgrading from a 35 mm film camera to a

The Function of Large Eyes in Protoceratops 309

FIGURE 22.2.

The advantage of large eyes. In a small eye (A), the image formed by the lens is projected onto relatively few photoreceptors. If the eye is scaled isometrically, the image is projected over more photoreceptors, allowing for increased resolution (B). Although the light is distributed over a larger area, the larger pupil increases the rate of light capture, keeping image brightness constant. However, the output of several photoreceptors can also be summed to create larger photoreceptive units (C), increasing the amount of light striking each photoreceptive unit. In this example, sensitivity is improved by summing photoreceptor output. The increase in the number of cells in the retina, however, means that the number of photoreceptor units is held constant, and visual acuity is not sacrificed.

4& — 5& large format camera: the larger film has more photo-

EVOLUTIONARY SIGNIFICANCE

sensitive silver grains, and therefore produces a higher resolution image. Increased resolution could be useful for detecting

It follows that larger eyes generally result in higher visual reso-

camouflaged prey, or for resolving objects from long distances.

lution, increased visual sensitivity, or some combination of the two. However, large eyes also come with costs. These in-

SENSITIVITY

clude the materials and energy needed to build and sustain a larger eye, and the costs of building and maintaining associ-

As the retina becomes larger, sensitivity increases without de-

ated tissues such as orbital bones, optic nerves, optic lobes,

grading image quality. This is accomplished by pooling the

ocular muscles, blood vessels, and eyelids. Larger eyes may

output of several photoreceptive cells such that they function

also require a larger orbit, which weakens the skull (Hender-

as a single, large photoreceptor unit (Fig. 22.2C; Warrant 1999;

son 2002). Therefore, large eyes should evolve only when the

Stevens 2006; Ross et al. 2007). The larger area covered by the

fitness-benefits outweigh the costs, when natural selection

photoreceptor units means it intercepts more photons per sec-

has favored the evolution of an eye with high acuity, high

ond, allowing the retina to form an image in environments

sensitivity, or both. These observations, along with casual ob-

with lower ambient light levels. Again, assuming that photore-

servation of bird skeletons (Fig. 22.3) argue that the relative

ceptor density is more or less constant, the larger eye will have

size of an animal’s eyes should provide useful information

a larger number of photoreceptor cells, allowing for receptor

about its ecology.

pooling while maintaining a large number of photoreceptive

This idea is corroborated by studies showing that relative

units, and therefore maintaining visual acuity (Fig. 22.2C).

eye size and ecology are correlated in extant vertebrates. Noc-

Sensitivity is also thought to be useful for increasing the rate at

turnal birds tend to have larger eyes than diurnal birds (Hall

which the image is formed (Warrant 1999), and therefore sen-

and Ross 2007; Ross et al. 2007), and a similar pattern is pres-

sitivity may be important for forming images when the ob-

ent in primates (Ross and Kirk 2007; Ross et al. 2007); preda-

server is moving rapidly (e.g., running or flying) or when the

tory birds also tend to have relatively larger eyes than her-

object of interest is moving rapidly (e.g., fleeing prey).

bivorous birds (Murray et al. 2003).

310 longrich

FIGURE 22.3. Sclerotic rings in several birds with comparable body masses (1.0 kg). From left to right: pheasant (Phasianus colchicus), an herbivore; raven (Corvus corax), a scavenger; red-tailed hawk (Buteo jamaicensis), a predator; owl (Strix), a nocturnal predator. The dramatic differences in the absolute size of the eye suggest that relative eye size can be used to predict ecology.

Given the correlation between relative eye size and habits in

Strigiformes, Caprimulgiformes; pers. obs.), but in others, the

extant organisms, relative eye size allows us to make infer-

eye fills a relatively small part of the orbit. For instance, in

ences about the habits of extinct animals. For instance, Russell

Alligator the eye is only 65% the diameter of the orbit (Stevens

and Seguin (1982) suggested that the large orbits of Troodon

2006). In particular, in many dinosaurs (e.g., Psittacosaurus,

could be an adaptation for crepuscular or nocturnal habits;

Saurolophus, Diplodocus), the sclerotic ring indicates an eye

Fiorillo and Gangloff (2000) further speculated that such an

much smaller than the orbit could potentially accommodate

adaptation might be useful during dark winters at high lati-

(pers. obs.). Another complicating factor is that, rather than

tudes of the Arctic, where Troodon appears to have been com-

simply being an ‘‘eye socket,’’ the orbit is an important struc-

mon. A study by Motani et al. (1999) looked at the large eyes

tural feature of the skull. Relatively small orbits may evolve to

of ichthyosaurs, and interpreted them as an adaptation to see-

increase the overall strength of the skull (Henderson 2002),

ing in the dim light of the deep sea. Although Horner (1994)

but relatively large orbits, as seen in hadrosaurs and sauro-

argued that the relatively small eyes of Tyrannosaurus rex indi-

pods, while weakening the skull, could function to reduce the

cated that it was a scavenger, he failed to consider allometric

weight of the skull thus improving the strength-to-weight

effects. In contrast, Stevens (2006) has argued that given the

ratio. There is also the problem that the orbit is irregularly

large size of the T. rex eye, and assuming the existence of optics

shaped in many taxa; for instance it is keyhole-shaped in ty-

and a retina with receptor densities comparable to those of

rannosaurids and many large theropods (Chure 1998) and

modern birds, T. rex would probably have had great visual

craniocaudally elongate in Ciconiiformes (pers. obs.), which

acuity. Murray et al. (2003) have argued that the small eyes of

complicates measurement and comparative analyses.

the giant, flightless mihirungs (Dromornithidae) of Australia

In this study these issues were circumvented by measuring

are consistent with herbivorous habits, rather than predatory

the diameter of the sclerotic ring itself. The sclerotic ring is an

habits. Finally, orbit size has been used to infer nocturnality

imbricating series of ossicles found in the sclera of the tetrapod

(Sieffert et al. 2005; Rossie et al. 2006; Ross et al. 2007) and

eye (Edinger 1929; Walls 1942; Romer 1956; Franz-Odendaal

diurnality (Simons 1997; Ni et al. 2003; Ross et al. 2007) in

and Vickaryous 2006). In the Sauropsida, the sclerotic ring

extinct primates.

is thought to function in maintaining the shape of the eye, in protecting the eye, and in the accommodation of the eye

The Sclerotic Ring as a Metric of Eye Size

(Walls 1942). Although the sclerotic ring is lost in snakes, crocodilians, and Mammalia, the structure is present in birds, tur-

While it seems logical that the bony orbit of the eye would

tles, lizards and the tuatara (Edinger 1929; Walls 1942; Franz-

provide a reliable estimate of eye size, in fact, eye size and orbit

Odendaal and Hall 2006; Franz-Odendaal and Vickaryous

size appear to be weakly correlated in archosaurs (Hall 2008;

2006). The sclerotic ring is probably primitive for Amniota,

pers. obs.). In some taxa, the eye fills the entire orbit (e.g.,

given its presence in the amniote sister group Diadecto-

The Function of Large Eyes in Protoceratops 311

morpha (Berman et al. 2004), as well as basal synapsids such as

Table 22.1. Avian Ecotypes Used in the Study of Avian Eye Size

dicynodonts (Sullivan and Reisz 2005), and numerous sauropsid taxa, including Ichthyosauria (Edinger 1929; Motani et al. 1999), Shuvosauridae (Nesbitt 2007), Pterosauria (Eaton 1910), and the Dinosauria (Edinger 1929).

Ecotype Nocturnal

Description Birds that primarily forage at night: Strigiformes (owls), Caprimulgiformes (Nighthawks), Nycticorax

Preliminary studies suggest that the dimensions of the eye in avians can be reliably estimated from the sclerotic ring

(Night Heron), Burhinus (Stone Curlews) Diomedea

(Schmitz et al. 2007; Hall, 2008), and given the similar design

immutabilis (Laysan Albatross)

of the eye and the sclerotic ring in various sauropsids (Walls

Predator

Visually oriented, diurnally hunting predators:

1942) the sclerotic ring should be useful in assessing eye size in

Falconiformes (falcons, hawks, and eagles), Laniidae

other taxa, including Ornithischia and non-avian Saurischia.

(shrikes), Ciconiidae (herons), Apodidae (swifts), Hirundinidae (swallows), Sterninae (terns), Lanius

The foregoing observations suggest that use of the sclerotic ring size to infer habits is a sound approach. However, for this

(kingbird), Geococcyx (roadrunner), Cariama (se-

study I did not attempt to reconstruct the dimensions of the

riema), Alcedinidae (kingfishers), etc.

soft tissues of the eye. Instead, I examined the correlation be-

Herbivore

Birds in which the diet primarily consists of plant

tween the sclerotic ring size (diameters) and activity patterns

matter including foliage, seeds, and/or fruit: An-

in archosaurs. This approach of inferring function directly

serinae (geese), Galliformes (game birds), Colum-

from osteology was undertaken because it involves fewer steps

biformes (pigeons), Psittacidae (parrots), etc.

than trying to infer function from soft tissues, which, in turn,

Probers and

Birds that do not use vision to acquire prey: Anas

are inferred mostly from osteology.

Filter Feeders

(dabbling ducks), Threskiornithidae (ibises and spoonbills), Numenius (curlews), Tringa (sandpipers), etc.

Materials and Methods

Scavengers

Birds that primarily scavenge for food: Falconiformes (Old World Vultures), Cathartidae (New World Vul-

DATA COLLECTION

tures), Giant Petrel (Macronectes), Raven (Corvus corax), sheathbill (Chionis)

To understand the correlation between eye size and ecology in living archosaurs, the aperture diameter and external diame-

Mixed

Birds that consume a combination of live prey, car-

ter of the sclerotic ring were measured for more than 130 spe-

Feeders

rion, and/or plant matter: Corvidae (magpies and

cies of bird. This sample encompasses a broad range of body

crows), Tadorna (shelducks), Procellariformes (pe-

sizes, ecologies, and clades. Skeletal specimens were measured

trels), Larinae (gulls)

at the University of Washington Burke Museum (Seattle) and the Smithsonian Museum of Natural History (Washington,

Aquatic

Predators that use vision to locate prey underwater:

Predator

Anhingidae (snakebirds), Phalacrocoracidae (cor-

D.C.). Measurements taken included maximum external di-

morants), Mergus (mergansers), Alcidae (auks),

ameter and the maximum internal diameter (aperture diame-

Sphenisciformes (penguins), Gaviiformes (loons),

ter) of the sclerotic ring, measured with digital calipers. In

Podocipediformes (grebes). Plunge divers that first lo-

most cases, body masses had been recorded for the selected

cate prey from the air have been (somewhat ar-

specimens before skeletonization; otherwise, body mass data

bitrarily) classified as diurnal predators

were sourced from the literature (Dunning 1993). Species were assigned to one of six general ecotypes: nocturnal, predatory, herbivorous, scavenger, or mixed feeder (Table 22.1), largely

ments come from different individuals of the same species,

based on Elphick et al. (2001).

but in most cases they were made from a single specimen.

The same measurements were taken for dinosaurs, using published figures, casts, and museum specimens. Figures were measured by scanning the figure and measuring it using the

Analysis

ImageJ image analysis program (Rasband 2007). Casts and

With respect to body mass, eye size scales with negative al-

specimens were measured either with digital calipers, or pho-

lometry (Fig. 22.4; Brooke et al. 1999; Hall and Ross 2007; Ross

tographed and measured in ImageJ. Body mass data were

et al. 2007). Thus, although larger animals tend to have abso-

sourced from previous studies (Paul 1988, 1997; Therrien and

lutely larger eyes, eye size increases at a lower rate relative to

Henderson 2007; Paul unpublished data) or calculated using

increasing body mass. The result is that small animals have

the regression model of Therrien and Henderson (2007). In

proportionately large eyes, and large animals have relatively

some cases, body mass estimates and sclerotic ring measure-

small eyes (e.g., Tyrannosaurus rex), regardless of their ecology.

312 longrich

FIGURE 22.4.

Bivariate plot of eye mass versus body mass for falcons, hawks, and eagles (Falconiformes). The relationship between eye size and body mass exhibits negative allometry: larger animals have eyes that are absolutely larger, but the eye represents a smaller percentage of the animal’s mass. Eye volume is approximated by 4/3pr3 where r is the radius of the sclerotic ring, and eye density was assumed to be 1.0 grams/cm3.

Eye size, therefore, must be examined in the context of al-

nocturnal birds plot out well above all other birds. Second,

lometry. We cannot simply consider the ratio of eye volume to

among diurnal birds, predators plot out above other birds,

body mass, instead, it is necessary to use an allometric equa-

such as mixed feeders, herbivores, and scavengers. These pat-

tion to examine whether the eyes are larger or smaller than

terns fit well with the expectation that large eyes should be

predicted for a given body mass.

seen where a bird needs high visual acuity, high sensitivity, or

Data were log-transformed and fit to a Reduced Major Axis

a combination of the two.

(RMA) regression, a linear model that accounts for measurement error in both the X and Y variables and uses PAST software (Hammer et al. 2001). For the allometric equation Y =

DINOSAUR DATA

aXb, the exponent b is found by the slope of the regression line

As with birds, the dinosaur eye size scales with negative al-

on a log-log plot, and the constant a is given by the Y-intercept

lometry (Table 22.2); aperture diameter scales with the .175

of the regression. Bivariate plots of the data were used to visu-

power of body mass, and ring diameter scales with the .177

alize the data trends.

power of body mass. Thus, the relatively small size of the sclerotic rings in animals such as Centrosaurus and Tarbosaurus is an expected consequence of large body size and does not nec-

Results

essarily say anything about their ecology. Likewise, the rela-

BIRD DATA

tively large eyes of many small dinosaurs are expected, and are not necessarily evidence of nocturnal or predatory habits. Ap-

As expected, when an allometric equation is fitted to the data,

erture diameter scales to the 1.02 power of sclerotic ring diam-

bird eyes show negative allometry (Table 22.2). Aperture scales

eter, suggesting that this relationship scales isometrically.

to the 0.235 power of body mass, and diameter scales to the

That dinosaurs tend to have significantly smaller eyes than

0.240 power of body mass, rather than to the 1⁄3 power ex-

birds (Fig. 22.6), as reflected in the lower slope of the regres-

pected for isometric scaling. Aperture scales to the .971 power

sions (T-test, p[.05), is not surprising. Birds typically cruise at

of ring diameter, suggesting that aperture diameter may scale

30–60 km/hr (Pennycuick 1997) and, at such speeds, good

isometrically with respect to overall diameter.

visual acuity and sensitivity are probably required. Birds are

Plots of the bird data (Fig. 22.5) reveal two patterns. First,

also under strong selective pressure to reduce mass. Whereas

The Function of Large Eyes in Protoceratops 313

Table 22.2. Regression Data for the Sclerotic Ring Dimensions of Extant Aves (Classified as per Table 22.1), and Dinosaurs (Including Archaeopteryx) Ecotype (birds) or kind (dinosaurs)

Parameter

N

Slope

Intercept

R

Aperature

17

0.23969

0.53495

0.85875

17

0.26571

0.6782

0.88127

0.94006

Birds Nocturnal

Diameter (AD) Nocturnal

External Ring Diameter (ED)

Herbivore

AD

30

0.23625

0.25008

Herbivore

ED

29

0.21789

0.50225

0.91683

Predator

AD

44

0.24219

0.38066

0.9409

Predator

ED

43

0.24431

0.57122

0.93243

Prober/Filter

AD

9

0.20247

0.33232

0.89346

Prober/Filter

ED

9

0.2172

0.45523

0.85882

Diver

AD

18

0.22628

0.33312

0.80956

Diver

ED

18

0.24005

0.45738

0.79874

Scavenger

AD

9

0.34587

Scavenger

ED

9

0.35723

0.058296

0.76337

Mixed

AD

7

0.18541

0.46937

0.83006

Mixed

ED

7

0.18455

0.62938

0.76268

All Aves

AD

134

0.23619

0.34757

0.73661

All Aves

ED

132

0.24261

0.52094

0.70629

Theropod

AD

12

0.20862

0.42792

0.90945

Theropod

ED

12

0.20229

0.68885

0.94754

Ornithischia +

AD

15

0.21325

0.19401

0.9032

ED

15

0.22064

0.39729

0.95185

All dinosaurs

AD

27

0.17481

0.476

0.88426

All dinosaurs

ED

27

0.17639

0.71004

0.91891

–0.1106

0.84928

Dinosaurs

Sauropodamorpha Ornithischia + Sauropodamorpha

bird postcrania exhibit adaptations to reduce mass (loss of the

Among non-theropods, the eyes of Protoceratops are excep-

tail and pneumatization of the skeleton) it may be difficult

tionally large (Table 22.3; Fig. 22.7 [‘‘Pa’’]). The eyes of a 164 kg

to reduce the mass of the eye without sacrificing acuity or

Protoceratops are 139% the diameter predicted by the regres-

sensitivity. Accordingly, this results in a higher eye size/body

sion for ornithischia + sauropodamorpha—about the size

mass ratio than in other vertebrates. Dinosaurs therefore can-

predicted for a 750 kg herbivorous dinosaur. The aperture is

not be evaluated using avian regressions.

180% the predicted value for a typical dinosaurian herbivore

Although dinosaurs do not fit the avian model some of the

of this size; in other words, an aperture of this diameter is

patterns seen in birds appear in dinosaurs as well. In general,

typical for a 250 kg theropod or a 2.6 ton herbivore. It would

for a given body mass, theropods (mostly consisting of preda-

be interesting to examine the relative eye size of other basal

tors and/or scavengers) have larger apertures and larger ring

neoceratopsians, such as Bagaceratops, but so far, sclerotic

diameters than do the herbivorous Ornithischia and Sauropo-

rings are known only in Protoceratops. The basal ceratopsian

damorpha (Fig. 22.7). Therefore, we cannot compare dino-

Psittacosaurus mongoliensis and the ceratopsid Centrosaurus ap-

saurs to birds, but we can compare dinosaurs to each other,

ertus plot out within the space occupied by other herbivorous

and the patterns seen in birds can help us understand the

dinosaurs (Fig. 22.7).

patterns in the dinosaurian data.

314 longrich

Why Did Protoceratops Have Large Eyes? As shown in the preceding section, relatively large eyes occur in extant archosaurian taxa that are active predators and/ or nocturnal. Among dinosaurs, Protoceratops has unusually large eyes, but is a taxon that is traditionally interpreted (or popularly pictured) as a diurnally active herbivore (Dodson 1996). Two possible reinterpretations of Protoceratops’s paleobiology that are supported by the presence of large eyes are assessed here: (1) Protoceratops was a predator; (2) Protoceratops was a nocturnal animal.

HYPOTHESIS I: PROTOCERATOPS WAS A PREDATOR One possible explanation for the large eyes of Protoceratops is that the animal was omnivorous, feeding on plants but supplementing its diet with live prey such as lizards and mammals. Ostrom (1966) interprets Protoceratops as an herbivore, but outside of his study, surprisingly little work has been done on the dietary habits of Protoceratops. For this reason, I feel that it is necessary to test the hypothesis that Protoceratops used its large eyes to hunt, rather than to dismiss the hypothesis outright simply because it goes against conventional wisdom. According to this hypothesis, large eyes are an adaptation to provide the high visual acuity needed to capture prey. If so, then we would expect to see other evidence for predatory habits in Protoceratops. In particular, we would predict that (1) Protoceratops would display adaptations facilitating prey capture, such as binocular vision, (2) Protoceratops would display adaptations for processing meat, and (3) Protoceratops would be relatively uncommon in its ecosystem, because animals tend to become rarer when they feed higher on the food chain. Binocular vision, where the visual fields of the two eyes overlap and allow the animal to be trained on a single object is frequently seen in predatory animals (Stevens 2006). Surprisingly, binocular vision appears to have been well-developed in Bivariate plots of sclerotic ring dimensions versus body mass. (A) Sclerotic ring aperture internal diameter (millimeters) versus body mass (grams) for extant birds; (B) sclerotic ring external diameter (millimeters) versus body mass for extant birds. For a given body mass, herbivores, scavengers, probers, and filter feeders have relatively small eyes compared to diurnal predators. Omnivores tend to be intermediate in eye size, and nocturnal birds have the largest eyes. To clarify trends among terrestrial forms, aquatic predators have been excluded from the plot. FIGURE 22.5.

Protoceratops; with approximately 30\ of arc covered by both the left and right eyes (Fig. 22.8). In this respect, the morphology of Protoceratops is consistent with predatory habits. However, binocular vision is not always correlated with predatory habits in extant taxa (Stevens 2006), as discussed below. The hypothesis that Protoceratops was predatory also predicts that the jaws and teeth would be specialized for processing meat. However, the jaws and teeth appear to have functioned for processing plant matter. Although the predentary has a moderately sharp tip, the upper beak is blunt and lacks the sharp hook seen in hawks, owls, and shrikes (Fig. 22.9B). Caniniform teeth are present (Fig. 22.9B) but these are blunt tusks, unsuitable for tearing prey. The posterior teeth are clearly those of an herbivore. The tooth batteries formed coarsely serrated occlusal surfaces (Fig. 22.9C–E), with small

The Function of Large Eyes in Protoceratops 315

FIGURE 22.6.

Bivariate plots of sclerotic ring (millimeters) versus body mass (grams) for dinosaurs (including Archaeopteryx) and extant birds. Upper plot is for ring aperture; lower plot is for ring diameter. Birds generally have larger eyes than non-avian dinosaurs and Archaeopteryx.

accessory serrations (Fig 22.9E), and these surfaces met at an

22.9E). These heavy wear facets are similar to the wear facets

oblique angle, shearing and crushing the food (Ostrom 1966).

seen in herbivorous dinosaurs (pers. obs.), rather than those

Among extant taxa, the arrangement is closely approached by

of carnivorous theropods such as dromaeosaurs and tyranno-

the herbivorous agamid Uromastyx (pers. obs.). As in hadro-

saurs, where the teeth exhibit small wear facets and con-

saurs and ceratopsians, but unlike carnivorous theropods, the

choidal fractures (Schubert and Ungar 2005; pers. obs.).

teeth have a thick enamel layer (Hwang 2005) on one side

Finally, the great abundance of Protoceratops in the Dja-

of the tooth, but were heavily worn before being shed (Fig.

dokhta Formation at Bayn Dzak, Tugriken Shireh, Ukhaa

316 longrich

Table 22.3. Data for Dinosaurs Examined in this Study Specimen

Taxon

Mass

Aperture

AQ

Diameter

DQ

JM 2257

Archaeopteryx recurva

220

5.9

0.77

11.2

0.85

HMN 1880

Archaeopteryx lithographica

252

7.6

0.96

13.4

0.99

IVPP V12733

Mei long

430

9.6

1.11

18.6

1.24

IVPP V13352

Microraptor gui

500

11.1

1.25

19.2

1.25

NGMC 91

cf. Sinornithosaurus

1130

14.8

1.45

23.3

1.31

NMMNH specimen

Coelophysis bauri

1301

12.5

1.19

26.3

1.45

Gaston Design cast

undescribed dromaeosaur

7020

19

1.35

32

1.31

MNA V2623

Coelophysis kayentakatae

12398

28.3

1.82

40.2

1.49

AMNH 5356

Dromaeosaurus albertensis

16290

24.7

1.51

35.6

1.25

GIN 100/13

Garudimimus brevipes

85000

22

1.01

41.2

1.08

DINO 11541

Allosaurus jimmadseni

460000

45.2

1.55

69.4

1.36

GIN 100/70

Tarbosaurus baatar

30

0.86

65

1.07

BMNH R8501

Lesothosaurus diagnosticus

2400

8.3

0.71

13

0.64

AMNH 6254

Psittacosaurus mongoliensis

12000

12.7

0.82

20.8

0.77

BMNH R2477

Hypsilophodon foxi

21000

17.4

1.02

25.2

0.85

ROM 804

Parksosaurus warreni

43000

20.2

1.05

37

1.1

BP-1-4779

Massospondylus carinatus

93000

13.4

0.61

24.2

0.63

AMNH specimen

Plateosaurus sp.

440000

17.2

0.59

37.4

0.74

CM 11338

Camarasaurus lentus

640000

22.5

0.73

46.6

0.86

AMNH 5351

Centrosaurus apertus

1400000

34

0.96

57.4

0.92

ROM 1218

Lambeosaurus lambei

2000000

32.6

0.86

52.2

0.79

SM 4037

Edmontosaurus annectens

2400000

44.9

1.15

80

1.17

AMNH 5338

Corythosaurus casuarius

2800000

43.5

1.09

70

1

AMNH 5220

Saurolophus osborni

2970000

53.7

1.33

83

1.17

AMNH 5386

Prosaurolophus maximus

6525090

43.2

0.93

76

0.93

Z. PAL MgD-I/9

Nemegtosaurus

8400000

42.7

0.88

85

1

CM 11161

Diplodocus longus

11400000

31.8

0.62

65.6

0.73

AMNH 6466

Protoceratops andrewsi

164000

36.5

1.5

49

1.15

1245451

Mass in grams, aperture diameter, and total ring diameter in millimeters. AQ: aperture quotient (actual/predicted by dinosaur regression); DQ: diameter quotient (actual/predicted by dinosaur regression).

Tolgod, and Bayan Mandahu (Dong and Currie 1993; Jerzykie-

abundance of Protoceratops are all consistent with the hypoth-

wicz et al. 1993; Fastovsky et al. 1997; Gao and Norell 2000;

esis that the animal was primarily, or exclusively, herbivorous.

Lambert et al. 2001) is consistent with it having been an herbivore rather than an omnivore or carnivore. In Mesozoic and Cenozoic terrestrial ecosystems, herbivores outnumber predators, often by an order of magnitude (Bakker 1972; Farlow

HYPOTHESIS II: PROTOCERATOPS WAS A NOCTURNAL ANIMAL

1990). Of the 108 dinosaurs collected by the Central Asiatic

Another possible explanation for the large eyes of Protocera-

Expeditions at Bayn Dzak, 101, or 94%, are Protoceratops (Fig.

tops is that Protoceratops was a nocturnal animal with eyes that

22.10). Given that collected samples of vertebrate fossils often

were adapted for seeing in low light conditions. The hypothe-

tend to preferentially include rare and unusual forms at the

sis is an intriguing one. Nocturnality has previously been pro-

expense of common forms, these numbers likely underesti-

posed for some dinosaurs (Russell and Seguin 1982; Vickers-

mate the abundance of Protoceratops at Bayn Dzak. Tapho-

Rich and Rich 1993; Fiorillo and Gangloff 2000), but little

nomic issues notwithstanding, had Protoceratops fed high on

compelling evidence has been presented so far. It remains,

the food chain as a predator, then its fossil remains would be

however, a plausible hypothesis. Given that nocturnal habits

expected to be considerably less common than they are.

have evolved multiple times in birds and squamates (Table

In summary, the large eyes and anteriorly inclined orbits of

22.4), it would be unsurprising if a group as long-lived and

Protoceratops could be read as adaptations for prey capture, but

diverse as the dinosaurs evolved nocturnal forms as well. If

other evidence is inconsistent with this hypothesis. Morphol-

Protoceratops was nocturnal, then we should expect to find

ogy of the teeth and dentition, tooth wear, and the relative

other features of the eye and orbits that are associated noc-

The Function of Large Eyes in Protoceratops 317

FIGURE 22.7.

Bivariate plot of sclerotic rings size versus body mass for Mesozoic dinosaurs (including Archaeopteryx). Aj: Allosaurus jimmadseni; Al: Archaeopteryx lithographica; Ar: Archaeopteryx recurva; Ca: Centrosaurus apertus; Cb: Coelophysis bauri; Cc: Corythosaurus casuarius; Ck: ?Coelophysis kayentakatae; Cl: Camarasaurus lentus; Da: Dromaeosaurus albertensis; Di: Diplodocus; Ea: Edmontosaurus annectens; Gb: Garudimimus brevipes; Hf: Hypsilophodon foxii; Ld: Lesothosaurus diagnosticus; Ll: Lambeosaurus lambei; Mc: Massospondylus carinatus; Mg: Microraptor gui; Ml: Mei long; Nm: Nemegtosaurus mongoliensis; Pa: Protoceratops andrewsi; Pm: Psittacosaurus mongoliensis; Pm: Prosaurolophus maximus; Pw: Parksosaurus warreni; Sm: Sinornithosaurus milleni; So: Saurolophus osborni; UD: undescribed dromaeosaur.

318 longrich

FIGURE 22.8.

Binocular vision in Protoceratops. (A) Anterior view, showing how both orbits are visible; (B) dorsal view. There is a large area ahead of the rostrum where the field of view of each eye overlaps. The orbits of Protoceratops are more rostrally inclined than in many predatory theropods. Drawn from a skull of Protoceratops sp. on display at the Mongolian Academy of Sciences, Natural History Museum, Ulan Bator.

FIGURE 22.9. Upper jaws and teeth of Protoceratops andrewsi. (A) Skull in right lateral view; (B) beak and premaxillary teeth; (C) lateral view of the maxillary tooth battery; (D) ventral view of the tooth battery, showing oblique orientation of the wear facets; (E) ventromedial view, showing the coarsely serrated cutting surface and accessory serrations (arrows) formed by enamel ridges. Scale bars are 20 mm.

The Function of Large Eyes in Protoceratops 319

FIGURE 22.10. Numbers and kinds of dinosaurs collected from Bayn Dzak (Flaming Cliffs) by the Central Asiatic Expeditions from 1922 to 1925. Protoceratops andrewsi comprises over 90% of the fossil collection. As one of the few large animals in the fauna, and the most common animal at Bayn Dzak, Protoceratops probably fed low on the food chain. Data from Brown and Schlaikjer (1940), Gilmore (1933), Osborn (1924), and Dashzeveg et al. (2005).

Table 22.4. Extant, Nocturnal Diapsids

turnal vision. Finally, there is the question of whether nocturnality is consistent with the known ecology of Protoceratops.

Aves Caprimulgiformes (nightjars,

Squamata Sphenodon (Tuatara)

Aperture Diameter. To increase the rate at which light enters the eye, forms adapted for vision in low light often have rela-

nighthawks, oilbirds, owlet-

tively large apertures (Caprimulgiformes—nighthawks and

nightjars)

allies—among birds, and Gekkota (geckos) and Xantusiidae

Nycticorax (night herons)

Creagrus furcatus (Galapagos

Gekkota (geckos; clade also con-

(night lizards) among lizards [pers. obs.]). Because aperture

tains secondarily diurnal taxa)

diameter scales isometrically with ring diameter, the size of an

Xantusiidae (night lizards)

swallowtailed gull)

aperature relative to body size can be described and compared between taxa simply by calculating the ratio of aperture diameter to sclerotic ring diameter; here termed relative aperture

Apteryx (kiwis)

Egernia striata, E. inornata (night skink and desert skink, respectively)

Burhinus (stone curlews)

size (RAS). For diurnal birds mean RAS is 0.64 (Fig. 22.11). For diving birds mean RAS is larger (0.68), perhaps to see in dimly lit

Lanthanotus borneensis (earless

underwater environments. The bizarre, tubular eyes of Strigi-

monitor)

formes (owls) actually have a mean RAS of 0.56, but when

Leiolopisma suteri (black shore

owls are excluded, the remaining nocturnal birds have a large

skink)

mean RAS of 0.71. Thus, in birds, large apertures appear to be

Diomedea immutabilis (Laysan al-

Corucia zebrata (monkey-tailed

related to seeing in low light conditions, although small aper-

batross)

skink)

tures are not necessarily indicative of diurnality.

Strigops habroptilus (kakapo)

Viperidae (vipers)

RAS ratio for Protoceratops is 0.74, the largest of any dino-

Acanthophis (death adder)

saur measured. This value also places Protoceratops in the up-

Hypsiglena (night snakes)

per quartile of the RAS range for nocturnal birds (excluding

Lampropeltis (kingsnakes)

Strigiformes).

Boiga (tree snakes)

320 longrich

As a group, dinosaurs have a low mean RAS (0.59). But the

Pythonidae (pythons and boas)

Binocular Vision. Although binocular vision is often inter-

FIGURE 22.11. Relative size of the aperture in birds and dinosaurs. The aperture varies considerably in size in nocturnal birds. Strigiformes (owls) have relatively small apertures, perhaps as a result of their telescopic eyes, but other nocturnal birds and diving birds have large apertures. The aperture for Protoceratops is above the mean for nocturnal (non-Strigiformes) birds, and larger than any other dinosaur measured.

preted as an adaptation to improve depth perception particu-

north-central China have shown consistently that the climate

larly in predators, it is not necessarily associated with either

of the region during the Late Cretaceous (Campanian) was

(Stevens 2006). A number of studies have shown that, in fact,

semi-arid to arid ( Jerzykiewicz and Russell 1991; Eberth 1993;

the presence of binocular vision is often associated with forag-

Jerzykiewicz et al. 1993; Fastovsky et al. 1997; Loope et al.

ing in low-light conditions. Nocturnal geckos and primates

1998; Dashzeveg et al. 2005). Analyses of paleoenvironments

tend to have well-developed binocular vision (Röll 2001; Ross

at Bayn Dzak and Bayan Mandahu—sites that have both pro-

and Kirk 2007; Ross et al. 2007) and many deep sea fish have

duced many tens to hundreds of Protoceratops skeletons—

highly developed binocular vision (Walls 1942; Warrant and

indicate basin margin settings characterized by coeval eolian

Locket 2004). Binocular vision improves sensitivity by allow-

dunes, alluvial fans, and ephemeral streams and lakes (Eberth

ing two eyes to focus on an object, doubling the number of

1993; Fastovsky et al. 1997), which may have experienced

receptors capturing light from the subject. The presence of

long-term climatic variations in wetness and aridity (Eberth

binocular vision in Protoceratops is therefore consistent with

1993; Loope et al. 1998). Bayan Mandahu and Ukhaa Tolgod

nocturnal habits.

preserve evidence of active and stabilized dunes (Eberth 1993;

Environmental Context. Understanding Protoceratops’s un-

Loope et al. 1998), whereas Tugriken Shireh—where articu-

usual paleoenvironmental associations offers insight into the

lated Protoceratops are particularly common—is dominated by

possible function of its eyes. Geological studies of Djadokhta

deposits of an active eolian dune field (Fastovsky et al. 1997).

Formation fossil localities across southern Mongolia and

Caliches are present at all of these localities, indicating that

The Function of Large Eyes in Protoceratops 321

Diversity of the Bayn Dzak dinosaur assemblage (Djadokhta Formation) compared to the diversity of the Nemegt dinosaur assemblage (Nemegt Formation). Bayn Dzak is dominated by Protoceratops, exhibits low diversity, and has few large animals. The Nemegt assemblage is characterized by higher diversity, and the presence of large herbivores (therizinosaurs, duckbills, and titanosaurs); ornithomimids and hadrosaurs are the dominant herbivores. The dinosaur diversity pattern from Bayn Dzak suggests that Protoceratops inhabited a stressed environment. Variations on these assemblages occur throughout the Upper Cretaceous of the Gobi Desert: the Protoceratopsid Fauna occurs in the semi-arid to arid aeolian/alluvial fan deposits at Bayn Dzak, Tugriken Shireh, Ukhaa Tolgod, and Bayan Mandahu (Djadokhta Formation), as well as Khulsan and Khermin Tsav (Barun Goyot Formation). An Ornithomimid-Hadrosaur fauna characterizes wetter, fluvial-dominated paleoenvironments, such as those inferred for the Nemegt Formation, the Baynshiree Formation, and Iren Dabasu. Nemegt data from pers. obs. of the Nemegt locality, collections at the Geological Institute of Mongolia, and P. Currie (pers. com. 2007). Djadokhta data from Osborn (1924), Gilmore (1933), Brown and Schlaikjer (1940), Tereschenko and Alifanov (2003), and Dashzeveg et al. (2005).

FIGURE 22.12.

seasonal aridity was associated with seasonally warm to hot

rannosaurs are rare or absent, and the medium-sized Protocera-

temperatures and seasonal evaporation/transpiration of sur-

tops is the largest animal common enough to be ecologically

face and ground waters (Retallack 1990; Birkeland 1999). The

important. A taphonomic bias toward preservation of smaller

abundance of protoceratopsids in such paleoenvironments

animals could explain the absence of adult tyrannosaurs,

(Fig. 22.10), and their rarity or absence in association with

hadrosaurs, and sauropods, but does not convincingly explain

the somewhat wetter paleoenvironments of the Upper Cre-

the absence of juveniles or eggs from these same taxa. In con-

taceous Bayn Shiree, Iren Dabasu, Javkhlant, Barun Goyot,

trast, in the Maastrichtian age Nemegt Formation, Tarbosaurus

and Nemegt formations (Fig. 22.12), provide strong evidence

is common, and giant herbivores such as Saurolophus, Tarchia,

that protoceratopsids (and associated taxa) were uniquely

Therizinosaurus, Deinocheirus, and Nemegtosaurus occur.

adapted to thrive under these warm to hot, semi-arid to arid conditions.

Small animals are very diverse in the Djadokhta, particularly lizards (Gao and Norell 2000) and mammals (Simpson 1925;

Faunal Context. The fauna of the Djadokhta Formation—the

Gregory and Simpson 1926; Clemens and Kielan-Jaworowska

‘‘Protoceratops fauna’’—shows abundance and diversity pat-

1979; Jerzykiewicz and Russell 1991) and is a pattern consis-

terns indicative of stressed and hot-to-warm, terrestrial paleo-

tent with those animals inhabiting a hot, arid environment; in

environments. As discussed above (Fig. 22.10), Protoceratops

particular, the extraordinary diversity of lizards parallels the

is extremely common whereas other taxa are comparatively

situation seen today in the hot deserts of Australia (Gao and

rare, overall species diversity is low, and large dinosaurs are not

Norell 2000).

well represented (Fig. 22.12). Hadrosaurs, sauropods, and ty-

322 longrich

Finally, aquatic animals are rare and low in diversity. One of

the few turtles found in the Djadokhta, Zangerlia, is a terres-

oilbirds (Griffin 1953; Konishi and Knudsen 1979) use active

trial, tortoise-like form (Brinkman and Peng 1996; Joyce and

sonar to navigate. If Protoceratops were nocturnal, then we

Norell 2005). Notwithstanding taphonomic biases, these pat-

might expect to find evidence of an unusually well-developed

terns agree well with the geologic data and interpretations

sense of smell and/or sense of hearing. Future studies of the

that during Djadokhta ‘‘time,’’ Central Asia was characterized

olfactory bulbs and ear, therefore, may provide a test of this

by extensive sandy deserts that had warm-to-hot, semiarid-to-

hypothesis, or indications of how well Protoceratops could see

arid climates during all or parts of the year.

in the dark.

The challenges for modern vertebrates living in warm to hot

The hypothesis of nocturnal habits raises other interesting

desert and desert margin settings are well known. During the

possibilities concerning the biology of Protoceratops. First, at

summer, subtropical deserts such as the Sonora Desert, Austra-

night, a desert can still be hot enough to pose thermoregula-

lian Desert, and Sahara Desert may see daytime air tempera-

tory challenges. Even after dark, Protoceratops may have needed

tures reach 40\–50\. As air temperature overtakes body tem-

to shed excess heat. Many nocturnal desert mammals pump

perature, evaporative cooling remains one of the few ways to

blood through radiative cooling devices to cool themselves; in

rid the body of heat (Walsberg 2000). However, evaporative

particular, jackrabbits, desert foxes, and desert hedgehogs have

cooling requires water, which, in a desert, is in short supply.

unusually large ears, which are thought to function as radia-

Accordingly many modern desert vertebrates have nocturnal

tors (Schmidt-Nielsen 1964; Nowak 1991). This suggests a pos-

activity patterns, seeking shade and cooler refuges during the

sible function for the frill and the long neural spines of the tail:

day, and foraging at night when cooling is less of a problem.

their large surface areas may have allowed them to shed heat

In the modern Sonoran Desert, more than 90% of the mam-

effectively once the air temperature had dropped below body

mal species are nocturnal (Walsberg 2000), and large desert

temperature, similar to the thermoregulatory function pro-

mammals such as peccaries and desert mule deer are noctur-

posed for the frill of Triceratops (Barrick et al. 1998). Although

nal during the hot summer months (Bigler 1974; Bissonette

the sexual dimorphism of the frill indicates that it served an

1978; Hayes and Krauseman 1993). As a nocturnal dinosaur,

intraspecific display function (Kurzanov 1972; Dodson 1976),

Protoceratops may have managed to thrive during times when

the frill may had multiple roles, just as modern ungulate horns

the Djadokhta paleoenvironments were harshest and where

function for thermoregulation (Taylor 1966), combat, and dis-

relatively few species of dinosaur were able to survive.

play (Happ this volume). There is also the question of where a nocturnal Protoceratops could have found shelter from the heat. Deer often shelter

Implications and Conclusions

from the heat in shady areas such as beneath rock overhangs

THE BIOLOGY OF PROTOCERATOPS

(pers. obs.), peccaries are known to shelter in caves (Bigler 1974), and desert cottaintails find shelter under bushes (Hinds

Protoceratops may have been nocturnal, resting through the

1973). It is unclear whether the sandy Djadokhta environ-

hot daylight hours and foraging and feeding at night. The

ment would have had many overhangs or cliffs, and Protocera-

combination of large eyes, a large aperture, and binocular vi-

tops would have required relatively large bushes to find shade.

sion would have improved the ability of the eyes to form im-

Another possibility is that Protoceratops may have taken shel-

ages at night. Among modern animals, a possible analogue for

ter in a scrape or dug into the sand. Many desert animals bur-

a nocturnal Protoceratops is the collared peccary, Tayassu ta-

row to find refuge from daytime heat (Nowak 1991), including

jacu. Peccaries dwell in the Sonoran Desert, use their robust

the nocturnal fennec fox (Fennecus zerda), kit fox (Vulpes mac-

jaws to feed on cactus and other desert plants (Corn and War-

rotis), and the aardvark (Orycteropus). Some desert-dwelling

ren 1985), and are primarily nocturnal during the hot summer

birds even use lizard burrows as thermal refuges (Williams et

months (Bigler 1974).

al. 1999).

This hypothesis that Protoceratops was nocturnal cannot be

Dinosaurs are not generally thought of as burrowers, but

considered proven, but it is testable. When the ability to see is

there is good evidence that some hypsilophodonts dug and

compromised, animals often compensate by relying on other

inhabited burrows (Varricchio et al. 2006). Typical burrowing

senses. Nocturnal and crepuscular birds tend to exhibit en-

adaptations, such as a large olecranon process and large claws

larged olfactory bulbs (Healy and Guilford 1990). In particular

(Hildebrand and Goslow 2001) are absent from the forelimb

the kakapo (Strigops habroptilus) and kiwi (Apteryx) are thought

of Protoceratops. Nevertheless, many burrow-dwellers, such as

to have a good sense of smell (Hagelin 2004; Martin et al.

rabbits, foxes, magellanic penguins (Spheniscus magellanicus),

2007), and the same also appears to be true for geckos (Dial

puffins (Fratercula spp.), and the burrowing parrot (Cyano-

and Schwenk 1996). Acoustic cues are important in other noc-

liseus patagonus) lack obvious fossorial adaptations.

turnal animals. Owls use passive echolocation (Payne 1970)

Another possibility is that Protoceratops used the powerful

swiftlets (Novick 1959; Griffin and Thompson 1982) and

hindlimbs to dig, as in the kiwi (Apteryx) and it is interesting,

The Function of Large Eyes in Protoceratops 323

FIGURE 22.13. Articulated left pes of Protoceratops andrewsi (AMNH 6476). (A) Dorsal view; (B) distal view. Note the short, massive foot, the long, robust toes, and the enlarged, shovel-shaped pedal unguals. Scale bar is 10 cm.

therefore, that the hindlimb of Protoceratops exhibits some anatomical features compatible with an adaptation for ‘‘backward burrowing.’’ The distal insertion of the caudofemoralis muscle would permit powerful retraction of the hindlimb, and the long, massive toes and shovel-shaped pedal unguals (Fig. 22.13) may have been effective in moving sand. From a taphonomic perspective, the interpretation of Protoceratops as a burrower could help explain why articulated skeletons of this form are so common in the Djadokhta Formation, and why they are often preserved in an upright stance. In this scenario, the animal facilitates its entry into the fossil record by burying itself. Animals that live and die underground, such a ground squirrels, often become buried by the collapse or filling-in of their burrows. Living Protoceratops may also have become entombed by the premature collapse or filling-in of their eolian-dune-hosted burrows. Over geologic time, occurrences of burrow collapse and filling-in could re-

Acknowledgments

Thanks to Jason Anderson, Demchig Badamgarav, Caleb Brown, Clive Coy, Phil Currie, Dave Eberth, David Evans, Tetsuto Miyashita, Anthony Russell, Michael Ryan, Jessica Theodor, Matt Vickaryous, and the Russell Lab for invaluable discussions, and to David Eberth and Kent Stevens for constructive reviews of the manuscript. Mick Ellison provided photos of NGMC-91; Kyo Tanoue and Peter Dodson provided photos of Psittacosaurus mongoliensis. Specimen access was provided by the Geological Institute of Mongolia, the Royal Tyrrell Museum, the Smithsonian Institution National Museum of Natural History, and the University of Washington Burke Museum. Nomadic Expeditions provided logistical support for fieldwork in the Gobi, which inspired many of the ideas laid out here. Finally, thanks to Michael Ryan for persuading me to abandon theropods long enough to pursue this project.

sult in the preservation of numerous Protoceratops in upright ‘‘standing’’ postures (e.g., Jerzykiewicz et al. 1993; Kirkland and Bader this volume), whereas ‘‘backward burrowing’’ could help to explain the occurrences of specimens that exhibit tails curved around and pointing forward (pers. obs. of a specimen at Tugriken Shireh). Accordingly, it seems reasonable that further studies of the taphonomy of in situ Protoceratops could help test this burrowing hypothesis. Undoubtedly, much remains to be learned about Protoceratops. However, what we can say is that from a paleobiological standpoint, the animal is every bit as extraordinary as the larger and more elaborately ornamented ceratopsids of North America. Protoceratops was not a primitive holdover, rather, it was a highly specialized animal that successfully exploited an extreme environment.

324 longrich

References Cited Bakker, R. T. 1972. Anatomical and ecological evidence of endothermy in dinosaurs. Nature 238: 81–85. Barrick, R. E., M. K. Stoskopf, J. D. Marcot, and D. A. Russell. 1998. The thermoregulatory functions of the Triceratops frill and horns: Heat flow measured with oxygen isotopes. Journal of Vertebrate Paleontology 18: 746–750. Berman, D. S., A. C. Henrici, R. A. Kissel, S. S. Sumida, and T. Martens. 2004. A new diadectid (Diadectomorpha) Orobates pabsti, from the Early Permian of central Germany. Bulletin of the Carnegie Museum of Natural History 35: 1–36. Bigler, W. J. 1974. Seasonal movements and activity patterns of the collared peccary. Journal of Mammalogy 55: 851–855. Birkeland, P. W. 1999. Soils and Geomorphology. New York: Oxford University Press.

Bissonette, J. A. 1978. Influence of extremes of temperature on activity patterns of peccaries. Southwestern Naturalist 23: 339– 346. Brinkman, D. B., and J.-H. Peng. 1996. A new species of Zangerlia (Testudines: Nanhsiungchelyidae). Canadian Journal of Earth Sciences 33: 526–540. Brooke, M. d. L., S. Hanley, and S. B. Laughlin. 1999. The scaling of eye size with body mass in birds. Proceedings of the Royal Society of London, Series B, 266: 405–412. Brown, B. 1917. A complete skeleton of the horned dinosaur Monoclonius, and description of a second skeleton showing skin impressions. Bulletin of the American Museum of Natural History 37: 281–306. Brown, B., and E. M. Schlaikjer. 1940. The structure and relationships of Protoceratops. Annals of the New York Academy of Sciences 40: 133–266. Chure, D. 1998. On the orbit of theropod dinosaurs. Gaia 15: 233–240. Clemens, W. A., Jr., and Z. Kielan-Jaworowska. 1979. Multituberculata. In J. A. Lilligraven, Z. Kielan-Jaworowska, and W. A. Clemens, Jr., eds., Mesozoic Mammals: The First Two-thirds of Mammalian History, pp. 99–149. Berkeley: University of California Press. Corn, J. L., and R. J. Warren. 1985. Seasonal food habits of the collared peccary in South Texas. Journal of Mammalogy 66: 155–159. Dashzeveg, D., L. Dingus, D. B. Loope, C. C. Swisher, T. Dulam, and M. R. Sweeney. 2005. New stratigraphic subdivision, depositional environment, and age estimate for the Upper Cretaceous Djadokhta Formation, Southern Ulan Nur Basin, Mongolia. American Museum Novitates 3498: 1–31. Dial, B. E., and K. Schwenk. 1996. Olfaction and predator detection in Coleonyx brevis (Squamata: Eublepharidae) with comments on the functional significance of buccal pulsing in geckos. Journal of Experimental Zoology 276: 415–424. Dodson, P. 1976. Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. Journal of Paleontology 50: 929–940. ———. 1996. The Horned Dinosaurs. Princeton: Princeton University Press. Dong, Z.-M., and P. J. Currie. 1993. Protoceratopsian embryos from Inner Mongolia, People’s Republic of China. Canadian Journal of Earth Sciences 30: 2248–2254. Dunning, J. B. 1993. Handbook of Avian Body Masses. Boca Raton, Fla.: CRC Press. Eaton, C. F. 1910. Osteology of Pteranodon. Memoirs of the Connecticut Academy of Arts and Sciences 2: 1–38. Eberth, D. A. 1993. Depositional environments and facies transitions of dinosaur-bearing Upper Cretaceous redbeds at Bayan Mandahu (Inner Mongolia, People’s Republic of China). Canadian Journal of Earth Sciences 30: 2196–2213. Edinger, T. 1929. Uber knocherne Scleralringe. Zoologische Jahrbucher. Abteilung fur Anatomie und Ontogenie der Tiere 51:163– 226. Elphick, C., J. B. Dunning, and D. A. Sibley. 2001. The Sibley Guide to Bird Life and Behavior. New York: Alfred A. Knopf.

Farlow, J. O. 1990. Dinosaur energetics and thermal biology. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 43–55. Berkeley: University of California Press. Farlow, J. O., and P. Dodson. 1975. The behavioral significance of frill and horn morphology in ceratopsian dinosaurs. Evolution 29: 353–361. Fastovsky, D. E., D. Badamgarav, H. Ishimoto, M. Watabe, and D. B. Weishampel. 1997. The paleoenvironments of TugrikenShireh (Gobi Desert, Mongolia) and aspects of the taphonomy and paleoecology of Protoceratops (Dinosauria: Ornithischia). Palaios 12: 59–70. Fiorillo, A. R., and R. A. Gangloff. 2000. Theropod teeth from the Prince Creek Formation (Cretaceous) of Northern Alaska, with speculations on arctic dinosaur paleoecology. Journal of Vertebrate Paleontology 20: 675–682. Franz-Odendaal, T. A., and B. Hall. 2006. Skeletal elements within teleost eyes and a discussion of their homology. Journal of Morphology 267: 1326–1337. Franz-Odendaal, T. A.,, and M. K. Vickaryous. 2006. Skeletal elements in the vertebrate eye and adnexa: Morphological and developmental perspectives. Developmental Dynamics 235: 1244–1255. Gao, K., and M. A. Norell. 2000. Taxonomic composition and systematics of Late Cretaceous lizard assemblages from Ukhaa Tolgod and adjacent localities, Mongolian Gobi Desert. Bulletin of the American Museum of Natural History 249: 1–118. Gilmore, C. W. 1933. Two new dinosaurian reptiles from Mongolia with notes on some fragmentary specimens. American Museum Novitates 679: 1–20. Gregory, W. K., and C. C. Mook. 1925. On Protoceratops, a primitive ceratopsian dinosaur from the Lower Cretaceous of Mongolia. American Museum Novitates 156: 1–9. Gregory, W. K., and G. G. Simpson. 1926. Cretaceous mammal skulls from Mongolia. American Museum Novitates 225: 1–20. Griffin, D. R. 1953. Acoustic orientation in the oil bird, Steatornis. Proceedings of the National Academy of Sciences 39: 884–893. Griffin, D. R., and D. Thompson. 1982. Echolocation by cave swiftlets. Behavioral Ecology and Sociobiology 10: 119–123. Hagelin, J. C. 2004. Observations on the olfactory ability of the Kakapo Strigops habroptilus, the critically endangered parrot of New Zealand. Ibis 146: 161–164. Hall, M. I. 2008. The anatomical relationships between the avian eye, orbit and sclerotic ring: implications for inferring activity patterns in extinct birds. Journal of Anatomy 212: 781–794. Hall, M. I., and C. F. Ross. 2007. Eye shape and activity pattern in birds. Journal of Zoology 271: 437–444. Hammer, Ø., D. A. T. Harper, and P. D. Ryan. 2001. Paleontological Statistics Software Package for Education and Data Analysis. Palaeontologia Electronica 4: 1–9. Happ, J. W. 2010. New evidence regarding the structure and function of the horns in Triceratops (Dinosauria: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 271–281. Bloomington: Indiana University Press. Hayes, C. L., and P. R. Krauseman. 1993. Nocturnal activity of

The Function of Large Eyes in Protoceratops 325

female desert mule deer. Journal of Wildlife Management 57: 897–904. Healy, S., and T. Guilford. 1990. Olfactory-bulb size and nocturnality in birds. Evolution 44: 339–346. Henderson, D. M. 2002. The eyes have it: The sizes, shapes, and orientations of theropod orbits as indicators of skull strength and bite force. Journal of Vertebrate Paleontology 22: 766–778. Hildebrand, M., and G. Goslow. 2001. Analysis of Vertebrate Structure. New York: John Wiley and Sons. Hinds, D. S. 1973. Acclimatization of thermoregulation in the Desert Cottontail, Sylvilagus audubonii. Journal of Mammalogy 54: 708–728. Horner, J. R. 1994. Steak knives, beady eyes, and tiny little arms (a portrait of T. rex as a scavenger). In G. D. Rosenberg and D. L. Wolberg, eds., Dino Fest, pp. 157–164. Paleontological Society Special Publication 7. Hwang, S. H. 2005. Phylogenetic patterns of enamel microstructure in dinosaur teeth. Journal of Morphology 266: 208–240. Jerzykiewicz, T., P. J. Currie, D. A. Eberth, P. A. Johnston, E. H. Koster, and J.-J. Zheng. 1993. Djadokhta Formation correlative strata in Chinese Inner Mongolia: An overview of the stratigraphy, sedimentary geology, and paleontology and comparisons with the type locality in the pre-Altai Gobi. Canadian Journal of Earth Sciences 30: 2180–2195. Jerzykiewicz, T., and D. A. Russell. 1991. Late Mesozoic stratigraphy and vertebrates of the Gobi Basin. Cretaceous Research 12: 345–377. Joyce, W. G., and M. Norell. 2005. Zangerlia ukhaachelys, new species, a Nanhsiungchelyid Turtle from the Late Cretaceous of Ukhaa Tolgod, Mongolia. American Museum Novitates 3481: 1–19. Kirkland, J. I., and K. Bader. 2010. Insect trace fossils associated with Protoceratops carcasses in the Djadokhta Formation (Upper Cretaceous), Mongolia. In M. J. Ryan, B. J. ChinneryAllgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 509–519. Bloomington: Indiana University Press. Konishi, M., and E. I. Knudsen. 1979. The oilbird: Hearing and echolocation. Science 204: 425–427. Kurzanov, S. M. 1972. Sexual dimorphism in protoceratopsians. Palaeontological Journal 1972: 91–97. Lambert, O., P. Godefroit, H. Li, C.-Y. Shang, and Z.-M. Dong. 2001. A new species of Protoceratops (Dinosauria, Neoceratopsia) from the Late Cretaceous of Inner Mongolia. Bulletin de L’Institut Royal Des Sciences Naturelles De Belgique 71: 5–28. Loope, D. B., L. Dingus, C. C. Swisher, and C. Minjin. 1998. Life and death in a Late Cretaceous dune field, Nemegt Basin, Mongolia. Geology Magazine 26: 27–30. Martin, G. R., K.-J. Wilson, J. M. Wild, S. Parsons, M. F. Kubke, and J. Corfield. 2007. Kiwi forego vision in the guidance of their nocturnal activities. PLoS ONE 2(2): e198 doi:10.1371/journal.pone.0000198 Motani, R., B. M. Rothschild, and W. M. Wahl. 1999. Large eyeballs in diving ichthyosaurs. Nature 402: 747. Murray, P. F., P. Vickers-Rich, and P. Vickers-Rich. 2003. Magnifi-

326 longrich

cent Mihirungs: The Colossal Flightless Birds of the Australian Dreamtime. Bloomington: Indiana University Press. Nesbitt, S. J. 2007. The anatomy of Effigia okeefeae. Bulletin of the American Museum of Natural History 302: 84. Ni, X., Y. Wang, Y. Hu, and C. Li. 2003. A euprimate skull from the early Eocene of China. Nature 427: 65–68. Novick, A. 1959. Acoustic orientation in the cave swiftlet. Bio logical Bulletin 117: 497–503. Nowak, R. M. 1991. Walker’s Mammals of the World, 5th ed. Baltimore: Johns Hopkins University Press. Osborn, H. F. 1924. Three new Theropoda, Protoceratops zone. American Museum Novitates 144: 1–12. Ostrom, J. H. 1966. Functional morphology and evolution of the ceratopsian dinosaurs. Evolution 20: 290–308. Payne, R. S. 1970. Acoustic location of prey by barn owls (Tyto alba). Journal of Experimental Biology 54: 535–573. Paul, G. S. 1988. Predatory Dinosaurs of the World. New York: Simon and Schuster. ———. 1997. Dinosaur models: The good, the bad, and using them to estimate the mass of dinosaurs. In D. Wolberg, E. Stump, and G. D. Rosenberg, eds., DinoFest International: Proceedings of a Symposium Held at Arizona State University, pp. 129–154. Philadelphia: Academy of Natural Sciences. Pennycuick, C. J. 1997. Actual and ‘‘optimum’’ flight speeds: Field data reassessed. Journal of Experimental Biology 200: 2355– 2361. Rasband, W. S. 2007. ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/. Retallack, G. I. 1990. Soils of the Past. Boston: Unwin-Hyman. Röll, B. 2001. Gecko vision-retinal organization, foveae and implications for binocular vision. Vision Research 41: 2043– 2056. Romer, A. S. 1956. Osteology of the Reptiles. Chicago: University of Chicago. Ross, C. F., M. I. Hall, and C. P. Heesy. 2007. Were basal primates nocturnal? Evidence from eye and orbit shape. In M. J. Ravosa and M. Dagosto, eds., Primate Origins: Adaptations and Evolution, pp. 233–256. New York: Springer. Ross, C. F., and E. C. Kirk. 2007. Evolution of eye size and shape in primates. Journal of Human Evolution 52: 294–313. Rossie, J. B., X. Ni, and K. C. Beard. 2006. Cranial remains of an Eocene tarsier. Proceedings of the Natural Academy of Sciences 103: 4381–4385. Russell, D. A., and R. Seguin. 1982. Reconstruction of the small Cretaceous theropod Stenonychosaurus inequalis and a hypothetical dinosauroid. Syllogeus 37: 1–43. Schmidt-Nielsen, K. 1964. Desert Animals. New York: Oxford University Press. Schmitz, L., R. Motani, and A. Milner. 2007. Diet activity pattern of Archaeopteryx. Journal of Vertebrate Paleontology 27(3, Suppl.): 142A. Schubert, B. W., and P. S. Ungar. 2005. Wear facets and enamel spalling in tyrannosaurid dinosaurs. Acta Palaeontologica Polonica 50: 93–99. Sieffert, E. R., E. L. Simons, W. C. Clyde, J. B. Rossie, Y. Attia, T. M.

Bown, P. Chatrath, and M. E. Mathison. 2005. Basal anthropoids from Egypt and the antiquity of Africa’s higher primate radiation. Science 310: 300–304. Simons, E. L. 1997. Preliminary description of the cranium of Proteopithecus sylviae, an Egyptian late Eocene anthropoidean primate. Proceedings of the National Academy of Sciences 94: 14970–14975. Simpson, G. G. 1925. A Mesozoic mammal skull from Mongolia. American Museum Novitates 201: 1–11. Stevens, K. A. 2006. Binocular vision in theropod dinosaurs. Journal of Vertebrate Paleontology 26: 321–330. Sullivan, C., and R. R. Reisz. 2005. Cranial anatomy and taxonomy of the Late Permian dicynodont Diictiodon. Annals of the Carnegie Museum 74: 45–75. Taylor, C. T. 1966. The vascularity and possible thermoregulatory function of the horns in goats. Physiological Zoology 39: 127– 139. Tereschenko, V. S., and V. R. Alifanov. 2003. Bainoceratops efremovi, a new protoceratopid dinosaur (Protoceratopidae, Neoceratopsia) from the Bain-Dzak locality (South Mongolia). Paleontological Journal 37: 71–80. Therrien, F., and D. M. Henderson. 2007. My theropod is bigger

than yours . . . or not: Estimating body size from skull length in theropods. Journal of Vertebrate Paleontology 27: 108–115. Varricchio, D. J., A. J. Martin, and Y. Katsura. 2006. First trace and body fossil evidence of a burrowing, denning dinosaur. Proceedings of the Royal Society B: 1–7. Vickers-Rich, P., and T. H. Rich. 1993. Australia’s polar dinosaurs. Scientific American 269: 50–55. Walls, G. L. 1942. The Vertebrate Eye and Its Adaptive Radiation. Bloomfield Hills, Mich.: Cranbrook Institute of Science. Walsberg, G. E. 2000. Small mammals in hot deserts: Some generalizations revisited. BioScience 50: 109–120. Warrant, E. J. 1999. Seeing better at night: Life style, eye design and the optimum strategy of spatial and temporal summation. Vision Research 39: 1611–1630. Warrant, E. J., and N. A. Locket. 2004. Vision in the deep sea. Biological Reviews 79: 671–712. Williams, J. B., I. Tieleman, and M. Shobrak. 1999. Lizard burrows provide thermal refugia for larks in the Arabian Desert. Condor 101: 714–717. You, H., and P. Dodson. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press.

The Function of Large Eyes in Protoceratops 327

23 A Semi-Aquatic Life Habit for Psittacosaurus TRACY L. FORD AND LARRY D. MARTIN

Psittacosaurus has been perceived historically as a ter-

fed in lakes or rivers, perhaps crawling in the mud in

restrial biped that lived in the seasonally dry alluvial to

search of aquatic plants.

desert paleoenvironments of early Cretaceous Asia. How-

Psittacosaurus is one of the most diverse and abundant

ever, because hundreds of specimens are now known

dinosaurs known. Its broad paleoenvironmental range

from lacustrine deposits in northeastern China, the pos-

(desert margin, seasonally dry alluvial plain, alluvial-

sibility that Psittacosaurus was a semi-aquatic dinosaur

lacustrine and volcano-lacustrine settings) is seen in

must be considered more seriously than it has been in

some other modern semi-aquatic animals including

the past.

agamid, scincid, and varanid lizards, as well as some

Our examination of specimens and review of the litera-

medium-size mammals such as castorids and caviids.

ture reveal that the morphology of Psittacosaurus is most parsimoniously interpreted as reflecting a semi-aquatic life habit. Psittacosaurus has been found exhibiting natu-

Introduction

ral resting and sprawling positions that suggest an ability

In the early twentieth century, sauropods were believed to

to lift the hindlimb in preparation for a swimmer’s

have been lumbering animals that buoyantly supported their

‘‘kick.’’ Other morphological features that support an in-

bulk by frequenting lakes (Matthew 1915; Mook 1918; Col-

terpretation for semi-aquatic life habits include (1) high

bert 1961, 1965). Likewise, hadrosaurs were believed to have

nares and orbits; (2) a forelimb motion that is consistent

frequented lakes and rivers to feed on aquatic vegetation

with a swimming stroke; (3) presence of muscle scars

(Matthew 1915; Colbert 1965). It wasn’t until the late 1960s

from strong flexor muscles on the back of the metatar-

and early 1970s—during the so-called Dinosaur Renaissance

sals; (4) presence of a long tail with deep chevrons;

(Bakker 1975; Olshevsky 1981)—that some of these interpre-

(5) ‘‘bristle-like’’ integumentary structure on the dorsal

tations were challenged. By the late 1970s a new orthodoxy

midline of the proximal tail that may have been covered

had been established that saw dinosaurs as highly active, but

by a layer of skin; and (6) gastroliths.

strictly terrestrial forms (Bakker 1986). Not only did the rein-

We propose that psittacosaurs could swim using a vari-

terpretation take sauropods and many hadrosaurs out of the

ety of techniques, including hindlimb kick, front limb re-

swamps, rivers and lakes, but it also showed that they weren’t

traction, and an anguilliform or undulating motion of

necessarily sluggish reptiles. In the last few decades, however,

the tail. As semi-aquatic animals, psittacosaurs may have

some other dinosaurs have been interpreted as semi-aquatic

328

forms, including the spinosaur, Baryonyx (Charig and Mil-

Psittacosaurs

ner 1986, 1997), some early Cretaceous North African sauropods (Lacovara et al. 2002), and even some ankylosaurs (Ford

Psittacosaurs are small marginocephalians with unique skele-

2002). Most recently, Tereschenko (2008), citing the morphol-

tal features (elevated nares and orbit, jugal protrusions, small

ogy of the vertebral column and, specifically, the caudal re-

horns/bumps on the top of the skull, and a four digit manus

gion, suggested that some protoceratopsians were capable of

with digit three being the longest [Sereno 1987]). Referral to

swimming, and were semi-aquatic, an idea that had been pre-

Ceratopsia is based on the rostral bone, long hindlimb, rela-

viously suggested by Brown and Schlaikjer (1940) and Bars-

tively short forelimbs and reduced manus (Sereno 1990). Most

bold (1974). It now seems safe to say that, indeed, some dino-

are small animals ranging in size from less than 1 m to just

saurs may have had a semi-aquatic mode of life during some

over 2 m.

part of their lives.

Psittacosaurus is the most abundant and probably the most

Recently, a Psittacosaurus specimen from Liaoning was found

speciose dinosaur known. It ranges in age from the Valangi-

with long ‘‘bristle-like’’ structures along the posterior midline

nian to Albian, and is known across central and eastern Asia

of the tail (Mayr et al. 2002; specimen SMF R 4970 [Forschungs-

from Xinjiang to Liaoning, and Siberia to Thailand (Dodson

institut Senckenberg, Frankfurt am Main, Germany]). The

1996; Lu et al. 2007; Horner pers. com. 2007). Ongoing dis-

structures have been interpreted as feather-like, quill-like, and

coveries from Liaoning and Mongolia continue to main-

ornamental in function. This specimen fired our imaginations

tain its numerical and taxonomic hegemony over other co-

and raised the possibility that psittacosaurs may have had

occurring dinosaurs. The deposits in Liaoning, China that

semi-aquatic life habits. Here we offer an alternative explana-

yield hundreds of psittacosaur skeletons were laid down in

tion for these ‘‘bristle-like’’ structures and, more importantly,

several large lake systems in a regional volcano-lacustrine set-

consider the possibility of a semi-aquatic, freshwater existence

ting that resulted from late Jurassic to early Cretaceous epi-

for some members of Psittacosaurus.

sodes of crustal rifting (Graham et al. 2001).

A semi-aquatic life habit interpretation for Psitttacosaurus is

The abundance, ubiquity and significant temporal range of

not a radical idea. Aside from some other dinosaurs now be-

Psittacosaurus in the early Cretaceous of Asia has resulted in its

ing regarded as semi-aquatic forms (see above), a semi-aquatic

use as a terrestrial bio-zone taxon (Lucas 2006) and, more im-

mode of life for Psittacosaurus has been postulated previously

portantly, identifies it as one of the best opportunities paleon-

by Rozhdestvensky (1955), Suslov (1983), Currie (1997), and

tologist have to thoroughly examine and understand the pa-

Averianov et al. (2006). These proposals, however, have been

leobiology of a non-avian dinosaur.

largely ignored, probably because psittacosaurs were historically associated with semi-arid and eolian paleoenvironments (Osborn 1923, 1924; Young 1958; Sereno 1990, 1997; Tereschenko 2008).

Previous Arguments in Support of Semi-Aquatic Psittacosaurs

Today, however, psittacosaurs are known from eolian, fluvio-

Rozhdestvensky (1955) states that the manual phalanges and

lacustrine, and volcano-lacustrine paleoenvironments (Rus-

unguals of psittacosaurs are dorsoventrally flattened and sug-

sell and Zhao 1996; Jerzykiewicz 2000; Brinkman et al. 2001;

gests that the living animal may have had webbing between

Qi et al. 2007), and it seems reasonable, therefore, that paleo-

its fingers, enabling it to swim. However, he also believed that

environmental associations alone are insufficient to support

because Psittacosaurus had sclerotic rings it was aquatic. This

an interpretation of psittacosaurs as obligate terrestrial forms.

latter argument fails spectacularly because sclerotic rings are

Institutional Abbreviations. AMNH: American Museum of

widely distributed within purely terrestrial vertebrates (nota-

Natural History, New York; BNHM: Beijing Natural History

bly birds) and are thus not exclusive to aquatic forms (see

Museum, Beijing; IG: Institute of Geology, Beijing; IVPP: In-

Longrich this volume).

stitute of Vertebrate Paleontology and Paleoanthropology,

Suslov, 1983, believed that psittacosaurs either lived near or

Beijing; JZMP: Jinzhou Museum of Paleontology, Jinzhou

in water because many specimens are found in lacustrine de-

City; PIN: Palaeontological Institute, Russian Academy of Sci-

posits of the Khamryn-Us locality in Mongolia. Currie (pers.

ences, Moscow; PKUP: Peking University Paleontological Col-

com. 1997) suggested that psittacosaurs may have employed

lections, Beijing; PM TGU: Paleontological Museum, Tomsk

their gastroliths as ballast in water rather than as a food pro-

State University, Tomsk; SMF: Forschungsinstitut und Natur-

cessing aid. In support of this interpretation he cited (1) the

museum Senckenberg, Frankfurt am Main; TF: Department of

association of psittacosaur fossils with lacustrine paleoenvi-

Mineral Resources, Bangkok; UGM: Urumqi Geological Mu-

ronments, (2) the presence of a relatively sophisticated dentary

seum, Urumqi; ZIN: Zoological Institute, Russian Academy of

battery and jaw mechanics, and (3) teeth with self-sharpening

Sciences, Saint Petersburg; ZMNH: Zhejiang Museum of Natu-

cutting edges, which would make food processing by gastro-

ral History, Hangzhou.

liths redundant. Averianov et al. (pers. com. 2006) agrees with

A Semi-Aquatic Life Habit for Psittacosaurus 329

FIGURE 23.1.

‘‘Resting’’ psittacosaurs showing the hyperflexed hind legs. (A) Psittacosaurus mongoliensis (AMNH 6245) modified from Osborn (1924) and in dorsal view; (B) right hind limb of Psittacosaurus xinjiangensis (UGM XG94Kh201) in dorsal and ventral views (after Brinkman et al. 2001); (C) left tarsus and pes of Psittacosaurus meileyingensis (IGV.330) in ventral view (after Sereno et al. 1988); (D) right tarsus of juvenile Psittacosaurus mongoliensis (AMNH 6536) in ventral view (after Coombs 1982). Scale bars are 10 cm (A), 2 cm (B, C), and 0.5 cm (D).

all preceding arguments and also supports the concept that psittacosaurs had a semi-aquatic mode of life.

formed after death and burial. In contrast, Faux and Padian (2007) determined that the skeletal posture of P. mongoliensis

As suggested by Suslov (1983), more psittacosaur specimens

is not consistent with opisthotony, but instead is a natural

are found in lacustrine deposits than alluvial or other non-

position. Similarly, Brinkman et al. (2001) described a referred

lacustrine settings. Specifically, most specimens are found as-

specimen of Psittacosaurus xinjiangensis that also exhibits

sociated with the sediments of three, large, Early Cretaceous

flexed hindlimbs and interpreted it as being in a resting posi-

lake systems. Two occur in northern China (Qingyang Lake

tion. We follow Brinkman et al. (2001) and propose that flexed

on the modern Ordos Plateau and a large central basin lake

hindlimbs were a natural resting position for psittacosaurs.

system in the Junggar Basin [Chen 1987; Averianov et al. 2006]), and the third occurs at Khamryn-Us, Mongolia (Suslov

SPRAWLING

1983). More recently many tens of specimens have been recovered from the volcano-lacustrine lake deposits in western

The preserved posture of the second psittacosaur specimen

Liaoning (e.g., Qi et al. 2007).

ever described, (= Protiguanodon mongoliensis Osborn 1923), included sprawling hind legs and outstretched arms (Fig. 23.2). Although Senter (2007) believes the arms are dislocated

Functional Anatomy Inferred from Articulated Psittacosaurs: Resting and Sprawling

from the glenoid, other psittacosaurs are found in a similar

RESTING

ferred IVPP. V. 739), P. xinjiangensis from Delunshan (Sereno

sprawling position, including P. sinensis (Young 1958) from the Laiyang beds of Shantung (IVPP V. 738 holotype, and reand Chao 1988) and numerous specimens from Liaoning seen

Terrestrial vertebrates that retain in-life resting postures as fos-

by the second author. Recently, an adult Psittacosaurus sp. was

sils are rare, but they provide powerful paleobiological in-

found with 34 articulated baby skeletons (Meng et al. 2004).

sights (e.g., Xu and Norell 2004; Jerzykiewicz et al. 1993; Fas-

Many of the babies have naturally sprawling humeri and

tovsky et al. 1994, 1997; Longrich this volume). The type

femora and flexed distal limb portions, suggesting that a

species of Psittacosaurus, P. mongoliensis (Osborn 1924), was

sprawling proximal limb posture was natural for this taxon.

found with its arms drawn to its sides, elbow slightly bent

Given that lizards, salamanders and crocodilians employ

with the palms up, and its hind legs flexed and abducted with

‘‘splayed’’ hind legs to assist in locomoting in aquatic settings,

its knee and ankle joints hyperflexed (Fig. 23.1; Sereno et al.

we propose that this preserved splayed posture is at least sug-

1988). Coombs (1982) believed that this posture was not re-

gestive of these psittacosaurs being able to splay their legs

flective of true resting behavior in Psittacosaurus and that it

in life.

330 ford & martin

FIGURE 23.2.

‘‘Sprawling’’ psittacosaurs. (A) Psittacosaurus mongoliensis (AMNH 6253) modified from Osborn (1924) and shown in dorsal view; (B) Psittacosaurus sinensis (IVPP V738) modified from Young (1958) and shown in dorsal view; (C) Psittacosaurus sinensis (IVPP V739) in dorsal view (after Young 1958); (D) juvenile Psittacosaurus sp. (Dalian Natural History Museum, D 2156) in dorsal view (after Meng et al. 2004); (E) Psittacosaurus xinjiangensis (IVPP V7698) in dorsal view (after Sereno and Chao 1988); (F) Psittacosaurus sibiricus (PM TGU 16/4-20) from Averianov (pers. com.). Scale bars are 10 cm (A, F) and 5 cm (B–E).

Anatomical Evidence for Semi-Aquatic Habits HINDLIMBS

cartilage cap. P. xinjiangensis (UGM XG94Kh201; Brinkman et al. 2001) show this in great detail. Ornithopods with a medially directed femoral head has the cartilage just on the head, whereas Psittacosaurus has the cartilage on the head and on

Because pittacosaurs have much larger hindlimbs than fore-

the dorsal edge of the femur. This arrangement would enable

limbs they are believed to have been primarily bipedal (Os-

the femur to be raised dorsally via muscles and ligaments into

born 1923; Maryanska ´ and Osmólska,1975; Sereno 1990;

a sprawling position.

Chinnery 2004). The femora of ornithopods, both large and

The knee had great mobility and the mesotarsal joint, with

small (Iguanodon and Hypsilophodon), has a large medially di-

the hyperflexibility of the ankle, permitted the entire pes to be

rected femoral head positioned at a right angle to its shaft (Fig.

placed on the ground when resting (Coombs 1982). Coombs

23.3; Norman 1980; Galton 1974), a morphology that is typi-

(1982) believed that the great range of movement at the hip,

cal for bipedal dinosaurs, in general. In these forms the proxi-

knee and ankle was due to small size. However, the alternative

mal end of the femur has the greater trochanter separated

interpretation favored here is that the great mobility of the

from the femoral head (slightly in Hypsilophodon and greatly

knee and hyperflexiblity of the ankle increased the ability of

in Iguanodon). The positions of the femoral head and greater

the animals to swim, giving it not only a greater range of mo-

trochanter both inhibited sprawling of the proximal portion

tion, but also increasing the force and stroke needed to swim

of the hindlimb. In Psittacosaurus the greater trochanter is

with the feet.

continuous with the head and along the posterior edge of the femur (Sereno 1987; Averianov et al. 2006). The shaft is bowed and has a pendant fourth trochanter on its proximal half, and

PES

a finger-like lesser trochanter that extends to the dorsal edge

As in all other bipedal dinosaurs, the pes functioned in a digi-

of the greater trochanter. the femoral head is small, though

tigrade fashion when the animal was walking. The shafts

P. sibericus (Averianov et al. 2006) has a larger medially di-

of the metatarsals were tightly bound together by ligaments

rected femoral head. Most psittacosaurs have a unique fem-

(Sereno 1987). There are 4 digits with digit I being the shortest

oral head that is set at a 30\ angle to the shaft with the head

(Fig. 23.4) out of contact with the substrate when walking

directed dorsally and assumed to have been covered with a

(Rozhdestvensky 1955). On the posterior side of the proximal

A Semi-Aquatic Life Habit for Psittacosaurus 331

FIGURE 23.3.

Ornithopod femora. (A) Right femur of Psittacosaurus xinjiangensis (UGM XG94Kh201) modified from Brinkman et al. (2001); (B) left femur of Psittacosaurus sibiricus (PM TGU 16/1-271) after Averianov et al. (2006); (C) Hypsilophodon foxii (BMNH R5830) after Galton (1974); (D) Iguanodon bernissartensis (IRSNB 1534) after Norman (1980); (E) proximal right femur of Psittacosaurus xinjiangensis (IVPP f.n. 64047) after Sereno and Chao (1988); (F) Psittacosaurus xinjiangensis (UGM XG94Kh201). Scale bars are 5 cm (A, F), 10 cm (B, C), 20 cm (D), and 2 cm (E).

shafts of the metatarsals 1–4 are large attachment scars, which

swimming animals such as penguins, seals and dolphins (Fig.

provide evidence for strong flexor muscles (Sereno 1987). This

23.6), though compared to those animals the longest digit is

suggests that the foot muscles were robust (and strong) and

toward the outside of the hand. In all of these forms the meta-

may have been used for more than just walking. Strong foot

carpals are parallel and positioned very close to one another

muscles would be ideal for digging or swimming, and in the

other. The phalanges are short and stocky. Accordingly, the

latter case, may have allowed for a strong kick. The toes are

manus of psittacosaurs may have been held together by thick-

longer than those of other ornithopods of the same size. The

ened skin (e.g., sea turtles).

pedal phalanges are flattened dorsoventrally. The pedal un-

Humerus. The humerus is large and the deltopectoral crest

guals are claw-like, long, narrow and slightly flattened dorso-

extends far laterally. Chinnery (2004) associates this condi-

ventrally (Rozhdestvensky 1955; Averianov et al. 2006).

tion in neoceratopsians with a partially sprawling posture. The glenoid cavity faces ventrally and slightly laterally when

FORELIMB

the scapula is horizontal (Sereno 1987). The deltopectoral crest on the humerus is robust and, together with the large

Manus. The manus is unique in psittacosaurs. It is asymmetri-

coracoid/proximal scapula, we infer the presence of relatively

cal in dorsal view with digit III being the longest and, overall,

large deltoid/pectoralis muscles in living adults that may have

more robust than in other ceratopsians (Fig. 23.5; Chinnery

been used in a swimming stroke.

2004). As in the pes, the metacarpals, phalanges and unguals

Ulna/Radius. The ulna and radius do not cross one another

are flattened dorsoventrally. In support of Rozhdestvensky’s

in Psittacosaurus (Senter 2007). The proximal and distal articu-

(1955) suggestion that the hand may have been webbed based

lations lack rolling surfaces that would allow for active pro-

on the presence of the flattened fingers, the short, fourth digit

nation and supination. In their natural position the palms

would have been useless in terrestrial locomotion (Senter

face submedially, another morphological feature that would

2007) or for grasping as suggested by Sereno (1990).

aid a swimming stroke.

Further evidence appears to argue against use of the manus for walking or grasping, but supports the possibility that it was used for swimming. The metacarpals have a distal ridge that

FORELIMB MOVEMENT

prevents digital hyperextension preventing their effective use

Senter (2007) assessed forelimb function in basal ceratopians

for walking (Senter 2007). During flexion, the first digit con-

and examined several basal ceratopians including Psittaco-

verged toward the centre of the palm. Thus, at rest it seems

saurus neimongoliensis, and P. mongoliensis (AMNH 6253, 6354,

that the palms of Psittacosaurus faced medially. If the manus

21752). He concluded that the range of motion of the shoul-

was webbed and used for swimming, flexion of the first digit

der of Psittacosaurus consists of a wide arc of parasagittal

may have resulted in folding of the web during the return

movement, but protraction was limited to not much more

stroke.

than a horizontal position (Fig. 23.7). The transverse move-

Overall, the manus shows some similarities with those of

332 ford & martin

ment of the humerus allows it to be raised to the subhorizon-

while P. sinensis has the shortest hind legs, though some of the Liaoning specimens that haven’t been described in detail apparently have even shorter hind legs. We can only speculate at this time that this difference may reflect more terrestrial habits in forms with relatively long hindlimbs (P. mongolieneis) versus more aquatic habits in forms with relatively shorter hindlimbs (P. sinensis and other taxa from Liaoning). Body Shape. In contrast to most ornithischians (e.g., Hypsilophodon and Iguanodon) the anterior to middle trunk of Psittacosaurus was neither deep, nor particularly wide. The rib cage was narrow and the ribs extended ventrally no farther than the extent seen in the ischium (Fig. 23.8). We suggest Pes of psittacosaurs. (A) Psittacosaurus neimongoliensis (IVPP 12-0888-2) after Russell and Zhao (1996); (B) Psittacosaurus sibiricus (PM TGU 16/1-200) after Averianov (2006); (C) dorsal view of Psittacosaurus mongoliensis (AMNH 6253) modified from Osborn (1924). Scale bars are 2 cm. FIGURE 23.4.

that these proportions resulted in a crudely fusiform body shape, similar to that of a crocodile, and may have helped in reducing drag while swimming.

GASTROLITHS Gastroliths have been reported from five psittacosaur specimens: 63 in AMNH 6544 (P. mongoliensis); more than 50 in AMNH 6253 (Sereno 1987); 36 in IVPP V 12165 (P. mazongshanensis; Xu 1997); numerous (number not given) in SMF R 4970 (P. sp; Mayr et al. 2002); numerous (number not given, nor accession number; P. sp.; Behrendt 2006). Several specimens collected recently from Liaoning also contain gastroliths (LMM pers. obs.). Gastroliths have two possible functions: (1) to aid in digestion; or (2) as ballast in aquatic animals. Gastroliths have been found in all the major groups of dinosaurs including avian and non-avian theropods (Lourinhanosaurus Mateus 1998),

Psittacosaurus mani. (A) Psittacosaurus mongoliensis (AMNH 6254) after Osborn (1924); (B) Psittacosaurus sibiricus (PM TGU 16/1-201) after Averianov (2006); (C) Psittacosaurus neimongoliensis (IVPP 12-0888-2) after Russell and Zhao (1996). Scale bars are 2 cm. FIGURE 23.5.

Sauropodomorpha (Stokes 1964; Sanders et al. 2001), and Ornithischia (Novas et al. 2004; Xu et al. 2006). They are also well known in other vertebrates including crocodylomorphs, birds, ichthyosaurs, sauropterygians (Taylor 1993). Psittacosaurus is generally believed to have used gastroliths for digestion, though this was challenged by Currie (1997) who pointed out that because the teeth have self-sharpening cut-

tal position. The elbow could not achieve full extension or

ting edges, these animals may have used their gastroliths

flexion and therefore could not be used in walking. Simi-

for ballast. With the exception of P. mazongshanensis, self-

larly, Senter’s results also do not support the use of the fore-

sharpening teeth appear to be a ubiquitous feature of psit-

limbs in digging or food gathering. Senter’s explanation for

tacosaurs. Following Currie (1997), we propose that in most

the unusual range of motion in the two species of Psittaco-

psittacosaurs, gastroliths were likely used as ballast when

saurus that he examined was that it was used for clutching

these animals were in water.

objects to the body. Alternatively, we propose that the range of motion that he documented is best interpreted as a swimming stroke.

TAIL The tail of Psittacosaurus is well preserved in many specimens.

BODY PROPORTIONS

The most complete specimen (AMNH 6254) has 43 caudal vertebrae. The tail is ‘‘long’’ and the anterior caudal centra are

Front versus Hindlimbs. The relative length of the forelimb

taller than long, whereas the posterior centra are longer than

versus the hindlimb varies among psittacosaur species (Table

tall (Sereno 1987). The anterior caudal transverse processes are

23.1). Psittacosaurus mongoliensis has the longest hind legs,

large and similar in shape to those of a crocodile. Tall chevrons

A Semi-Aquatic Life Habit for Psittacosaurus 333

FIGURE 23.6.

Generalized psittacosaur front limb (A) compared to front limbs of known swimming vertebrates. (B) dolphin; (C) sea lion; (D) sea turtle; (E) penguin (B–E after Hildebrand 1974). Not to scale.

The anteriormost caudals have zygapophysial articulations that are essentially vertical. According to Sereno (1987), such articulations would have restricted the lateral motion of the tail. However, this interpretation contrasts sharply with the presence of laterally curled tails in many articulated specimens (LMM pers. obs.).

NARES AND ORBITS The dorsally high position of the nares and orbits also support an interpretation for semi-aquatic life habits for some psittacosaurs. High nares and orbits are typical for many modern aquatic and semi-aquatic animals, allowing them to have their eyes and nose above the water while their bodies are mostly submerged (e.g., crocodilians, Hippopotamus, capybara).

SKIN Skin impressions often provide insights into what a fossil vertebrate may have looked like in life. Many psittacosaurus speciProposed range of movement (indicated by arcs) in the left front limb of Psittacosaurus neimongoliensis. (A) Cranial view; (B, D) lateral view. Proposed range of natural movement for right femur (C). All after Senter 2007. Scale bar is 10 cm. FIGURE 23.7.

mens from Liaoning are preserved with skin impressions ( Ji and Bo 1998; Mayr et al. 2002; Ji 2004; Lingham-Solair 2008). The left hindlimb of SMF R 4970 (Mayr et al. 2002) is flexed and has skin impressions associated with the feet and leg. These give the impression that the soft tissue was thick along the posterior part of the hind leg (from the ankle to the ilium). In our view, this may have increased the surface area when

are present making the tail quite deep. From our perspective, these features would have provided the tail with larger areas for muscle attachment—again, as in crocodilians.

using the hindlimbs for swimming. Based on specimen MV53 (Lingham-Solair 2008), the skin of Psittacosaurus was thick and strong. This specimen has

The neural spines are proportionately tall in all species and

soft tissues impressions along its ventrolateral region, and a

are particularly tall in P. sinensis. In P. mongoliensis and P. sinen-

cross section through the skin shows that it had seven layers

sis distal neural spines are flattened side-to-side, and fan-

of collagen fibers. Although Lingham-Solair (2008) believes

shaped (antero-distally). Thus, the tail may have been laterally

that this thick skin may have been a barrier against attack, it

compressed, which would help in swimming as in some mod-

also may have helped to strengthen the limbs and tail for

ern lizards (Bauer and Jackman 2008).

swimming.

334 ford & martin

Table 23.1. Sizes (in mm) and Ratios of Hindlimbs versus Forelimbs in Select Specimens of Psittacosaurus Psittacosaurus specimen P. mongoliensis,

Hindlimb/ Humerus

Radius

Femur

Tibia

forelimb

119

90

162

179

1.3

126

90

158

167

1.5

105

69

144

141

1.6

AMNH 6254 P. mongoliensis

and feet could have been used to push backward and swim. The front legs had a great range of lateral movement and could have been used in maneuvering and swimming, similar to some lift-based labriform fish, wrasses, parrotfishes, sturgeons, and chimeras (Webb 1973; Sfakiotakis et al. 1999; Helfman et al. 1997). The ulna and radius did not cross over and could not be used in pronation and supination, thus also aid-

AMNH 6253 P. neimongoliensis

aquatic lizards and mammals. It is also likely that the lower leg

IVPP

ing swimming. The manus and pes have dorsoventrally flattened phalanges and unguals adding surface area for paddling/kicking. The manus may have been webbed as suggested by Rozhdestven-

12-0888-2 P. major LH PV1

149

81

172

197

1.6

P. sp. JZMP-V-11

145

95

165

185

1.4

sky (1955). The small digit I could have been folded reducing drag in the return stroke, important for a swimming animal (Hildebrand, 1974). The strong foot muscles could have help given a stronger kick while swimming. The long flattened toes may have also added in walking on soft slippery substrate

BRISTLE-LIKE STRUCTURES

(Rozhdestvensky 1955). We propose that Psittacosaurus would have used its tail as a

Mayr et al. (2002) described the presence of approximately 100

source of propulsion and maneuvering, similar to crocodil-

bristle-like integument structures on the tail of SMF R 4970

ians. Animals with long bodies and tails usually swim by un-

(Fig. 23.9). The structures are associated only with the proxi-

dulating (Fish 1984) or in an anguiliform manner (eels, sirens,

mal and midline portion of the tail, and they terminate prox-

snakes, crocodilians; Gillis 1996; Sfakiotakis et al. 1999). The

imally at the vertebrae. Thus, they appear to have been buried

large transverse processes on the caudal vertebrae and large

deep in the skin and firmly attached to the tail along its dorsal

chevrons of psittacosaurs would have allowed a large contact

midline. Because they were so deeply anchored in the skin and

area for the caudofemoralis muscle whose side to side contrac-

there is some evidence that these bristles were keratinzed, it is

tions could have provided significant propulsion. The deeply

likely that these structures would have been quite stiff. The

imbedded bristle-like integumentary structures along the top

structures lack a shaft and were interpreted as being feather-

of the tail (if present in other psittacosaurs) may have sup-

like stage one by Xu (2006). Mayr (et al. 2002) interpreted

ported a tall tail fin, similar to that of a salamander.

these structures as more tubular and stiffer than the ‘‘filaments’’ found in Liaoning theropods (e.g., Sinosauropteryx), and speculated that they may have been used for display.

Conclusions

We suggest that these bristles may have supported a caudal

Psittacosaurus has been historically perceived as a terrestrial

fin that was somewhat analogous to the caudal fin in modern

biped living in seasonally dry to desert paleoenvironments.

amphibians, such as the Hellbender (Cryptobranchus; Reese

However, because hundreds of specimens are now known

1906), the Mexican axolotl (Sirenodon; Willemse 1977), sala-

from lacustrine deposits in northeastern China, the possibil-

manders (Azizi 2005), and tadpoles (Doherty et al. 1998). In

ity that Psittacosaurus was a semi-aquatic dinosaur must be

contrast to the flexible collagenous tails of amphibians, how-

considered more seriously than it has in the past.

ever, the relatively large size of psittacosaurs may have se-

Our examination of specimens and a review of the literature

lected for a stiffer structure in order to support the large tail

show that Psittacosaurus has anatomical features that, to-

and enable its use in swimming.

gether, are most parsimoniously interpreted as reflecting semiaquatic life habits for this taxon. Psittacosaurus has been found

The Nature of Psittacosaur Swimming

exhibiting natural resting and sprawling positions. In particular, the frequently encountered sprawling posture suggests

In our view, most psittacosaurs could have used both their

that these animals may have been able to lift their hind legs in

limbs and tail to swim. A femoral head at the dorsal edge

preparation for a ‘‘swimmer’s kick.’’ Other anatomical features

of the femur having a cartilaginous cap may have enabled

that support semi-aquatic life habits include (1) high nares and

the femur to have been raised from vertical for bipedality

orbits; (2) a forelimb motion that is consistent with a swim-

to lateral/horizontal (or sprawling) for swimming. With its

ming stroke; (3) hindlimbs with a large range of movement

sprawling hindlimbs, a backward-oriented kicking motion

including the ability to be lifted into the horizontal plane;

could have been used for propulsion, similar to some semi-

(4) presence of muscle scars from strong flexor muscles on the

A Semi-Aquatic Life Habit for Psittacosaurus 335

Psittacosaurus skeletons reconstructed in left lateral view and all to the same scale. (A) Psittacosaurus mongoliensis (AMNH 6254) modified from Osborn (1924); (B) Psittacosaurus mongoliensis (AMNH 6253) modified from Osborn (1924); (C) Psittacosaurus sinensis (IVPP V738) modified from Young (1958); (D) Psittacosaurus neimongoliensis (IVPP 12-0888-2) after Russell and Zhao (1996); (E) Psittacosaurus sp. (SMF R 4970) modified from Mayr et al. 2002; (F) Psittacosaurus xinjiangensis (IVPP V7698) after Sereno and Chao (1988); (G) Psittacosaurus sinensis (BNHM PPV.149) modified from Chao (1963); (H) Psittacosaurus major (LH PV1) after Sereno et al. (2007); (I) Psittacosaurus sibiricus (PM TGU 16/4-20) from Averianov (pers. com.). Scale bars are 10 cm. FIGURE 23.8.

back of the metatarsals; (5) presence of a long tail with deep

or rivers, perhaps crawling in the mud in search of aquatic

chevrons; (6) ‘‘bristle-like’’ integumentary structures that may

plants (A. O. Averianov pers. com. 2008). However, a variety of

have been covered by a layer of skin; and (7) gastroliths.

forelimb to hindlimb relative lengths suggest that some psit-

We propose that pisttacosaurs swam using a variety of techniques including a hindlimb kick, paddling using the front limbs, and an anguilliform or undulating motion of the tail. As semi-aquatic animals psittacosaurs may have fed in lakes

336 ford & martin

tacosaurs were likely more terrestrial than others (e.g., Psittacosaurus mongoliensis). Psittacosaurus is one of the most diverse and abundant dinosaurs known. It inhabited a wide variety of environments in-

FIGURE 23.9.

Bristle-like structures of Psittacosaurus sp. (SMF R 4970). (A) Caudal vertebrae, scales and bristles; (B) detail of the proximal area of the bristles; (C) skeleton in matrix. All after Mayr et al. (2002). Scale bars are 2.5 cm (A), 0.5 cm (B), and 10 cm (C).

cluding desert margins, and alluvial to alluvial-lacustrine settings. This environmental range is similar for agamid, scincid, and varanid lizards (Bauer and Jackman 2008), as well as some medium size mammals such as castorids (Martin 1987) and caviids. Acknowledgments

We thank Alexander O. Averianov, Phil Currie, David Burnham, Ralph Molnar, and David Eberth for valuable suggestions and editorial assistance. References Cited Averianov, A., A. V. Voronkevich, S. V. Leshchinskiy, and A. V. Fayngertz. 2006. A ceratopsian dinosaur Psittacosaurus sibiricus from the Early Cretaceous of West Siberia, Russia and its phylogenetic relationships. Journal of Systematic Palaeontology 4: 359–395. Azizi, E. 2005. Biomechanics of salamander locomotion. Ph.D. diss., University of Massachusets, Amherst. Bakker, R. T. 1975. Dinosaur Renaissance. Scientific American 232(4): 58–78. ———. 1986. The Dinosaur Heresies, New Theories Unlocking the Mystery of the Dinosaurs and their Extinction. New York: William Morrow and Co. Barsbold, R. 1974. Duelling dinosaurs. Priroda 2: 81–83. [In Russian.] Bauer, A. M., and T. Jackman. 2008. Global diversity of lizards in freshwater (Reptilia: Lacertilia). Hydrobiologia 595: 581–586. Behrendt, M. 2006. Feature fossil (Psittacosaurus sp). Fossil News, Journal of Avocational Paleontology 12: 1–2. Brinkman, D. B., D. A. Eberth, M. J. Ryan, and P.-J. Chen. 2001. The occurrence of Psittacosaurus xinjiangensis Sereno and Chow, 1988 in the Urho area, Junggar Basin, Xinjiang, People’s Republic of China. Canadian Journal of Earth Sciences 38: 1781– 1786. Brown, B. B., and E. M. Schlaikjer. 1940. The structure and relationships of Protoceratops. Annals of the New York Academy of Science 40(3): 133–266.

Chao, S. T. 1963. New Species of Psittacosaurus from Laiyang, Shantung. Vertebrata PalAsiatica 6: 349–360. Charig, A. J., and A. C. Milner. 1986. Baryonyx, a remarkable new theropod dinosaur. Nature 324: 359–361. ———. 1997. Baryonyx walkeri, a fish-eating dinosaur from the Wealden of Surrey. Bulletin of The Natural History Museum, Geology Series 53: 11–70. Chen, P.-J. 1987. Cretaceous paleogeography in China. Palaeogeography, Palaeoclimatology, Palaeoecology 59: 49–56. Chinnery, B. J. 2004. Morphometric analysis of evolutionary trends in the ceratopsian postcranial skeleton. Journal of Vertebrate Paleontology 24: 591–609. Colbert, E. H. 1961. Dinosaurs: Their Discovery and Their World. New York: E. P. Dutton and Co. ———. 1965. The Age of Reptiles. New York: W. W. Norton & Co. Coombs, W. P. 1982. Juvenile specimens of the Ornithischian Dinosaur Psittacosaurus. Palaeontology 25: 89–107. Currie, P. J. 1997. Gastroliths. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, p. 270. San Diego: Academic Press. Dodson, P. 1996. The Horned Dinosaurs. Princeton: Princeton University Press. Doherty, P. A., R. J. Wassersug, and J. M. Lee. 1998. Mechanical properties of the tadpole tail fin. Journal of Experimental Biology 201: 2691–2699. Fastovsky, D. E., D. Badamgarav, H. Ishimoto, M. Watabe, and D. B. Weishampel. 1994. Paleoenvironments of Tugrikin-Shire (Late Cretaceous: Mongolia), and Protoceratops (Dinosauria: Ornithischia). Journal of Vertebrate Paleontology 14(3, Suppl.): 24A. ———. 1997. The Paleoenvironments of Tugrikin-Shireh (Gobi Desert, Mongolia) and aspects of the taphonomy and paleoecology of Protoceratops (Dinosauria: Ornithischia). Palaios 12: 59–70. Faux, C. M., and K. Padian. 2007. The opisthotonic posture of vertebrate skeletons: Postmortem contraction or death throes? Paleobiology 33: 201–226. Fish, F. E. 1984. Kinematics of undulatory swimming in the American Alligator. Copeia 1984: 839–843. Ford, T. L. 2002. Marine dinosaurs? Why are there so many

A Semi-Aquatic Life Habit for Psittacosaurus 337

hadrosaurids and ankylosaurids found in marine deposits? Mesa Southwest Museum Symposium: unnumbered abstract. Galton, P. M. 1974. The Ornithischian dinosaur Hypsilophodon from the Wealden of the Isle of Wight. Bulletin of the British Museum (Natural History), Geological Series 25: 3–152. Gillis, G. B. 1996. Undulatory locomotion in elongate aquatic vertebrates: Anguilliform swimming since Sir James Gray. American Zoologist 36: 656–665. Graham, S. A., M. S. Hendrix, C. L. Johnson, D. Badamgarav, G. Badarch, J. Amory, M. Porter, R. Barsbold, L. E. Webb, and B. R. Hacker. 2001. Sedimentary record and tectonic implications of Mesozoic rifting in southeast Mongolia. Geological Society of America Bulletin 113: 1560–1579. Helfman, G. S., B. B. Collette, and D. E. Facey. 1997. The Diversity of Fishes. Malden, Mass.: Blackwell. Hildebrand, M. 1974. Analysis of Vertebrate Structure. New York: John Wiley & Sons. Jerzykiewicz, T. 2000. Lithostratigraphy and sedimentary settings of the Cretaceous dinosaur beds of Mongolia. In M. J. Benton, M. A. Shishkin, D. Unwin, and E. N. Kurochkin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. 279–296. Cambridge: Cambridge University Press. Jerzykiewicz, T., P. J. Currie, D. A. Eberth, P. A. Johnston, E. H. Koster, and J.-J. Zheng. 1993. Djadokhta Formation correlative strata in Chinese Inner Mongolia: An overview of the stratigraphy, sedimentary geology, and paleontology and comparisons with the type locality in the pre-Altai Gobi. Canadian Journal of Earth Sciences 30: 2180–2195. Ji, S. 2004. Preliminary report on the dinosaurian skin impressions from the Yixian Formation of Lingyuan, Liaoning. Geological Review 50: 170–174. Ji, S.-A., and Bo, H.-C. 1998. Discovery of the psittacosaurid skin impressions and its significance. Geological Review 44: 603– 606. Lacovara, K. J., M. Lamanna, J. B. Smith, B. Grandstaff, and J. Smith. 2002. Concentration and preservation potential of vertebrate fossils within coastal lithosomes: Examples from the Upper Cretaceous, Bahariya Formation in Egypt. Journal of Vertebrate Paleontology 22(3, Suppl.): 77A. Lingham-Soliar, T. 2008. A unique cross section through the skin of the dinosaur Psittacosaurus from China showing a complex fibre architecture. Proceedings of the Royal Society of London, Series B 275: 775–780. Longrich, N. 2010. The function of large eyes in Protoceratops: A nocturnal ceratopsian? In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 308–327. Bloomington: Indiana University Press. Lu, J., Y. Kobayashi, Y.-N. Lee, and Q. Ji. 2007. A new Psittacosaurus (Dinosauria: Ceratopsia) specimen from the Yixian Formation of western Liaoning, China: The first pathological psittacosaurid. Cretaceous Research 28: 272–276. Lucas, S. G. 2006. The Psittacosaurus biochron, Early Cretaceous of Asia. Cretaceous Research 27: 189–198. Martin, L. D. 1987. Beavers from the Harrison Formation (Early Miocene) with a revision of Euhapsis. Dakoterra 3: 73–91.

338 ford & martin

Maryanaska, ´ T., and H. Osmólska. 1975. Protoceratopsidae (Dinosauria) of Asia. Palaeontologica Polonica 33: 135–181. Mateus, O. 1998. Lourinhanosaurus antunesi, a new Upper Jurassic Allosauroid (Dinosauria: Theropoda) from Lourinha, Portugal. Memorias da Academia de Ciencias de Lisboa 37: 111–124. Matthew, D. 1915. Dinosaurs, With Special Reference to the American Museum of Natural History. New York: American Museum of Natural History. Mayr, G., D. S. Peters, G. Plodowski, and O. Vogal. 2002. Bristlelike integumentary structures at the tail of the horned dinosaur Psittacosaurus. Naturwissenschaften 89: 361–365. Meng, Q., J. Liu, D. J. Varricchio, T. Huang, and C. Gao. 2004. Parental care in an ornithischian dinosaur. Nature 431: 145– 146. Mook, C. C. 1918. The habitat of the sauropodous dinosaurs. Journal of Geology 26: 459–470. Norman, D. B. 1980. On the ornithischian dinosaur Iguanodon bernissartensis of Bernissart (Belgium). Institut royal des Sciences Naturelles de Belgique, Memore 178: 1–103. Novas, F. E., A. V. Cambiaso, and A. Ambrosio. 2004. A new basal iguanodontian (Dinosauria, Ornithischia) from the Upper Cretaceous of Patagonia. Ameghiniana 41: 75–82. Olshevsky, G. 1981. Dinosaur Renaissance. Science Digest 89(7): 34–43. Osborn, H. F. 1923. Two Lower Cretaceous Dinosaurs of Mongolia. American Museum Novitiates 95: 1–10. ———. 1924, Psittacosaurus and Protiguandon: Two Lower Cretaceous iguanodonts from Mongolia. American Museum Novitates 127: 1–15. Qi, Z., P. M. Barrett, and D. A. Eberth. 2007. Social behaviour and mass mortality in the basal ceratopsian dinosaur Psittacosaurus (Early Cretaceous, People’s Republic of China). Palaeontology 50: 1023–1029. Reese, A. M. 1906. Anatomy of Cryptobranchus allegheniensis. American Naturalist 166: 287–326. Rozhdestvensky, A. K. 1955. New data on PsittacosaurusCretaceous ornithopods. Questions on the Geology of Asia 2: 783–788. Russell, D. A., and X.-J. Zhao. 1996. New psittacosaur occurrences in Inner Mongolia. Canadian Journal of Earth Science 33: 637–648. Sanders, F. H., K. Manley, and K. Carpenter. 2001. Gastroliths from the Lower Cretaceous sauropod Cedarosaurus weiskopfae. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 166–180. Bloomington: Indiana University Press. Senter, P. 2007. Analysis of fore limb function in basal ceratopsians. Journal of Zoology 273: 305–314. Sereno, P. C. 1987. The ornithischian dinosaur Psittacosaurus from the Lower Cretaceous of Asia and the relationships of the Ceratopsia. Ph.D. diss., Columbia University, New York. ———. 1990. Psittacosauridae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 579–592. Berkeley: University of California Press. ———. 1997. Psittacosauridae. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, pp. 611–613. San Diego: Academic Press.

Sereno, P. C., and S. Chao. 1988. Psittacosaurus xinjiangensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northwestern China. Journal of Vertebrate Paleontology 8: 353–365. Sereno, P. C., S. Chao, Z. Cheng, and C. Rao. 1988. Psittacosaurus meileyingensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of Northeastern China. Journal of Vertebrate Paleontology 8: 366–377. Sereno, P. C., X. Zhao, L. Brown, and L. Tan. 2007. New psittacosaurid highlights skull enlargement in horned dinosaurs. Acta Palaeontologica Polonica 58: 275–284. Sfakiotakis, M., D. M. Lane, and B. C. Davies. 1999. Review of fish swimming modes for aquatic locomotion. Journal of Oceanic Engineering 24: 237–252. Stokes, W. L. 1964. Fossilized stomach contents of a sauropod dinosaur. Science 143: 576–577. Suslov, J. V. 1983. The locality of Psittacosaurus in Khamrin-us (East Gobi, MPR). Transactions of the Joint Soviet Mongolian Paleontological Expedition 24: 118–121. Taylor, M. A. 1993. Stomach stones for feeding or buoyancy? The occurrence and function of gastroliths in marine tetrapods. Philosophical Transactions of the Royal Society of London Series B 341: 163–175. Tereschenko, V. S. 2008. Adaptive features of protoceratopoids

(Ornithischia: Neoceratopsia). Palaeontological Journal 42(3): 273–286. Webb, P. W. 1973. Kinematics of pectoral fin propulsion in Cymalogaster aggregate. Journal of Experimental Biology 59: 697–710. Willemse J. J. 1977. Morphological and functional aspects of the arrangement of connective tissue and muscle fibres in the tail of the Mexican axolotl, Siredon mexicanum (Shaw) (Amphibia, Urodela). Acta Anatomica 97: 266–85. Xu, X. 1997. A New Psittacosaur (Psittacosaurus mazongshanensis sp. nov. ) from Mazongshan Area, Gansu Province, China. In Z.-M. Dong, ed., Sino-Japanese Silk Road Dinosaur Expedition, pp. 48–67. Beijing: China Ocean Press. ———. 2006. Feathered dinosaurs from China and the evolution of major avian characters. Integrative Zoology 1: 4–11. Xu, X., C. A. Forster, J. M. Clark, and J. Mo. 2006. A basal ceratopsian with transitional features from the Late Jurassic of northwestern China. Proceedings of the Royal Society of London, Series B 273: 2135–2140. Xu, X., and M. A. Norell. 2004. A new troodontid dinosaur from China with avian-like sleeping posture. Nature 431: 838–841. Young, C.-C. 1958. The Dinosaurian remains of Laiyang, Shantung. Palaeontologia Sincia, new series C, 142: 1–138.

A Semi-Aquatic Life Habit for Psittacosaurus 339

24 Habitual Locomotor Behavior Inferred from Manual Pathology in Two Late Cretaceous Chasmosaurine Ceratopsid Dinosaurs, Chasmosaurus irvinensis (CMN 41357) and Chasmosaurus belli (ROM 843) ELIZABETH REGA, ROBERT HOLMES, AND ALEX TIRABASSO

we correlate progressive degenerative joint disease

integrity of the joint cartilage and which provide an un-

of the first digit of the manus with inferred habitual fore-

usually clear picture of abnormal joint angulation. Given

limb locomotor behavior of two mature chasmosaurine

the apparent prevalence of this condition in mature

ceratopsid dinosaurs from the Late Cretaceous of Alberta,

chasmosaurs and aspects of the etiology that are similar

Canada. A complete articulated right manus from the

to a condition in humans known as hallux valgus, we

holotype of Chasmosaurus irvinensis (CMN 41357) in-

propose that habitual forced adduction of the pollex led

cludes a first digit that is markedly abnormal in shape

to the observed joint angulation deformity with ad-

and orientation. The external and joint surfaces of distal

vanced age. Digital animated reconstruction of three al-

metacarpal I and the proximal phalanx are highly rough-

ternative locomotor cycles from scanned skeletal

ened and rugose, with pitting and ridges disrupting the

elements of CMN 41357 suggests that humeral eversion

normally smooth surfaces. Careful matching of joint sur-

varying from 20 to 32\ during the locomotor cycle, ac-

faces yields a reconstruction in which the metacarpo-

companied by strong humeral rotation resulting in man-

phalangeal joint is medially deviated, causing the pha-

ual pronation or rolling during the terminal stance

langes of the pollex to curve sharply laterally toward the

phase, best accounts for the observed changes.

second digit. The first metacarpo-phalangeal articulation deviates approximately 34\ from the inferred normal position. Strikingly similar abnormal morphology, also

Differential Diagnosis in Paleopathology

involving the first digit, is exhibited by the right and left

Paleopathological ‘‘diagnosis’’ is necessarily an approximate

manus of Chasmosaurus belli (ROM 843), sister taxon to

activity. Diagnosis is important to the clinician who is at-

C. irvinensis. Gross appearance of both hands is consis-

tempting accurately identify a disease. But paleopathologists

tent with superimposed pathological processes: chronic

are not attempting to ‘‘cure’’ fossils—instead, they are occu-

sub-periosteal inflammation and subsequent degenera-

pied with looking for clues to past life events, and a specific

tive changes in articular bone surfaces and joint capsule

diagnosis applied with undue confidence potentially obscures

of the first digit, the latter due in part to repetitive

more than it reveals. Rarely are bone lesions pathognomic-

trauma. Precise etiology of the former is unknown but of

meaning that one identifiable type of lesion is caused by one

less importance than the signal of life history presented

and only one causative agent (Kelley and Eisenberg 1987; Salo

by the secondary degenerative changes, which affect the

et al. 1994). Moreover, strict diagnosis of disease in the past is

340

necessarily committed to finding similarities to disease man-

weighting of the manus during the step cycle that provide the

ifestations in the present, when certainly both pathogens and

evidence of greatest significance.

hosts have evolved considerably. The response of fossil organ-

In this chapter, we review the evidence for ceratopsid man-

isms may differ considerably even from that of their extant

ual locomotor position, provide a detailed description of the

descendants, as it has demonstrably changed for the human

pathological hands in question, and explore a broad differen-

species during its comparatively short period of existence

tial diagnosis based in the avian, reptilian, and mammalian

(Buikstra 1981). Assertion of diagnostic authority or uncritical

literature. We finish with the evidence provided by locomotor

wholesale comparison with the human medical literature can-

simulations based upon the observed pathologies.

not obviate the uncertainty with which this process is fraught. For comparative purposes, phylogenetic bracketing by examining bone disease in members of the extant crown clades (bracketing the pathology in question) is the most conserva-

The Ceratopsid Forelimb During Locomotion

tive approach. Indeed, examples drawn from the avian and

The feet of large tetrapods are subjected to enormous stress,

crocodilian literature can aid in narrowing the range of proba-

the particulars of which are largely determined by mass and

ble etiologies for dinosaurian pathology (Rega and Brochu

the parameters of the step cycle. The stance and step cycle

2001). However, confounding factors such as scale, differ-

of ceratopsid dinosaurs have historically been the subject of

ences in bone mass, and bone structure and development,

much debate (see Thompson and Holmes 2007). In early com-

as well as unknown qualities of ecology and habitat—among

posite skeletal reconstructions (e.g., Marsh 1891), these mas-

other issues—make direct comparison only an approxima-

sive animals were depicted with both front and hindlimbs

tion. Without precise knowledge of the metabolic and immu-

held vertically under the body, implying columnar support

nological characteristics of extinct taxa, the uncertainty of

throughout step cycle, similar to that of the elephant. More

any diagnosis of pathology in the fossil record can only be

detailed study of articulated material (e.g., Gilmore 1905;

partly ameliorated by comparison with living relatives.

Sternberg 1927; Lull 1933; Osborne 1933; Erickson 1966) indi-

However, in assessing the etiology and significance of a par-

cated that although the hindlimb approximated a graviportal

ticular pathological occurrence in the fossil record, there is a

stance, the forelimbs did not. Rather, the humerus appeared

clear hierarchy of inferential confidence (Rega and Holmes

to be oriented in a more horizontal plane, projecting at right

2006). Without the necessity of committing diagnostically to

angles to the body axis, producing a sprawling posture with

a specific pathogen, processes such as trauma and subsequent

the front feet tracking much further from the sagittal plane

healing, degenerative changes, chronic patterns of bone for-

than the hind feet.

mation, and repetitive injury lend themselves to more secure

After the recent rehabilitation of large dinosaurs into poten-

conclusions about the interplay of these pathologies with the

tially endothermic (e.g., Bakker 1987) fast growing active ani-

lifeways of extinct forms than diagnoses of specific infectious

mals (Padian et al. 2004, Lee and Werning 2008), upright front

diseases, metabolic abnormalities, or dietary deficiency.

limbs were once again hypothesized for ceratopsids, although

In the cases presented here, we advance the hypothesis that

this was not universally accepted (e.g., Johnson and Ostrom

secondary degenerative changes due to locomotor stresses

1995). Even within the proponents of an upright stance for

have left a clear signal on three pathological Chasmosaurus

ceratopsids, there is considerable variation in specific hypoth-

thumbs. Our differential diagnosis of the primary disease is

eses (e.g., Bakker 1986; Dodson and Farlow 1997; Paul and

necessarily broad, with parallels drawn from both the avian

Christiansen 2000). That such a wide range of hypotheses

and mammalian literature, including humans. Comparisons

could be generated reflects the paucity of complete, articu-

must be approached critically, as ‘‘the character, quality, and

lated ceratopsid limb material available to test them.

progression of inflammatory reactions in non-mammalian

Recently, two important developments have provided an

vertebrates differ from the mammalian paradigm more mark-

opportunity to reevaluate these competing hypotheses: (1) the

edly than any other aspect of general pathology’’ (Therio

discovery of ceratopsid trackways (Lockley and Hunt 1995),

2004). Birds obviously must figure heavily, as the only extant

permitting the manus to be positioned accurately on the sub-

dinosaurs. The appropriateness of mammals for comparison

strate for the first time, and (2) the availability of three-

to Chasmosaurus is based upon similarities of ornithosuchian

dimensional digital scans and scale models of a ceratopsid

bone (sensu Padian et al. 2004) characteristic of large dino-

skeleton with a complete, articulated forelimb that could

saurs to that of mammals, particularly in terms of rapid

be manipulated to explore various possible stances and step

growth, sub-periosteal bone deposition, and extensive Haver-

cycles (Thompson and Holmes 2007). Based upon the integra-

sian remodeling. The primary bone disease in these specimens

tion of these two lines of evidence, the forelimb stance in

is likely caused by infection, trauma, or a combination of both

Chasmosaurus cannot be accurately described as either upright

factors. But it is the secondary changes reflecting position and

(e.g., Bakker 1987) or sprawling (e.g., Sternberg 1927), but is

Habitual Locomotor Behavior Inferred from Manual Pathology 341

FIGURE 24.1.

Chasmosaurus irvinensis (CMN 41357). Dorsal view of right manus in normal articulation, ignoring surface angles of digit I metacarpophangeal (MCP) joint.

closer to the semi-erect stance hypothesized by Dodson and

thology conveys a reliable locomotor signal for these taxa is

Farlow (1997) and Paul and Christiansen (2000), although the

greatly increased.

total range of motion appears to have been much more limited than previously suggested (Thompson and Holmes 2007).

Materials and Methods

Preliminary analysis of this stance, which has no close analogues in other known tetrapods, suggests that this produces

We examined the hands of two completely prepared speci-

a peculiar rolling gait in which the weight shifts toward

mens of Chasmosaurus. The type of Chasmosaurus irvinensis

the preaxial (medial) side of the manus as the stance (pro-

(CMN 41357) is a largely complete specimen that preserves,

pulsive) phase of the step cycle progresses (Thompson and

among other elements, a complete right forelimb, including

Holmes 2007).

scapula-coracoid, humerus, ulna, radius, and an articulated

Three articulated hands from two specimens of Chasmo-

carpus and manus (Holmes et al. 2001; Thompson and Holmes

saurus of advanced age show strikingly similar degenerative

2007), but is lacking the majority of the left front and hind-

changes in the first digital ray that provide additional evi-

limbs and tail. The right manus of CMN 41357 is exceptionally

dence for the role of the manus in locomotion (Rega and

well preserved with all five digits present (Fig. 24.1), but be-

Holmes 2006). The sub-periosteal and joint surfaces of the

cause the left is entirely missing, no intra-individual com-

thumb and associated metacarpal preserve evidence of inflam-

parison is possible. This specimen was excavated by W. Lang-

mation and subsequent chronic degenerative joint disease.

ston, Jr., in 1958 and prepared by D. Stoffregen, who drafted a

Because the significance of this pathology is the readily ob-

detailed specimen map during preparation that confirms the

servable response of the bone to biomechanical stresses pro-

relative placement of the forelimb elements (Thompson and

duced during locomotion over the course if its life, identifica-

Holmes 2007).

tion of the processes that created the condition is possible

Chasmosaurus belli (ROM 843), also a relatively complete

without the necessity of committing to a specific diagnosis or

skeleton, preserves elements of both forelimbs, including

etiology in the clinical sense. Moreover, because these changes

both complete hands and associated carpi (see McGowan

are manifest on two individuals separated in time by at least a

1991: 365). This specimen was excavated by Levi Sternberg in

half a million years, one bilaterally, confidence that this pa-

1926. Although no specimen map was created, Sternberg’s

342 rega, holmes, & tirabasso

field description indicates the state of articulation and expo-

2003) and avoids the problems associated with undue diag-

sure of the entire specimen, and suggests that the forelimb

nostic confidence. Thereafter, we advance hypotheses con-

elements were not scattered (Seymour pers. com. 2008). De-

cerning the cause of these conditions based upon similarities

spite the lack of direct histological measures of age at death

to existing conditions in appropriate animal models.

for either specimen, their large size, and in particular, complete resorption of the orbital horncores indicates that they were individuals of advanced age (see Sampson et al. 1997; Ryan and Russell 2005).

Pathologies in the Holotype of Chasmosaurus irvinensis

Articulated hands of comparable completeness and quality,

CMN 41357 displays a number of pathological and/or degen-

although rare, are known for Centrosaurus apertus (= Mono-

erative conditions. The cranial pathology is all due to resorp-

clonius, AMNH 5351, Brown 1917; YPM 2015 in Lull 1933)

tion of bone tissue, while the postcranial pathology results

and Protoceratops (various specimens, Brown and Schlaikjer

from mixed destructive and proliferative osseous responses.

1937). However, because these specimens are mounted dis-

These manifestations are summarized in Table 24.1, where a

play pieces, manipulation and thorough examination of all

brief description and hypothesized etiology is offered for all

aspects of the individual skeletal elements are not currently

the pathological lesions on this specimen. Thorough descrip-

possible. In addition, not all mounted skeletons are as com-

tion and differential diagnosis will be limited to the pathology

plete as they appear. For example, a pair of Chasmosaurus skel-

of the first manual digit.

etons, for many years on exhibit at the Canadian Museum of Nature in Ottawa (Sternberg 1927), have been widely regarded as being largely complete. However, recent study has shown

The Case of the Deviated Thumb

that much of both skeletons—including almost all of the

The entire metacarpal I and a proximal third of the proximal

bones of the manus and pes—is restored in plaster (Mallon

phalanx of CMN 41357 manifest highly rugose external bone

and Holmes 2006).

surfaces, in marked contrast to the smooth, even texture of

The phalangeal formula for both specimens is 2-3-4-3-2, as

the adjacent metacarpals and phalanges of the other digits.

reported for Centrosaurus apertus (= Monoclonius, Brown 1917).

The bone surface is most markedly abnormal on metacarpal I,

Terminal phalanges of digits I–III in both cases show longitu-

affecting lateral, medial, dorsal, and ventral aspects (Figs.

dinally striated bone texture and morphology indicative of

24.2A, B) and includes alterations of the distal joint surface,

keratinous hooves (Thompson and Holmes 2007). The most

but not the proximal surface. The overall shape of the shaft of

distal phalanges of digits IV and V each bear what appears to

metacarpal I is distorted, being more angular as opposed to

be a terminal articular facet. As in the similarly articulated and

cylindrical and with a distal joint surface that is sharply an-

well-preserved manus of Centrosaurus apertus (Brown 1917),

gled at the inferio-medial border away from the joint (Fig.

there is no trace of a hoof-bearing terminal phalanx associated

24.3). Significant resorption has eroded over 45% of the medi-

with either digit, suggesting they were either absent in vivo or

odistal joint margin. The distal joint surface is clearly dis-

unossified. Terminal phalanges of the first and second digits

rupted by a process that is both resorptive and proliferative. A

are distinctly larger than that of the third digit, suggesting

large 8 mm diameter subcircular island of roughened porous

that the preaxial (medial) side of the manus bore most of the

bone is present in the center of the rugose joint surface. Nu-

weight or sustained more stress during locomotion through

merous 2 mm diameter, smooth-margined irregular pits are

these digits (Thompson and Holmes 2007).

present, primarily on the medial joint surface where the shape

Because of the rarity and completeness of these specimens, destructive methods such as thin sectioning for histological

change is most marked (Fig. 24.4D). The proximal joint surface is unremarkable (Fig. 24.4B).

analysis could not be performed. Because CT scanning of the

The external bone surface of metacarpal I is highly uneven,

terminal phalanges of CMN 41357 failed to discern internal

with roughened nodules elevated above the level of the nor-

morphology, no additional scans were attempted. Joint angles

mal cortex. Although raised, the texture of these nodular foci

were measured using a goniometer positioned along long

is rugose, as would be expected in a chronic condition. There

bone axes.

is no evidence of the finely porous or spongy texture that is

In the specimens examined, we commence by formulating

indicative of rapid sub-periosteal bone deposition, which can

diagnostically neutral descriptions focusing on the under-

be caused by acute infection or contusion occurring premor-

lying osseous processes and overall skeletal pattern of bone

tem with insufficient time for remodeling before death. In-

resorption and deposition, rather than asserting a diagnosis

stead, the lesions are sclerotic, as if the cortex contains multi-

based upon pattern matching of lesions observed in extant

ple foci of chronic proliferation. A small (approximately 7 —

organisms. This is the best practice in paleopathology (Ortner

7 mm) patch on the lateral aspect of the proximal metacar-

Habitual Locomotor Behavior Inferred from Manual Pathology 343

Table 24.1. Summary of Pathology in CNM 41537 Element(s) Right postorbital horncore

Description

Probable etiology

A smooth round deep lytic lesion with smooth

Is likely the result of the normal aging

margins set into a low, rugose patch of bone; Has

process in ceratopsians (see Sampson et

appearance of a well-resolved chronic process,

al. 1997; Ryan and Russell 2005).

similar to that seen ROM 843 (Chasmosaurus belli, Godfrey and Holmes 1995) and CMN 344 (Styracosaurus albertensis—Ryan et al. 2007). Right squamosal above the lower

Manifests a similar lesion completely perforating

Tanke and Farke (2007) documented

temporal opening

the outer and inner surfaces of the bone.

these lesions as a widespread phenomenon in ceratopsians, likely due to chronic resorption as a part on normal ontogeny and not due to healed puncture trauma.

Cervical vertebrae C6–7; slight

An extensive osteoproliferative mass resembling

The differential diagnosis includes many

involvement of C8, C4, C5–C9

‘‘cauliflower’’ with cervical rib fusion and C5–C9

neoplastic conditions; chondrosarcoma

intervertebral disc space; C6–7 ribs

intervertebral disc space obliteration and fusion;

possibly arising from multiple prior osteochondromas at this location is most likely.

Right hindfoot, digit IV 3rd phalanx

A smooth surfaced, small circular excrescence on

Benign slow-growing button osteoma.

lateral aspect of midshaft. Right metacarpal I & proximal phalanx

Rugose mixed osteoproliferative and osteolytic

(See text)

lesions with considerable shaft shape change and degenerative alterations of both joint surfaces.

pal II shaft also displays this altered bone texture. This portion

came into close apposition. The distal end of the phalanx ap-

of the metacarpal II shaft nests closely into the adjacent meta-

pears normal in shape and texture (Fig. 24.4C). The com-

carpal I and the altered bone texture here could be due to

plementarity of the metacarpal-phalangeal joint surfaces al-

infectious spread from the adjacent metacarpal 1. Three

lows a reconstruction of the angle of deflection of this first

grooves, 1–3 mm wide, extend mediolaterally across the en-

metacarpo-phalangeal joint at 34\, as measured by a goniome-

tire dorsal surface of metacarpal I. These are consistent with

ter aligned with the long axis of each bone (see Fig. 24.3).

the interpretation as blood vessel impressions left as subperiosteal bone deposition caused as inflammation raised the bone profile (see Fig. 24.2). The proximal phalanx of the first digit of CMN 41357 shows

Pathologies in Chasmosaurus belli (ROM 843)

similar alterations in the proximal third of the shaft, with

Similar to the case of Chasmosaurus irvinensis, this specimen of

more extensive lesions on the dorsal and lateral aspects; how-

C. belli displays a number of pathologies of the cranium and

ever, the textural change is not as marked as seen on metacar-

postcranium, including traumatic, degenerative, and devel-

pal I. The medial proximal border is missing bone at an angle

opmental conditions. As in CNM 41357, the horncores are

complementary to the resorption of the articulating metacar-

marked by advanced resorption and there is also a similar well-

pal I. The proximal joint surface shows two resorptive patho-

resolved resorption fenestra on the left squamosal (Tanke and

logical lesions (Fig. 24.4A) One lesion is a large (4 mm diame-

Farke 2007), pointing to the advanced age of the animal. Ad-

ter) pit with smooth margins on the lateral portion of the

ditional pathologies include marked discrete osseous excres-

surface; the second is medial to this and is a linear lesion ap-

cences on the right metacarpal IV (see Fig. 24.5A) and on the

proximately 3 — 9 mm. The shape is broadly complementary

left eighth and ninth thoracic ribs (McGowan 1991). These

to that of the opposing metacarpal I surface. These appear to

latter conditions will be discussed in greater detail below, as

indicate degeneration and narrowing of the articular cartilage

they may bear upon the etiology of the deviated thumbs.

cap and chronic alteration of the articular surfaces as they

344 rega, holmes, & tirabasso

The thumb malformation is the most striking pathological

FIGURE 24.2.

Chasmosaurus irvinensis (CMN 41357). Dorsal (A) and ventral (B) views of right metacarpal I, and articulated proximal phalanx with respect to MCP joint surface morphology.

similarly shared by these two remarkable speciments. The

found in the distal joint surface of CNM 41357 and clearly

right and left first metacarpals and proximal phalanges of

differs from the normal texture of the adjacent proximal met-

ROM 843 both display varying degrees of rugose surface mor-

acarpal articular surfaces displayed by right metacarpals II–V.

phology similar to that manifest on CNM 41357, with the

In addition, this roughened articular surface extends 4 mm

appearance of mixed osteoproliferative/osteolytic lesions ac-

onto the dorsal aspect of right metacarpal I, perhaps indicat-

cumulated chronically. While the severity of the surface dis-

ing slight hyper-extension of the first digit on the carpus.

tortion of the right metacarpal I/proximal phalanx complex

The distal articular surface is similarly pitted and the shape

is comparable in degree to that of CNM 41357, the elements

of this articular surface is distorted in a similar manner to that

on the left are considerably less affected (see Fig. 24.5A, B).

of the distal joint surface of the CNM 41357 metacarpal I.

Right metacarpal I is the most severely affected of all the ele-

However, the deformity is more severe in this specimen, with

ments. In addition to the rugose surface texture apparent on

over 65% of the superior margin truncated from the superio-

the dorsal, ventral, lateral, and medial aspects, the shaft shape

medial to inferio-lateral aspect. The proximal right phalangeal

changes and degenerative alterations of both proximal and

articular surface is complementary to that of the distal meta-

distal joint surfaces are marked. The proximal joint surface

carpal I, with considerable pitting. The angle of the right

is roughened with multiple osteolytic foci similar to those

metacarpo-phalangeal joint measures 42\. The entire dorsal

Habitual Locomotor Behavior Inferred from Manual Pathology 345

FIGURE 24.3.

Chasmosaurus irvinensis (CMN 41357).

Dorsal view of right manus in articulation; digit I articulated with respect to MCP joint surface morphology.

and medial surfaces of this phalanx are severely modified. The

by means of differential diagnosis, during which the observed

dorsomedial edge displays two parallel 1–2 mm thick stria-

pathology is compared with suspected diagnoses. In paleopa-

tions running along the long axis of the phalanx. These re-

thology, this must include consideration of similar lesions in

semble the putative ‘‘blood vessel’’ markings seen on the dor-

related species.

sal metacarpal of CNM 41357 and are likely indicative of a

The rugose hyperostotic bone surface texture of the most se-

raised periosteum. The proximal medial edge adjacent to the

verely affected elements has the appearance of sub-periosteal

joint surface has a hook-like morphology resembling the over-

bone deposition accumulated and remodeled over a substan-

hanging edges of remodeled periarticular osteophytes.

tial period of time. Chronic periosteal inflammation and the

The left metacarpal I and phalanx show alterations of the

supracortical bone deposition subsequent to periosteal eleva-

bone surface morphology and angulation of the metacarpo-

tion can be attributed to multiple sub-periosteal abscesses,

phalangeal joint (20\). However, they are less severely dis-

these ultimately due to a variety of potential causes: single

torted than those on the right side. Only the appositional

or repeated trauma, venous stasis, and/or osteomyelitic in-

surfaces of the metacarpo-phalangeal joints display a degen-

fection by direct inoculation or blood borne spread of patho-

erative roughening and pitting. The external surface of the

gens, which can include bacteria and fungi. Inflammation

proximal phalanx is almost normal in surface texture, except

(synovitis) and pathogen invasion of the associated joint

for three pits at the dorsal distal margin adjacent to the inter

capsule would account for the alterations of the metacarpo-

phalangeal joint: an irregular 8 — 3–4 mm pit, a smaller 3 —

phalangeal joint. The joint could have been first infected and

3 mm pit proximal to it, and a tiny 1 mm circular pit proximal

spread subsequently to the subchondral bone, as is common

to this. The depth of these pits ranges from 2–3 mm.

in cattle (Verschooten et al. 2000), or the sequence could have been reversed. Similar rugose surface texture with deep

Pathological Detective Work

erosions on the proximal joint surface of an Allosaurus pedal phalanx was noted by Hanna (2002), who attributed the con-

Given the clues provided above, how do we identify the cul-

dition to chronic suppurative osteomyelitis, or an oozing in-

prit in these pathologies? The field of possibilities is narrowed

fection of the bone and associated periosteal membrane. A

346 rega, holmes, & tirabasso

Chasmosaurus irvinensis (CMN 41357). (A) Pathological proximal and (C) normal distal joint surfaces of proximal manual pollex; (D) pathological distal and (B) normal proximal joint surfaces of right metacarpal I. FIGURE 24.4.

manual phalanx from the same individual manifested similar surface alterations, but lacked joint involvement.

OSTEOMYELITIS AND INFECTIOUS SYNOVITIS Osteomyelitis is well documented, but not especially common in dinosaurs. A complete specimen of the basal ceratopsid, Psittacosaurus, was found to be affected by chronic osteomyelitis of the fibula that did not spread to other bones (Lü et al. 2007). The tibia of a hypsilophodontid manifested classic signs of chronic osteomyelitis in a partial pelvis and complete hindlimb, which affected the length and shape of the tibia as it grew, sparing the elements proximal and distal to it (Gross et al. 1993). The Tyrannosaurus rex ‘‘Sue’’ (FMPR 2081) manifests extensive osteomyelitis of the entire left fibular diaphysis, sparing the joint surfaces and surrounding bones (Rega and Brochu 2001; Brochu 2003). These cases con-

FIGURE 24.5. Chasmosaurus belli (ROM 843). (A) Dorsal view of right and (B) left mani in articulation; digit I articulated with respect to MCP joint surface morphology.

form to the ‘‘classic’’ manifestation of chronic osteomyelitis, where a thick layer of enveloping bone (involcrum) covers a necrotic interior core of bone (sequestrum), with external

hadrosaur tibia CMN 41201 (Tanke and Rothschild 1997) and

evidence of pus-draining holes (cloacae). Osteomyelitis is

hadrosaur manual elements (Moodie 1926).

more commonly noted in theropod skeletons, particularly Al-

Osteomyelitis commonly occurs on the elements lying di-

losaurus manual and pedal elements (Peterson et al. 1972;

rectly underneath the skin surface unprotected by muscle

Madsen 1976; Molnar 2001; Hanna 2002). Pathology attrib-

mass, primarily affecting the extremities and long bones of

uted to osteomyelitis or infectious periostitis in ornith-

the distal limb segments, such as the tibia, fibula, and radius

ischians is reported in a ceratopsian scapula (Rothschild and

(Ortner 2003). It frequently localizes in the metaphyseal bone

Martin 1993) and isolated Triceratops coronoid (Rothschild

of young animals and the vertebrae of adults ( Johnson 1994).

1997). Several cases of ancient osteomyelitis are proposed as

The tibiometatarsus is commonly affected in birds (Wyers et

secondary to primary trauma (Marshall et al. 1998; Mc-

al. 1991). Osteomyelitic infection is often related to trauma,

Whinney et al. 1998; Molnar 2001; Hanna 2002). Additional

either to the skin and underlying periosteum in the involved

cases attributed to osteomyelitis can be found in an isolated

bone, but can result from hematogenous (blood-borne) spread

Habitual Locomotor Behavior Inferred from Manual Pathology 347

from a distant site. While extant mammals mount an ag-

transportation and breeding (Wyers et al. 1991). Dyschondro-

gressive inflammatory response, it has been suggested that

plasia and the resultant condition osteochondrosis, which re-

birds manifest a suppurative reaction less frequently (Montali

sult from errors in growth plate endochondral bone forma-

1988). Osteomyelitic infection becomes systematic less fre-

tion and cartilage development, are common conditions in

quently in birds as compared to mammals (Stocker 2005).

birds, including terrestrial ratites (Tully and Shane 1996). Dys-

A few specific infectious agents among the myriad possibili-

chondroplasia is extremely common in some varieties of do-

ties include systemic bacterial pathogens, such as Staphlococ-

mestic fowl, such as broiler chickens, ducks, and turkeys, af-

cus aureus, a common cause of osteomyelitis in birds as well as

fecting up to 30% of flocks (Crespo and Shivaprasad 2003).

mammals (Stocker 2005) and Salmonella, commonly found in

Dyschondroplasia can also account for the some of the cystic

squamates (Ramsey et al. 2002). Escherichia coli and Streptococ-

changes observed in the Chasmosaur thumbs and resulting in

cus sp. have been isolated from avian osteomyelitic lesions

bone and joint deformity. Several lesions of the postcranial

(Clark et al. 1991). Actinomyces, a bacterium with fungal char-

skeleton of Chasmosaurus belli (ROM 843) already suggest that

acteristics, found in the environment and typical body flora,

this animal was affected by one of the sequalae of dyschondro-

was the cause of an extensive outbreak of osteomyelitis in

plasia and growth plate dysplasia, known as osteochondroma.

turkeys, and is known to cause osteomyelitis with certain en-

ROM 843 manifests marked discrete osseous excrescences on

vironmental conditions ( Johnson 1994). Fungal conditions

the left metacarpal IV (see Fig. 24.5A) and the left eighth

that can result in osteomyelitis are many, and include blasto-

and ninth thoracic ribs (McGowan 1991). While the latter has

mycosis, coccidiodomycosis, and dermatophytosis ( Johnson

been attributed to bony proliferation subsequent to rib frac-

1994; Ramis et al. 1998). Aspergillis is a common fungal patho-

ture (McGowan 1991), the truncated nature of rib eight to-

gen of wild and domestic birds (Stocker 2005) and has been

gether with the rounded morphology of both the eighth rib

isolated from osteomyelitic lesions of birds (Bauck 1994), rep-

and manual exostosis is strongly suggestive of osteochon-

tiles (Heatley et al. 2001), and mammals (Bajracharya et al.

droma. The location, as well as the rounded form of the ex-

2006; Sherman et al. 2006). Environmental fungi not typically

ostosis and inferred age of the individual, are all consistent

pathological can cause severe infectious synovitis and os-

with a benign exostosis of developmental endochondral ori-

teomyelitis when introduced through traumatic inoculation

gin. That these occur in the two different locations in the same

(Destino et al. 2006). Systematic infection with the parasitic

animal could indicate a condition commonly affecting many

protozoan causes bone infection. Trichomonas can cause skull

animals and humans known as multiple hereditary car-

and beak erosions in birds (Stocker 2005) and Hepatozoa has

tilaginous exostoses, which is due to a well-documented auto-

been demonstrated to cause marked osteomyelitic lesions in

somal dominant mutation (Schmale et al. 1994; Shupe et al.

coyotes (Kocan et al. 2000). Although hepatozoans are preva-

1981). Multiple exostoses can also occur from mechanical dis-

lent in reptiles and lissamphibians, they do not typically cause

ruption or ingestion of excessive vitamin A (Lynch et al. 2002).

symptomatic disease as they do in mammals, where Hepatozoa is well known for causing hepatozoonosis and osteomyelitis.

The differential diagnosis for osteochondroma necessitates the consideration of bone spurs. These are differentiated based

Tuberculosis and blastomycosis (Kelley and Eisenberg 1987)

on their position; in a joint capsule attachment (marginal cap-

should be included in the differential diagnosis; however,

sular osteophyte), in a ligament (syndesmophyte), an ossify-

these conditions tend to be primarily destructive (osteolytic)

ing area of muscle attachment (enthospathy), or forming a

in extant species, not proliferative, and are therefore less likely

tumor (neoplasm). The shape of the growths on CMN 843

the cause of the manifestations considered in this case. This

makes the former unlikely, and their form/location make it

observation also renders the specific diagnosis of tuberculosis

unlikely to be either a syndesmophyte or an enthesophyte

as the cause of the osteomyelitis in Psittacosaurus less likely (Lü

(Mann and Hunt 2005).

et al. 2007). OTHER PATHOLOGICAL CONSIDERATIONS DYSCHONDROPLASIA AND OSTEOCHONDROMA

Osteochondritis dessicans, or subchondral lytic lesions, is also

Osteomyelitic infection can accompany other underlying

common in the joints of birds (Crespo and Shivaprasad 2003)

conditions in birds. Dyschondroplasia, a relatively common

and is consistent with the observed pathological joint mor-

and usually benign developmental defect characterized by ab-

phology. These are an unknown etiology and may be due in

normal masses of avascular cartilage below the growth plate

part to trauma. Healed subchondral fractures should also

(Crespo and Shivaprasad 2003), was associated the widespread

be considered as they can cause progressive chronic shape

occurrence of osteomyelitis deformity and lameness in the

change in the joint surfaces and therefore alter articulation

legs in turkeys, itself exacerbated by the stress associated with

(Resnick et al. 1995).

348 rega, holmes, & tirabasso

Diaphyseal micro fracture should be part of the differential diagnosis. In a putatively osteomyelitic Allosaurus manus pha-

overwhelming preponderance of avian cases of gout is visceral, not articular (Klasing and Austic 2003).

lanx (Hanna 2002), scanning electron micrography of a transverse cross section of the shaft revealed the presence of fibro

Does Diagnosis Matter?

lamellar bone uniting an incomplete oblique-longitudinal fracture through the zonal lamellar bone of the cortex. In this

We have demonstrated that a number of potential causes exist

case, both the osteomyelitis and the fracture may be the result

in appropriate animal models that could account for the mor-

of the same process, namely trauma.

phological changes observed in these Chasmosaurus thumbs.

Stress fracture has also been documented in ceratopsian

Infection and trauma are factors in virtually all of these patho-

phalanges. An isolated ceratopsian phalanx unattributed to

logical scenarios, perhaps exacerbating an underlying de-

species (CNM 40741) displays a radiographically verified in-

velopmental defect of cartilage. The cause may be unknow-

ternal stress fracture with considerable smooth remodeled

able. However, for the purposes in understanding the role of

bone lying external to it (Rothschild 1988). The appearance of

the forelimb in locomotion, the degenerative angulation of

the external bone in this specimen is very different than that

the joint surfaces remains the truly critical observation.

of the chasmosaur cases currently under consideration.

Whatever the nature of the underlying primary pathology,

Finally, without radiographic or CT evidence, the differential diagnosis for the thumbs should be broadened to include

it is the palimpsest of the secondary degenerative changes that contain the ‘‘smoking gun’’ for the locomotor signal. The angulation of all three digits reflects the deformation of diseased

§ Fibro-osseous tumor (radiograph CT would indicate or

bone and joints over time and the habitual forces at work in

eliminate this possibility, as the endosteal cavity will

Chasmosaurus. In this way it is reminiscent of the condition

show large osteolytic foci surrounded by dense

called hallux valgus in humans (Salter 1999), which is, in

sclerotic bone (Ortner 2003)

part, due to forced adduction of great toe and/or inborn weak-

§ Cystic lesions of secondary hyperparathyroidism

ness of transverse ligaments of metatarsal heads. As the dis-

(Watson and Dixon 1977)

tal phalanx pushes toward the midline of foot it forces the

§ Unicameral bone cyst (Ortner 2003)

metatarsal-phalangeal joint medially, resulting in the deformation of the bone termed a bunion and severe angulation of

The pattern of the pathology (identical elements in two indi-

the joint.

viduals in one case bilateral) renders the last two on the list amongst the least likely. Finally, no evaluation of erosive lesions adjacent to joints is complete without consideration of gout. Gout is characterized

Reconstructing Locomotion Causing Thumb Deformation

by monosodium urate crystal deposition. In mammals this is

The deformation of the thumbs serves to further constrain

primarily an articular condition. A diagnosis of gout is itself a

ceratopsid locomotor models previously discussed. Digital

microcosm representing the problems with paleontological

scans of forelimb elements of CMN 41357 were made avail-

differential diagnosis because the articular form is diagnosed

able to the authors by the Canadian National Museum for use

in ancient remains based upon the supposed pathognomicity

in reconstructing Chasmosaurus locomotion.

of the lesion form. This diagnosis has been advanced on morphological grounds alone in dinosaurs and a ratite museum specimen based upon the presence of a solitary spherical per-

SCALE MODEL MANIPULATION

iarticular bone lesion with sharp margins and minimal reac-

Thompson and Holmes (2007) suggest a forelimb stance for

tive bone (Rothschild and Rühli 2007). Based on this criterion,

Chasmosaurus intermediate between sprawling and gravipor-

the morphology of the bone changes in the Chasmosaurus

tal closer to the semi-erect stance hypothesized by Dodson

thumbs does not match the classic pathognomic dry bone

and Farlow (1997) and Paul and Christiansen (2000; Fig.

descriptions of gout lesions, due to the presence of reactive

24.6A, B). The former study utilized scale models, created

bone.

from the scanned limb, limb girdle, and trunk elements from

The literature however, presents a more complicated picture

CNM 41357. These models were carefully aligned with a neo-

of gout. Human studies suggest, in fact, that periosteal reac-

ceratopsian trackway (Lockley and Hunt 1995), paying special

tive bone can be in found cases of articular gout (Liote and Ea

attention to digit position and articular surface matching. Al-

2006). Moreover, the articular lesion form of gout in birds may

though the trackway was probably created by the much larger

be subcutaneous, not within the joint capsule (Beach 1962).

neoceratopsid Triceratops, similarities with respect to pectoral

Interestingly, although gout is very common in birds, the

girdle and limb anatomy and the lack of other clearer pub-

Habitual Locomotor Behavior Inferred from Manual Pathology 349

DIGITAL RECONSTRUCTION OF ARTICULATION AND MOVEMENT In the present study, the scanned elements of the manus and forelimb were digitally ‘‘rearticulated’’ and animated. We also approached the articulation from ‘‘the ground up,’’ using the same published trackway data (Lockley and Hunt 1995) as a starting point for hand orientation and utilizing threedimensional scanned files from CNM 41357. The orientation of the manus was determined following the methodology of Thompson and Holmes (2007) by aligning the digits with the individual toe prints preserved in the trackway. With the manus in place, the humerus, ulna, and radius were attached, carefully matching the positions of the articular surfaces. The distal ends of the epipodials forms a natural arch that can be aligned with the arch formed by the proximal end of the articulated metacarpus. A gap for the cartilaginous carpals was left, which seemed reasonable and does not differ appreciably from other reconstructions. Thereafter, the torso was moved along the midline of the track way until all bones attained a ‘‘best fit’’ with the limb in approximately mid-stance. Given the lack of close congruence of the joints due presumably to large cartilage end caps, there Reconstruction of inferred position of forelimb in Chasmosaurus. (A) Lateral view; (B) frontal view. FIGURE 24.6.

was some degree of play in the positioning of individual elements and the degree of humeral eversion. These parameters were subjected to multiple articular scenarios during digital animated step cycles until they produced the joint angulation

lished trackways dictated its choice. As a result of matching

observed in the pathology of the first digit.

this trackway, the distal epipodium (i.e., the combined heads

The combined proximal articular surface of the articulated

of the ulna and radius), metacarpus, and manus of CNM

metacarpus forms a broad arch. As typically reconstructed,

41357 are reconstructed as strongly arched.

the digits are distinctly splayed from this arch. Thompson

Thompson and Holmes (2007) note the following param-

and Holmes (2007) determined the degree of digital splay

eters for the forelimb posture based on their analysis: (1) the

by matching the trackway data of Lockley and Hunt (1995).

humerus has a limited excursion, moving from a fully pro-

However, Gatesy et al. (1999) demonstrated that trackways do

tracted position of 43\ from horizontal at beginning of the

not necessarily reflect an unmodified pedal morphology. The

stance phase to about 24\ at terminal stance—a total excursion

articulation and step cycle presented in this paper proposes

of just under 20\; (2) the humerus remains distinctly everted

that the digitigrade manus deforms during weight bearing—

throughout, ranging from 30 to 32\ to the parasagittal plane,

such deformation is the rule in virtually all terrestrial tetra-

positioning the elbow and manus further from the midline

pods. We construct a slightly more gathered metacarpus when

than the hind foot; (3) the elbow moves from strongly flexed—

unweighted; this reconstruction is supported by the observa-

an 87\ angle between humerus and ulna—to slightly more

tion that there is a complementarity to the lateral and medial

extended at 114\, and its movement provides a significant

aspects of metacarpals I–III that suggests a slightly more gath-

proportion of the propulsive thrust; (4) the lateral digits IV

ered hand posture during at least part of the step cycle, and

and V must be lifted off the substrate at terminal stance in

spreading of the digits during weight bearing.

order to maintain articulation between the metacarpus and

The elements of the first digit of CMN 41357 are so severely

ulna; and (5) humeral rotation along the long axis is necessary

deformed that the orientation of the metacarpal I cannot be

to maintain articulation of the elements and trackway match-

determined solely on the basis of morphology. Multiple artic-

ing. This unique stance produces a slight metacarpal roll as

ular scenarios were attempted to find a best fit for these ele-

the stance phase progresses, resulting in the elevation of the

ments, including articulating metacarpal I upside down (i.e.,

outer toes toward the end of the phase, and a concomitant

with the presumptive flexor surface in extensor position). This

shift of weight toward the preaxial (inner) side of the manus

‘‘articulation’’ results in a laterally hyper-extended ‘‘splaying’’

(Thompson and Holmes 2007).

of the thumb. However, the right manus of CMN 41357 was

350 rega, holmes, & tirabasso

preserved in close articulation and is documented by Stoffre-

walk cycle not only served to further lessen oscillation of the

gen’s 1959 specimen map. There is no doubt from this map

center of mass but increased slightly the rolling action of

that all of the digits lay in natural relation to each other, and if

the manus, which resulted in displacement of the metacarpal-

metacarpal I had been rotated 180\ around the long axis dur-

phalangeal joint of the first digit that was consistent with the

ing preservation, it was the only bone of the metacarpus or

observed pathology. Given the size of the deltopectoral crest

manus to have experienced such a reorientation. These fac-

in ceratopsids, this figure may even be an underestimation of

tors have led us to accept our current articulation as the most

the actual humeral rotation produced during locomotion.

likely. No documentation on the original in situ relations of

During these walk cycles, the elements of the thumb were

the pedal elements of ROM 843 has survived to evaluate the

clearly affected by varying proximal locomotor parameters,

validity of this alternative articulation for this specimen.

with the thumb displaying an automatic ‘‘hyper-extension’’

The chasmosaur 3D rig created from the scanned elements

of the metacarpo-phalangeal joint accompanying the varia-

of CMN 41357 is constructed as a hierarchal linked system,

tion of the degree of humeral eversion. To more closely dupli-

where the individual elements of the manus are hierarchically

cate the observed pathological joint angulation, another in-

linked from child to parent; that is, the most distal element

verse kinematic control was placed between the metacarpal I

(ungual) to the proximal element (metacarpal). The metacar-

and the first phalanx. This control was not manipulated man-

pals of the manus were then linked to the wrist from the rest of

ually (such as the other controls), but was allowed to be influ-

the arm in another hierarchal linked system (Rega et al. in

enced by the progression of the walk cycle. As a result, the

prep). The major pivot points of the entire arm or hierarchal

digit ‘‘buckled’’ during the stance phase in a way that matched

linked system were the wrist, elbow, and shoulder (humerus/

the observed pathology.

scapula-coracoid joint). A maximum range of articulation was applied to every ‘‘3D bone’’ so that the bone and associated poly-mesh element did not rotate past their acceptable range,

Discussion

as determined by joint morphology. The elbow, shoulder, and

Reconstruction of the forelimb posture of Chasmosaurus ir-

wrist joints were allowed to translate and rotate.

vinensis, using both a scale model combined with footprint data (Thompson and Holmes 2007), and computer simulation using digital scans (this study) indicates that both the limb

Results

stance and kinematics of the ceratopsid forelimb are unique

Using clues from the first digit pathologies that indicated re-

within tetrapods. Neither the upright stance and parasagittal

petitive injuries, three walk-cycle animation models were cre-

limb movement characteristic of cursorial quadrupedal mam-

ated for CNM 41357 to test multiple locomotor hypotheses

mals and dinosaurs nor a primitive sprawling posture typical

while accounting for the pathological findings. Once the ap-

of basal amniotes was possible, as indicated by the articular

propriate walk cycle was selected, further trial and error cre-

morphology and rib cage size. Instead, the humerus was ever-

ated an animation that best accounted for the observed man-

ted throughout the entire step cycle. Thompson and Holmes

ual pathology.

(2007) proposed a model in which the elbows were moderately

In the first walk cycle animation, the humerus maintained a

everted, and the long axis of the humerus formed an angle of

constant eversion of 20\ from the long axis of the body. Al-

about 30\ to the parasagittal. This configuration resulted in a

though this insured that articular surfaces remained reason-

slight rolling of the hand, causing the weight to shift toward

ably congruent, this resulted in an unacceptable ‘‘jerkiness’’ as

the preaxial side as the stance phase progressed, stressing the

the center of mass moved over the planted manus and the

pollex at an angle to the digital axis. The animated digital

rolling action of the manus and contact of the first digit on the

reconstruction in this study refines these parameters. The hu-

substrate was minimal. In the second walk cycle, the humeral

meral eversion varies from 20 to 32\ during the stance phase,

eversion ranged from 20 to 32\, with the maximum eversion

and a 3\ medial rotation along the long axis of the humerus

reached at the terminal swing phase just before initial stance.

during the transition from terminal swing to initial stance

This maximum approached the amount of eversion estimated

phase causes further stress and deformation of the pollex.

by Thompson and Holmes (2007). During the mid-stance

A consideration of the severe deformation of the first meta-

phase, when the manus was planted and the body moving

carpal and the adjacent first phalanx as well as their joint

forward, the humeral eversion was only 20\, the point of max-

surfaces must form part of the functional explanation. 3D ani-

imum forelimb ‘‘uprightness.’’ The resulting animation dem-

mation demonstrates that locomotor forces acting on the

onstrated that the change in the degree of humeral eversion

thumb result in stresses that could induce a malformation

over the course of the step cycle effectively produced a ‘‘rolling

strikingly similar to a human condition called hallux val-

compression’’ experienced by the manus. Adding just 3\ of

gus. The severely deformed metacarpus I and metacarpal-

humeral rotation along the humeral long axis in the third

phalangeal joint is consistent with the forces impinging on

Habitual Locomotor Behavior Inferred from Manual Pathology 351

the manus observed in simulations derived from both manipulation of a scale model and digital images. If, for the sake of argument, an opposite orientation of metacarpal I is assumed (i.e., that the bone was oriented with its extensor surface facing the substrate in our preferred scenario), then the first digit deviates strongly away from, rather than towards, the remaining digits. In this case, the direct comparison with hallux valgus is lost, but an analogous malformation, in which the digit simply deviates in the opposite direction, remains. Regardless of which of the two possible metacarpus I orientations is assumed, the highly deformed condition of the bone surface of the pollex and metacarpo-phalangeal joints and consequent peculiar orientation of the first digit of both specimens is consistent with chronic alteration of the joint surfaces in old individuals caused by the stress created by the unique rolling gait of these animals. The rolling locomotion likely subjected the thumbs to repeated trauma from the substrate, and subsequent inoculation with pathogens could have caused a chronic osteomyelitic infection that made the thumbs all the more susceptible to degenerative alteration. This condition may have been exacerbated by an underlying developmental defect, such as dyschondroplasia, which is suggested by other lesions on ROM 843. This apparently maladaptive trait may have had little or no effect on the reproductive health of the species, since such repetitive stress injuries tend to be most prevalent in older individuals that exhibit declining reproductive rates. The anatomy of the ribs, pectoral girdle, and forelimb of Chasmosaurus irvinensis is closely comparable to that described in other ceratopsids (Brown 1917; Lull 1933; Dodson et al. 2004), and therefore these results may be extrapolated more broadly to other ceratopsid dinosaurs. Acknowledgments

Thanks to K. Seymour of the Royal Ontario Museum for information on ROM 843 and for loan of this specimen’s hands. P. Bloskie, 3D scanning technician at CMN, performed the 3D digitization of CNM 41357 elements. The Montfort Hospital in Ottawa and Dr. D. Tenaschuk and S. Boyle facilitated CT scanning of CNM 41357. D. Stoffregen provided expert preparation of CMN 41357 and drafted the associated specimen map. K. Shepherd and M. Feuerstack provided access to CMN 41357. S. Fujiwara engaged in discussions regarding the orientation of the first manual digit of Triceratops. P. Dodson, M. Ryan, A. Chew, and L. Rush provided valuable input in the revision of the manuscript. S. Sumida assisted with production of the figures and with oversight of Darwin and Owen during research and manuscript preparation. And finally, we are deeply grateful to Wann Langston, Jr., for the discovery and excavation of C. irvinensis, and to all those who continue to excavate new dinosaur material.

352 rega, holmes, & tirabasso

References Cited Bajracharya, S., M. Jayaram, M. P. Singh, and G. K. Singh. 2006. Fungal osteomyelitis of the tibia and fibula with bony ankylosis of ankle, intertarsal and tarsometatarsal joints—a rare presentation. Nigerian Journal of Orthopaedics and Trauma 5(2): 56–57. Bakker, R.T. 1986. The Dinosaur Heresies. New York: William Morrow. ———. 1987. The return of the dancing dinosaurs. In S. J. Czerkas and E. C. Olson, eds., Dinosaurs Past and Present, Vol. 1, pp. 38– 69. Seattle: University of Washington Press. Bauck L. 1994. Mycoses. In B. W. Ritchie, G. J. Harrison, and L. H. Harrison, eds., Avian Medicine: Principles and Application, pp. 997–1006. Lake Worth, Fla.: Wingers Publishing Inc. Beach, J. E. 1962. Diseases of Budgerigars and other cage birds. A survey of post-mortem findings, Part. II. Veterinary Record 74: 63–68. Brochu, C. A. 2003. Osteology of Tyrannosaurus rex: Insights from a nearly complete skeleton and high-resolution computed tomographic analysis of the skull. Society of Vertebrate Paleontology Memoir 7: 1–138. Brown, B. 1917. A complete skeleton of the horned dinosaur Monoclonius and description of a second skeleton showing skin impressions. Bulletin of the American Museum of Natural History 37: 281–306. Brown, B., and E. M. Schlaikjer. 1937. The skeleton of Styracosaurus with the description of a new species. American Museum Novitates 955: 1–12. Buikstra, J. E. 1981. Prehistoric tuberculosis in the Americas. Archaeological Program Scientific Papers No. 5. Evanston: Northwestern University. Clark, S. R., H. J. Barnes, A. A. Bickford, R. P. Chin, and R. Droul. 1991. Relationship of osteomyelitis and associated soft-tissue lesions with green liver discoloration in tom turkeys. Avian Diseases 35: 139–146. Crespo, R., and H. Shivaprasad. 2003. Developmental, metabolic and other noninfectious disorders. In Y. M. Saif, H. J. Barnes, J. R. Glisson, A. M. Fadly, L. R. McDougald, and D. E. Swayne, eds., Diseases of Poultry, 11th ed., pp. 1055–1102. Ames: Iowa State University Press. Destino, L., D. A Sutton, A. L. Helon, P. L. Havens, J. G. Thometz, R. E. Willoughby, Jr., and M. J. Chusid. 2006. Severe osteomyelitis caused by Myceliophthora thermophila after a pitchfork injury. Annals of Clinical Microbiology and Antimicrobials 5: 1–21. Dodson, P., and J. O. Farlow. 1997. The forelimb carriage of ceratopsid dinosaurs. In D. L. Wolberg, E. Stump, and G. D. Rosenberg, eds., Dinofest International: Proceedings of a Symposium Held at Arizona State University, pp. 393–398. Philadelphia: Academy of Natural Sciences. Dodson, P., C. Forster, and S. Sampson. 2004. Ceratopsidae. In D. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Erickson, B. R. 1966. Mounted skeleton of Triceratops prorsus in

the Science Museum. Scientific Publications of the Science Museum 1: 1–16. Gatesy, S. M., K. M. Middleton, F. A. Jenkins, Jr., and N. H. Shubin. 1999. Three-dimensional preservation of foot movements in Triassic theropod dinosaurs. Nature 399: 141–144. Gilmore, C. W. 1905. The mounted skeleton of Triceratops prorsus. Proceedings U.S. National Museum 29: 433–435. Godfrey, S. J., and R. Holmes. 1995. Cranial morphology and systematics of Chasmosaurus (Dinosauria: Ceratopsidae) from the Upper Cretaceous of western Canada. Journal of Vertebrate Paleontology 15: 726–742. Gross, J. D., T. H. Rich, and P. Vickers-Rich. 1993. Dinosaur bone infection. National Geographic Research and Exploration 9(3): 286–293. Hanna, R. R. 2002. Multiple injury and infection in a sub-adult therapod dinosaur (Allosaurus fragilis) with comparisons to allosaur pathology in the Cleveland-Lloyd Dinosaur quarry collection. Journal of Vertebrate Paleontology 22: 76–90. Heatley, J. J., M. A. Mitchell, J. Williams, J. A. Smith, and T. N. Tully, Jr. 2001. Fungal periodontal osteomyelitis in a chameleon, Furcifer pardalis. Journal of Herpetological Medical Surgery 11(4): 7–12. Holmes, R. B., C. A. Forster, M. Ryan, and K. Shepherd. 2001. A new species of Chasmosaurus from the Dinosaur Park Formation of southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Johnson K. A. 1994. Osteomyelitis. In S. J. Birchard and R. G. Sherding, eds., Saunders Manual of Small Animal Practice, pp. 1091–1095. Philadelphia: W. B. Saunders. Johnson, R. E., and J. H. Ostrom. 1995. The forelimb of Torosaurus and an analysis of the posture and gait of ceratopsid dinosaurs. In J. J. Thomason, ed., Functional Morphology in Vertebrate Palaeontology, pp. 205–218. Cambridge: Cambridge University Press. Kelley, M. A., and L. Eisenberg. 1987. Blastomycosis and tuberculosis in early American Indians: A biocultural view. Midcontinental Journal of Archaeology 12: 89–116. Klasing, K. C., and R. E. Austic. 2003. Nutritional diseases. In Y. M. Saif, H. J. Barnes, J. R. Glisson, A. M. Fadly, L. R. McDougald, and D. E. Swayne, eds., Diseases of Poultry, 11th ed., pp. 1027–1053. Ames: Iowa State University Press. Kocan, A. A., C. A. Cummings, R. J. Panciera, J. S. Matthew, S. A. Ewing, and R. W. Barker. 2000. Naturally occurring and experimentally transmitted Hepatozoan americanum in coyotes from Oklahoma. Journal of Wildlife Disease 36(1): 149–153. Lee, A. H., and S. Werning. 2008. Sexual maturity in growing dinosaurs does not fit reptilian growth models. Proceedings of the National Academy of Science 105(2): 582–587. Liote, F., and H. Ea. 2006. Gout: An update on some pathogenic and clinical aspects. Rheumatic Disease Clinics of North America 32: 295–311. Lockley, M. G., and A. P. Hunt. 1995. Ceratopsid tracks and associated ichnofauna from the Laramie Formation (Upper Cretaceous: Maastrichtian) of Colorado. Journal of Vertebrate Paleontology 15(3): 592–614. Lü, J., Y. Kobayashi, Y. N. Lee,and Q. Ji. 2007. A new Psit-

tacosaurus (Dinosauria: Ceratopsia) specimen from the Yixian Formation of western Liaoning, China: The first pathological psittacosaurid. Cretaceous Research 28(2): 272–276. Lull, R. S. 1933. A revision of the Ceratopsia, or horned dinosaurs. Memoirs of the Peabody Museum of Natural History 3: 1–175. Lynch, M., H. McCracken, and R. Slocombe. 2002. Hyperostotic bone disease in red pandas (Ailurus fulgens). Journal of Zoo and Wildlife Medicine 233(3): 263–272. Madsen, J. H. 1976. Allosaurus fragilis: A revised osteology. Utah Geological and Mineral Survey Bulletin 109: 1–163. Mallon, J. C., and R. Holmes. 2006. A reevaluation of sexual dimorphism in the postcranium of the chasmosaurine ceratopsid Chasmosaurus belli (Dinosauria: Ornithischia). Canadian Field Naturalist 120: 403–412. Mann, R. W., and D. R. Hunt. 2005. Photographic Regional Atlas of Bone Disease. St. Louis: Charles C. Thomas. Marsh, O. C. 1891. Restoration of Triceratops and Brontosaurus. American Journal of Science Series 3, 41: 339–342. Marshall, C., D. Brinkman, R. Lau, and K. Bowman. 1998. Fracture and osteomyelitis in PII of the second pedal digit of Deinonychus antirrhopus (Ostrom) an Early Cretaceous ‘‘raptor’’ dinosaur, p. 16. Paleontological Association 42nd Annual Meeting. University of Portsmouth. McGowan, C. 1991. Dinosaurs, Spitfires, and Sea Dragons. Cambridge: Harvard University Press. McWhinney, L. A., B. M. Rothschild, and K. Carpenter. 1998. Post-traumatic chronic osteomyelitis in Stegosaurus dermal spikes. Journal of Vertebrate Paleontology 18(3, Suppl.): 62A. Molnar, R. E. 2001. Therapod paleopathology: A literature survey. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 337–363. Bloomington: Indiana University Press. Montali, R. J. 1988. Comparative pathology of inflammation in the higher vertebrates (reptiles, birds, mammals). Journal of Comparative Pathology 99: 1–26. Moodie, R. L. 1926. Studies in paleopathology: II. Excess callus following fractures of the forefoot. Annals of Medical History 8: 73–77. Ortner, D. J. 2003. Identification of Pathological Conditions in Human Skeletal Remains. San Diego: Academic Press. Osborne, H. F. 1933. Mounted skeleton of Triceratops elatus. American Museum Novitates 654: 1–14. Padian, K., J. R. Horner, and A. de Ricqlès. 2004. Growth in small dinosaurs and pterosaurs: The evolution of archosaurian growth strategies. Journal of Vertebrate Paleontology 24(3): 555– 571. Paul, G. S., and P. Christiansen. 2000. Forelimb posture in neoceratopsian dinosaurs: implications for gait and locomotion. Paleobiology 26: 450–465. Petersen, K., J. I. Isakon, and J. H. Madsen, Jr. 1972. Preliminary study of paleopathologies in the Cleveland-Lloyd dinosaur collection. Utah Academy Proceedings 48: 44–47. Ramis, A., J. Fernandez-Moran, X. Gibert, and H. FernandezBellon. 1998. Dermatophytosis in a Hyacinth Macaw (Anodorhynchus hyacinthinus)—a case report. Proceedings of

Habitual Locomotor Behavior Inferred from Manual Pathology 353

International Virtual Conferences in Veterinary Medicine: Diseases of Psittacine Birds. www.vetuga.edu/vpp/ivcvm/1998/ramis/ index.php. Ramsey, E. C., G. B. Daniel, B. W. Tryon, J. I. Merryman, P. J. Morris, and D. A. Bemis. 2002. Osteomyelitis associated with Salmonella enterica SS arizonae in a colony of ridgenose rattlesnakes (Crotalis willardi ). Journal of Zoo and Wildlife Medicine 33(4): 301–310. Rega, E. A., and C. A. Brochu. 2001. Paleopathology of a mature Tyrannosaurus rex. Journal of Vertebrate Paleontology 21(3, Suppl.): 92A. Rega, E. A., and R. Holmes. 2006. Manual pathology indicative of locomotor behavior in two chasmosaurine ceratopsid dinosaurs. Journal of Vertebrate Paleontology 26(3, Suppl.): 114A. Resnick, D., T. G. Goergen, and G. Niwayama. 1995. Physical injury: Concepts and terminology. In D. Resnick, ed., Diagnosis of Bone and Joint Disorders, pp. 2561–2692. Philadelphia: W. B. Saunders. Rothschild, B. M. 1988. Stress fracture in a ceratopsian phalanx. Journal of Paleontology 62: 302–303. ———. 1997. Dinosaurian paleopathology. In J. Farlow and M. K. Brett-Surman, eds., The Complete Dinosaur, pp. 426–448. Bloomington: Indiana University Press. Rothschild, B. M., and L. Martin. 1993. Paleopathology: Disease in the Fossil Record. Boca Raton, Fla.: CRC Press. Rothschild, B. M., and F. R. Rühli. 2007. Comparative frequency of osseous macroscopic pathology and first report of gout in captive and wild-caught ratites. Journal of Veterinary Medicine, Series A, 54(5): 265–269. Ryan, M. J., R. Holmes, and A. P. Russell. 2007. A revision of the Late Campanian centrosaurine ceratopsid genus Styracosaurus from the western interior of North America. Journal of Vertebrate Paleontology 27: 944–962. Ryan, M. J., and A. P. Russell. 2005. A new centrosaurine ceratopsid from the Oldman Formation of Alberta and its implications for centrosaurine taxonomy and systematics. Canadian Journal of Earth Sciences 42: 1369–1387. Salo, W. L., A. C. Aufderheide, J. E. Buikstra, and T. A. Holcomb. 1994. Identification of Mycobacterium tuberculosis DNA in a PreColumbian Peruvian mummy. Proceedings of the National Academy of Sciences 91: 2091–2094. Salter, R. B. 1999. Textbook of Disorders and Injuries of the Musculoskeletal System, 3rd ed. Philadelphia: Lippincott Williams and Wilkins. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Cera-

354 rega, holmes, & tirabasso

topsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Schmale, G. A., E. U. Conrad, and W. H. Raskind. 1994. The natural history of hereditary multiple exostoses. Journal of Bone and Joint Surgery (American version) 76(7): 986–92. Sherman, K. M., G. D. Myhre, and E. I. Heymann. 2006. Fungal osteomyelitis of the axial border of the proximal sesamoid bones in a horse. Journal of the American Veterinary Medical Association 229(10): 1607–1611. Shupe, J. L., N. C. Leone, E. J. Gardner, and A. E. Olson. 1981. Hereditary multiple exostoses. Hereditary multiple exostoses in horses. American Journal of Pathology 104(3): 285–288. Sternberg, C. M. 1927. Horned dinosaur group in the National Museum of Canada. Canadian Field Naturalist 4: 67–73. Stocker, L. 2005. Practical Wildlife Care. Abingdon: Blackwell Science Ltd. Tanke, D. H., and A. A. Farke. 2007. Bone resorption, bone lesions, and extra cranial fenestrae in ceratopsid dinosaurs: A preliminary assessment. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 319–347. Bloomington: Indiana University Press. Tanke, D. H., and B. M. Rothschild. 1997. Paleopathology. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, pp. 525– 529. New York: Academic Press. Therio, K. A. 2004. Comparative inflammatory responses of nonmammalian vertebrates: Robbins and Coltran for the birds. 55th Annual Meeting of the American College of Veterinary Pathologists ACVP and 39th Annual Meeting of the American Society of Clinical Pathology ASVCP: 1225–1104, www.ivis.org (International Veterinary Information Service). Thompson, S., and R. Holmes. 2007. Forelimb stance and step cycle in Chasmosaurus irvinensis (Dinosauria: Neoceratopsia). Palaeontologia Electronica 10(1): 5A:17 Tully, T. N., and S. M. Shane. 1996. Ratite Management, Medicine and Surgery. Malabar, Fla.: Krieger Publishing Company. Verschooten, F., D. Vermieren, and L. Devriesse. 2000. Bone infection in the bovine appendicular skeleton: A clinical, radiographic and experimental study. Veterinary Radiology and Ultrasound 41(2): 250–260. Watson, A. D. J., and R. T. Dixon. 1977. Cystic bone lesions in related Old English Sheepdogs. Journal of Small Animal Practice 18: 561–571. Wyers, M. Y., Y. Cherel, and G. Plassiart. 1991. Late clinical expression of lameness related to associated osteomyelitis and tibial dysplasia in male breeding turkeys. Avian Diseases 35: 408–424.

25 Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance D A R R E N H . TA N K E A N D B R U C E M . R O T H S C H I L D

larger horned dinosaurs (ceratopsids) are often por-

of nonlethal flank-butting behavior in the family. While

trayed in the media, popular science, and paleontologi-

ceratopsids are often envisioned to have been pug-

cal communities as highly aggressive animals, frequently

nacious and regularly injured themselves during intra-

injuring themselves during encounters with predators

specific encounters, the evident rarity of

and intraspecific rivals. Thousands of ceratopsian speci-

paleopathologies related to such activities does not sup-

mens in the field and in museum collections were exam-

port this suggestion.

ined for examples of paleopathology. This study focused on cranial and postcranial elements recovered from Cen-

Introduction

trosaurus, Styracosaurus, and Pachyrhinosaurus monodominant bonebeds in Alberta, as well as additional

Paleopathology, the study of ancient animal disease and in-

isolated and articulated ceratopsid material from the

jury and related topics, has been sporadically applied to dino-

province. The survey yielded a variety, but relatively few

saur research since 1836 (Tanke and Rothschild 2002). In

numbers of paleopathologies such as healing fractures,

the past two decades dramatic growth and maturation has

fusions, bone infections, and other conditions of un-

occurred in the field of dinosaur paleopathology, as seen in

determined nature. While some lesions may be related to

growing interest and numbers of students, published works,

intraspecific agonistic encounters, many are here inter-

and available specimens, all supported by rapid advances in

preted as benign in nature or related to individual infec-

medical technologies allowing a non-destructive view of af-

tions, age, accidents, teratisms, or other causes. Most

fected specimens.

injuries present in the animals were received during the

Paleopathology is an important field as it not only tells us

adult stage of life. If one type of osteopathy could epito-

about ancient disease and disease processes in dinosaurs, but

mize Albertan ceratopsids as a group, it would be fract-

the occurrences of injuries (i.e., avulsions, fractures, healing

ured posterior dorsal ribs. Nearly half of the entire

bite marks) on bones can provide fascinating clues on possible

sample of osteopathy in Pachyrhinosaurus recovered from

dinosaur behaviors.

one bonebed was located on the mid-dorsal and pos-

Late Cretaceous Albertan ceratopsian osteopathy received

terior dorsal ribs; similar injuries are also known in Cen-

little detailed attention in the distant past. Most early reports

trosaurus and Chasmosaurus. Regularly observed healing

were simply passing observations within larger anatomical de-

fracture injuries to the posterior dorsal ribs are suggestive

scriptions. Parks (1921) was the first, mentioning a possible

355

pathological condition affecting a Centrosaurus parietal. Lull

ceratopsian taxa like Torosaurus and Triceratops could ‘‘lock’’

(1933) briefly noted and figured two fused cervical vertebrae

together (Sampson 1997, 2001; Farke 2004; Krauss et al. 2007,

(V and VI) in Centrosaurus. The pathological condition, acro-

this volume), and that the thickened anterior orbital rim in

megaly, was cited in Anonymous (1934) as possibly explain-

some centrosaurines may help block the nasal horncore of a

ing the evolution of ‘‘freak’’ features such as frills found in

shoving opponent (Sampson 1997: 386). Hypothesized cera-

ceratopsians. Brown and Schlaikjer (1937) noted two fused

topsian fighting behavior is often based on modern mammal

dorsal vertebrae in their study of a Styracosaurus skeleton.

models, for which caution is urged (Farke 2004).

Sternberg (1940) describes a subadult centrosaurine skull

A large sample of Albertan ceratopsian bones is now avail-

(Monoclonius lowei ) from southeastern Alberta possibly with

able for study to the contemporary paleopathologist. Begin-

an injured frill. Sternberg (1950) considered a deep circular

ning in 1979, the Palaeontology Division of the Provincial

depression on the top of a Pachyrhinosaurus skull as possibly

Museum of Alberta (which eventually formed the basis of the

abnormal. Langston (1975) posed the question that the vari-

Royal Tyrrell Museum of Paleontology [TMP]; opened 1985)

ety of pits and depressions afflicting centrosaurine skulls was

began rigorous field programs and conducted extensive ex-

‘‘. . . pathological (fungal?) corrosion and without genetic

cavations of monodominant Campanian-aged ceratopsian

basis.’’ Tyson (1977) included a short description of a de-

dinosaur bonebeds (Table 25.1); much of the material de-

pressed puncture mark with strain cracks in a radial pattern on

scribed herein was derived from these activities. The field pro-

the left squamosal of a Centrosaurus skull (UALVP 11735) and

gram was also active in locating and collecting isolated mate-

further noted an infected circular patch of surface bone on the

rial, always on the lookout for any pathological specimens.

antero-dorsal region of the left squamosal in Chasmosaurus

Most of the recovered bonebed and isolated specimens are cen-

UALVP 40. Norman (1985) and Dixon et al. (1988) posited

trosaurine, and much of this report will discuss osteopathy in

that the distinctive rugose nasal boss in Pachyrhinosaurus

that subfamily. Most of the material was collected from Centro-

might be a pathological feature, resulting from a large nasal

saurus and Styracosaurus bonebeds at Dinosaur Provincial Park,

horncore being snapped off. Rothschild (1988) reported a

southern Alberta, as well as from the Pipestone Creek Pachy-

stress fracture in a ceratopsian phalanx; the first paper focused

rhinosaurus bonebed in the west-central region of the province.

solely on Albertan ceratopsian osteopathy. Tanke (1989b) was

Chasmosaurine bonebeds in Alberta are few in number (Tanke

the first attempt to review osteopathy in the group and corre-

2007), and none have yet been systematically excavated.

late these to possible injury-incurring behaviors. These speci-

Specimens underwent gross examination in the field and

mens are reconsidered here and new pathologic material col-

TMP collections by the senior author for evidences of osteopa-

lected over the past three decades described.

thy. Specimens with paleopathologies were collected and pre-

Many of the more recent descriptions and discussions of pa-

pared (mostly by DHT). Specimens were determined as patho-

leopathologies in ceratopsids and dinosaurs have been greatly

logic by the presence (or in some cases, absence) of a ‘‘normal’’

assisted by the advent of CT- or MRI-scanning and other ad-

anatomy as deduced from an examination of numerous speci-

vanced medical technologies, as well as computer modeling

mens in TMP collections; presence of fracture callus or disrup-

and a greater understanding of osteopathy and its effects on

tive surface bone textures and other anomalous conditions.

bones (Rothschild and Tanke 1992, 1997; Rothschild and Mar-

The second author has observed the TMP dinosaur pathology

tin 1993, 2006; Dodson 1996; Rothschild 1997; Tanke and

collection several times over the past twenty years. A substan-

Rothschild 1997, 2002, 2007; Rothschild et al. 1999; Tanke

tial amount of the centrosaurine bonebed material is now pre-

and Farke 2003, 2007; Rothschild et al. 2003; Halperin 2004;

pared and housed at the TMP, making it the repository for the

Rega and Holmes 2006, Rega et al. 2007; Thompson and

world’s largest dinosaur paleopathology collection.

Holmes 2007; Holmes et al. 2007; Ryan et al. 2007).

This paper is not intended to be analytical in nature, but to

A number of ceratopsian papers and popular dinosaur publi-

present a general overview or survey of osteopathy known in

cations make reference to paleopathologies and horn arma-

Albertan Ceratopsidae to date. In some cases we do infer pos-

ment in neoceratopsians and relate these features to intraspe-

sible aspects of intraspecific behavior. Here, we have elected

cific behavior (see Tanke and Rothschild 2002 for a review). It

to organize our review of paleopathologies by skeletal re-

has been suggested that the nasofrontal boss in Pachyrhino-

gion. Some of the more interesting or problematic examples

saurus was utilized for head butting or head to head/body

are figured.

shoving matches (Farlow and Dodson 1975; Colbert 1981) in a

Institutional Abbreviations. AMNH: American Museum of

manner similar to that which has been popularly suggested in

Natural History, New York; CMN: Canadian Museum of

Late Cretaceous pachycephalosaurids (Galton 1970; but see

Nature, Ottawa; DPP: Dinosaur Provincial Park, Alberta;

Carpenter 1997) or Late Paleozoic dinocephalians (Barghusen

ROM: Royal Ontario Museum, Toronto; TMP: Royal Tyrrell

1975). It has also been suggested that enlarged orbital horns in

Museum of Palaeontology, Drumheller; UALVP: University

356 tanke & rothschild

of Alberta, Edmonton; YPM: Yale Peabody Museum, Yale University, New Haven.

Cranium ROSTRUM Prominent notches of bone missing from the cutting edge of the rostrum are known in two specimens (DPP pachyrhinosaur TMP 2002.76.1 [Fig. 25.1A; see Ryan et al. this volume], and Pachyrhinosaurus TMP 89.55.188 [Fig. 25.1B]). In the first specimen (Fig. 25.1A), an estimated 60–70 mm of the tip of the rostrum has been lost on both sides, more so on the right. On the left side, loss of part of the ventral edge of the rostrum gives it the appearance of having a small hook-like process directed ventrally. Normal rostra have a smooth internal surface, however, adjacent to the defect, the internal bone surface is rugose, with a texture similar to that found on the rostrum externally. In the second specimen (Fig. 25.1B), the notch is

FIGURE 25.1. Possible pathologic rostra in pachyrhinosaurs. (A) Mounted specimen of DPP Pachyrhinosaurus-like form in right lateral view (Ryan et al. this volume; TMP 2002.76.1); (B) Pachyrhinosaurus (TMP 89.55.188) in left lateral view. Both specimens show missing bone (arrows) in the antero-ventral margin of their rostra. Scale bar is 5 cm.

present, and the bone is smooth. Although it is difficult to determine if these features are taphonomic artifact or true osteopathy, the bone morphology

Injuries involving fractures to the nasal and supraorbital

and texture immediately adjacent to the rostra appear normal

horncores are both rare in TMP collections and the literature.

in both specimens, and the unusual internal rostral texture in

CMN 8800 includes evidence for trauma to the top of the skull

TMP 2002.76.1 suggests that they are pathologies. Interpreted

in Chasmosaurus that resulted in deformed frontal and na-

as pathologies, these features may relate to bone loss due to

sals, which, in turn, altered the nasal horncore shape (Stern-

biting tough food.

berg 1940). One adult cf. Centrosaurus nasal fragment (TMP 82.16.289; Fig. 25.3A, B) has a low, rounded dome of pathological bone instead of the normal prominent nasal horncore.

NASAL BOSS / NASAL AND SUPRAORBITAL HORNCORES

Whereas generally resembling a pachyrhinosaur nasal boss, this specimen probably represents a nasal horncore that was

Norman (1985) and Dixon et al. (1988) suggested that the

snapped off at its base in life. The dorsal surface bears an irreg-

nasofrontal boss in Pachyrhinosaurus was a pathological fea-

ular but fairly smooth and hummocky texture suggestive of a

ture caused by the breaking off of a nasal horncore. While

pseudoarticulation surface.

regarded as unusual when first discovered, this boss is now

Another nasal, TMP 2007.35.77, appears to show simi-

recognized as normal anatomy and not pathologic. Even so,

lar osteopathy suggesting premortem nasal horn loss may

TMP 89.55.899 (Fig. 25.2B, C) is the right half of an adult

not have been a rare occurrence. Recently Horner and Good-

Pachyrhinosaurus nasal boss that is unusually modified. Here

win (2008) reported examples of pathologic nasal boss de-

the sutures between the premaxilla and circumorbitals over-

velopment in several adult Triceratops. In these cases how-

lap on the external surfaces of the bones, instead of exhibiting

ever, the boss development was related to premortem loss

a standard tongue-and-groove type articulation seen in other

of the unfused epinasal, a separate ossification not found in

specimens from the Pipestone Creek Pachyrhinosaurus bone-

centrosaurines.

bed. Furthermore, along the internal surface, where rough tra-

TMP 91.36.488, an isolated subadult cf. Centrosaurus right

becular bone normally occurs, there is smooth-surfaced cor-

nasal, bears several small (largest 18 x 8 mm) openings postero-

tical bone with low ridges and deep pits. These anomalies

ventral to the posterior edge of the nasal horncore. These

suggest that part or the entire floor of the original nasal boss

openings are surrounded by bone with a normal texture, sug-

was hollow, possibly opening to the palate and oral cavity

gesting that the anomalies were benign or developmental (vas-

below. Because the cortical bone does not appear pathologic,

cular?) in origin.

however, these features may represent an extreme case of

Healed supraorbital horncore fractures have been reported

bone resorption in an adult, as documented in centrosaurine

in Triceratops where up to half the total length was lost in life

postorbitals (Sampson et al. 1997; Tanke and Farke 2007).

(Hatcher et al. 1907; Moodie 1923; Happ 2008). A less severe

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 357

FIGURE 25.2.

Pachyrhinosaurus skull and nasal. (A) Skull (TMP 89.55.1234) in left lateral view (reversed) to serve as a template to show interpreted life position of nasal (box). White arrow indicates an unusual circular raised bony lesion of unknown etiology. (B) Partial right nasal (TMP 89.55.899) in external view; (C) internal view of B. The unusual profile of the ventral edge in B and C (arrows) is the result of bone resorption and remodeling. Scale bar is 10 cm.

358 tanke & rothschild

horncore marks differ in having smooth, thickened, and raised rims encircling each mark, suggestive of reactive bone formation after the initial trauma. Such rimmed morphology is not present around normal vascular openings. Either these features represent healing bite trauma or, possibly, some other form of osteopathy. The specimen was found near the surface in a soft organic shale, so the quality of the cortical bone has been compromised with a loss of instructive bone texture.

MAXILLA Pachyrhinosaurus skull TMP 89.55.1234 bears a raised circular lesion of uncertain etiology on the thickened ramus of the left maxilla (Fig. 25.2A). This skull bears a number of other interesting pathologies (large hole in the side of the face; ‘‘punchedout lesions’’ on the medial left squamosal, and resorption of the ventral edge of the left squamosal); see Rothschild and Tanke (1997) and Tanke and Farke (2007) for more details.

SKULL ROOF Sternberg (1950) suggested that a deep circular depression atop the type Pachyrhinosaurus canadensis skull CMN 8867 represented either a folded frontal (postfrontal fontanelle) complex, present in other centrosaurines, or an abnormal feature. The presence of postfrontal fontanelles in the Drumheller P. canadensis skull (Langston 1967) and Pipestone Creek Pachyrhinosaurus bonebed materials (Currie et al. 2008) confirms the former interpretation is likely valid.

JUGAL-EPIJUGAL FIGURE 25.3. cf. Centrosaurus nasal fragment (TMP 82.16.289).

(A) Dorsal view; (B) left lateral view. Anterior is to the left. Scale bar is 5 cm.

Normally, ceratopsid jugals preserve the ventral margin of the orbit and bear prominent sutural contacts for the squamosal (posteriorly) and maxilla (anteriorly). In the right jugal (TMP

case is known in a cf. Anchiceratops supraorbital horncore tip

2002.68.46; Fig. 25.4A) a lesion-related remodeling has re-

(TMP 89.12.8). The distal 100 mm was broken and separated

sulted in modifications such that the squamosal sutural con-

from the remaining horn, but apparently was retained within

tact forms a deep longitudinal ridge medially, and the orbital

the keratinous sheath, thus forming a pseudoarthrosis. The

rim and maxilla suture are notably absent. Instead, a broad

side of the horncore shows a patch of finely pitted texture in-

curving rim arises from the front of the element at mid-height

dicative of mild osteomyelitis (Rothschild and Tanke 1992: Pit-

and arches dorsoposteriorly through about 60\. The leading

ting on tips of horncores (especially centrosaurine postorbital

edge of this rim bears a series of small but prominent rounded

horns or bosses) have occasionally been assessed as pathologic.

and bluntly pointed exostoses along its length. The bone is

However, in centrosaurine postorbitals these pits appear to be

thickened throughout. The anterior edge (rim) is also greatly

related to advanced age and non-pathological bone loss/re-

thickened, up to 31.5 mm compared to 17 mm in a normal,

modelling (Sampson et al. 1997; Tanke and Farke 2003, 2007).

adult-sized jugal (TMP 2002.68.41; Fig. 25.4B) from the same

The description of Eotriceratops (Wu et al. 2007) includes

bonebed. The great degree of modification and bone remodel-

notation of several ‘‘bite mark’’ traces clustered at the external

ing makes it impossible to determine the etiology.

base of the left supraorbital horncore. These differ from clas-

Centrosaurine epijugals sometimes show resorption pitting

sic, large-theropod tooth marks in not resembling a depressed

on their external surfaces. These occur as small, shallow ([3

puncture mark or tooth-drag furrows (Erickson and Olson

mm deep) circular depressions with a floor texture resem-

1996; Chure et al. 1998; Jacobsen 1998). The Eotriceratops

bling trabecular bone (Pachyrhinosaurus TMP 89.55.1091;

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 359

FIGURE 25.4. Centrosaurus jugals in external views. (A) Right pathologic (TMP 2002.68.46); (B) left normal (TMP 2002.68.41 [reversed]). Scale bars are 10 cm.

non-pathologic?) or as deep (20 mm), relatively smooth-

Farke 2007). The peg growth may have caused difficulties dur-

floored circular depressions (Pachyrhinosaurus TMP 86.55.304;

ing mastication, as this feature would have been in close con-

Fig. 25.5). The latter example appears to be an extreme case of

tact with the articular. Continued growth may have caused

bone resorption pitting (cf. Centrosaurus postorbitals; Samp-

eburnation, though there is no trace of grooving or polishing

son et al. 1997; Tanke and Farke 2007) and possibly not

of the bone surface. The second specimen (TMP 89.55.1072) bears two patholo-

osteopathy.

gies. The first is a circular depression on the articulating surQUADRATE

face of the medial condyle. The mildly rugose bone texture on the floor of the depression suggests an origin due to infection.

Two pathological quadrates are known in Pachyrhinosaurus.

The second pathology is a crack, up to 10 mm deep, that runs

TMP 87.55.101 (Fig. 25.6A, B) exhibits three anomalies. The

diagonally between the two distal condyles. At midlength,

distal articulating surface, normally relatively smooth and

a thin wall of bone separates the crack into two sections. This

convex (Fig. 25.6C), is strongly concave and moderately ru-

feature is similar to recent bovid phalangeal cracks of an un-

gose. Secondly, a low, blunt and smooth bony peg protrudes

determined but presumed non-pathological circumstance

ventrally from the articular surface of the medial condyle. The

(Baker and Brothwell 1980). Alternatively, the crack may rep-

third feature occurs above this on the anterior face and con-

resent a fracture with advanced healing and bone remodeling.

sists of a roughly circular pit, 20 mm in diameter and up to 13 mm deep with a roughly textured floor. These features probably related but their etiology is unknown. The pathology has

SQUAMOSAL

the appearance of resulting from something that pushed into

In Pachyrhinosaurus skull TMP 86.55.111, a smooth, flattened

the quadrate (forming the pit), and possibly causing the peg to

lobe of bone 40 mm long and 37 mm wide, with a rounded

extrude ventrally. The bone texture in the floor of the pit is

distal end, arises from the underside of the antero-ventral

similar to that seen in ceratopsian resorption pits (Tanke and

edge of the left squamosal. Viewed from the side, the lobe

360 tanke & rothschild

protrudes for a distance of approximately 20 mm below the ventral edge of the squamosal. The bone surfaces of this feature and the surrounding bone are normal. Thus, it may have resulted from anomalous growth or an avulsion injury. Partial Pachyrhinosaurus skull TMP 89.55.1234 preserves both squamosals, the left one of which shows severe resorption of the ventral margin. There is other osteopathy on this skull (Rothschild and Tanke 1997; Tanke and Farke 2007), but the relationship it has with the squamosal is unclear. Centrosaurus squamosal TMP 90.36.411 exhibits a similar marked loss along the ventral margin (Tanke and Farke 2003, 2007). TMP 81.16.362, an isolated adult centrosaurine left squamosal fragment, has pathological features in the form of roughened, disrupted bone surface at and near the ventral margin that is suggestive of osteomyelitis. In this specimen the ventral edge has also been resorbed and the exoccipital wing from the braincase has fused to the squamosal, an atypical condition. A small number of other centrosaurine squamosals with osteopathy or pathology-like features are known, but these are described more fully in Tanke and Farke (2007). The osteopathy affecting many of these squamosals resembles that seen on a Triceratops squamosal from South Dakota (Farke and Alley 2006). In that case, a previous injury with subsequent bone loss to the ventral margin is implicated. Pachyrhinosaurus jugal fragment with attached epijugal (TMP 86.55.304) in external view. The epijugal exhibits extreme bone resorption pitting (arrows). Scale bar is 5 cm. FIGURE 25.5.

In Albertaceratops, Ryan (2007: 383) noted an anomalous deep ovoid foramen on the external surface of the left squamosal in TMP 2001.26.1 posterior to the jugal process. The lesion includes a ‘‘small drainage channel.’’ He did not offer an interpretation, but concluded that the lesion did not resemble other lesions seen in ceratopsian squamosals. Our inspection of the specimen reveals that immediately dorsal to the lesion is a shallow three-sided pit that resembles ‘‘punched out lesions’’ or resorption pitting seen in other ceratopsian skulls (Tanke and Farke 2007). Also, the proximal blade of the quadrate is exposed, but that element itself is not affected by the new opening. The close proximity of this lesion to the one noted by Ryan (2007) suggests that the latter is a resorption pit that fully penetrated the bone. Lull (1933) and Farlow and Dodson (1975) have suggested cranial injuries in centrosaurines would be borne upon the squamosals from horn thrusts directed by intraspecific aggressors, but the present rarity of such specimens would indicate otherwise. Squamosals, many in perfect condition, are common occurrences in centrosaurine bonebeds and none show any indication of a pathological condition attributable to agonistic behavior. A tour of major North American museums

Pachyrhinosaurus quadrates. (A) Anterior view of left quadrate (TMP 87.55.101) showing protruding peg of bone (black arrow) and circular pit (white arrows); (B) ventral view of TMP 87.55.101 showing anomalous concave surface; (C) ventral view of left quadrate (TMP 87.55.77) showing normal convex morphology of articulating surface. Scale bars are 5 cm. FIGURE 25.6.

containing most of the collected centrosaurine material also did not reveal any squamosal injuries (M. Ryan and S. Sampson pers. com.). Pathological squamosals in chasmosaurines are also uncommon. A number of Chasmosaurus skulls and others in the subfamily bear supernumerary fontanelles that

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 361

FIGURE 25.7. Two normal adult centrosaurine posterior parietals in dorsal view showing symmetry. (A) Centrosaurus (CMN 971) modified slightly from Hatcher et al. (1907: plate 24) with parietal processes numerated and reconstructed left side; (B) Pachyrhinosaurus (TMP 89.55.141). Scale bar is 10 cm.

fully pierce the squamosals, with centrosaurines also showing

FIGURE 25.8. Centrosaurus (TMP 64.5.191) posterior parietal in dorsal view showing pronounced asymmetry, the most observed in this study. The long black line indicates the midline. Scale bar is 10 cm.

such openings affecting various regions of the skull. Examination of these lesions suggests that whereas some may truly reflect intraspecific strife (see Farke 2004 for a discussion in

to show a transverse fracture across the element some 22 cm

Triceratops) many others are the result of disease or bone re-

posterior to its point of insertion into the skull. Finished

sorption (Tanke and Farke 2003, 2007) rather than intraspe-

bone occurs in places along the broken edge that joins the up-

cific horn thrust injury (Rothschild and Tanke 1997).

per and lower sides of the element, suggesting long-standing pseudoarthrosis. An adult (TMP 64.5.191; Fig. 25.8) exhibits a

PARIETAL

pathological condition showing severe bilateral asymmetry with the posterior ramus of the element 40\ from the perpen-

A number of parietal pathologies occur and are treated here by

dicular. In addition, the left side of this bone thins rapidly and

genus. The normal condition (Fig. 25.7) for an adult centro-

is possible evidence of atrophy.

saurine parietal is as follows. The medial parietal bar is posteri-

Styracosaurus. Holmes et al. (2007) and Ryan et al. (2007)

orly directed ending in a greatly thickened, transverse dorsal

note an unusual condition affecting the bases of several left

ramus (90\ to the medial bar) at the back and top of the frill.

frill spikes (P4 and P5) in the type specimen (CMN 344). Here

This ramus bears a complex variety of horns, hooks, spikes

the spikes are co-joined and partially overlapping at their

or hornlets. Also present are two anteriorly directed lateral

bases. The authors suggest a fracture of the parietal in life re-

rami, which bear a variety of spikes, hornlets, or epiparietals

sulting in the base of P4 overriding the base of P5.

(revised epoccipital terminology sensu Horner and Goodwin 2008).

Pachyrhinosaurus. TMP 89.55.883 is a parietal bar from an adult collected from the Pipestone Creek Pachyrhinosaurus

Centrosaurus. Parks (1921) noted a ‘‘step’’ in the lateral ra-

bonebed. Adult-sized proximal parietal bars from that site

mus of the parietal in ROM 767, which was correctly diag-

bear 1–3 upward-pointing horns on the dorsal midline (Cur-

nosed by him as pathologic. This lesion may represent a well-

rie et al. 2008). TMP 89.55.883 differs in lacking a horn and

healed fracture. An adult parietal bar (TMP 66.35.1) appears

preserving a single low rounded ‘‘boss.’’ It is presently un-

362 tanke & rothschild

FIGURE 25.9. Pachyrhinosaurus posterior parietals in dorsal view exhibiting asymmetry. (A) TMP 87.55.210; (B) TMP 89.55.1085, note unusual curvature of the right P3 process; (C) TMP 89.55.1503, note asymmetry of right P3 spike. Scale bars are 10 cm.

known if this deviation represents osteopathy, extreme re-

similar to the Centrosaurus example noted above (Fig. 25.8).

sorption of the dorsal regions of a previous horn, or simply

The first (TMP 87.55.210; Fig. 25.9) lacks the left P3 frill spike

reflects individual variation.

(see Fig. 25.7A for an explanation of frill ornamentation num-

Two subadult parietals of Pachyrhinosaurus exhibit patho-

bering). Viewed anteriorly, extensive premortem bone remod-

logical features. TMP 88.55.90 was apparently affected by an

eling resulted in a strongly asymmetrical frill (Fig. 25.9A), with

unidentified disease (osteomyelitis?), which eroded a large

the transverse ramus tilted from its normal position perpen-

patch (up to 7 mm deep) into the dorsal and right lateral sur-

dicular to the medial bar, and thus forming angles of 38\–142\

faces of the median parietal bar. TMP 89.55.125 consists of

relative to the medial bar. Also viewed anteriorly, the left side

the midline posterior parietal bar section with a transverse

of the frill terminates in a smooth rounded point, suggesting

fracture through the parietal bar, some 130 mm anterior to the

a pseudoarticulation (Fig. 25.9A). Ventrally, two round and

posterior edge. Along this break, instead of continuing on a

roughly textured scars mark the P3 horncore base and contact

straight plane, the bone flares ventrally and to a lesser de-

point for the left lateral ramus of the parietal. These two scars

gree dorsally and has smooth rounded edges. The face of

may represent pseudoarticulation surfaces, but owing to the

the break shows a smooth-textured and undulating surface,

incompleteness of the specimen this cannot be confirmed.

nearly identical in texture to pseudoarthrosis surfaces on

The presence of a P3 frill spike base in this specimen leaves no

hadrosaur caudal neural spines in the TMP collections, sug-

question that the parietal was injured when the individual

gesting an improperly healed fracture or pseudoarthrosis

had already attained adult size.

across this element. The injury was apparently noninfectious

The other parietal (TMP 89.55.1085, Fig. 25.9B) includes

at the time of death, although ventrally a 20 mm wide texture

both P3 frill spikes. The left one exhibits the usual outward

change is apparent posterior to the break.

and forward curvature and circular cross section, but the right

Two adult Pachyrhinosaurus parietals apparently suffered

P3 horn (now a flattened oval in cross section) points dorsally

massive fracture injuries and unusual post-trauma healing,

and then curves anteriorly; that side of the frill is also asym-

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 363

metrical, measuring 15\–105\ relative to the medial bar. Both parietals have unusual shapes, but no other anomalies are evident (e.g., extra growths, diseased bone, or disrupted bone surface texture), suggesting that the injuries occurred a long time prior to death. Finally, an adult Pachyrhinosaurus parietal (TMP 89.55.1503; Fig. 25.9C) shows a normal left P3 horn and a right P3 of markedly larger and different shape. Minor asymmetry is suggested, but it is not as noticeable as in the previously described specimens. It is unknown if this specimen reflects well-healed osteopathy, a developmental anomaly or individual variation. Specimen of Uncertain Taxonomic Position. The subadult centrosaurine skull, CMN 8790, described by Sternberg (1940) as Monoclonius lowei shows possible trauma to the right side of the parietal. The external border is malformed and a thin sheet of bone has grown over and completely covered the parietal fenestra on that side. However, such damage might be the result of post-mortem crushing (S. Sampson and M. Ryan pers. com.). As the frill contributes much to the overall shape of the skull, frill asymmetry as described above would have imparted to the individuals a most unusual appearance (see also Farke and Alley 2006). In each case it appears that post-trauma movement of the bones (mediated by soft tissue) shifted portions of the frill into these unusual positions, where the bone then healed. Most current students of the Ceratopsia would agree that the frill probably served multiple functions in life. Much attention was raised by earlier workers who stressed the elongated frills main function as attachment points for long and powerful jaw musculature (Lull 1908; Russell 1935; Haas 1955; Ostrom 1964; Currey 1984). If true, these parietal traumas would severely impact mastication. However, revised interpretations indicate shorter jaw muscles with more anterior insertion positions on the parietal (Dodson and Currie 1990; Dodson et al. 2004: 512). Such revisions imply that even severe frill fractures may not have impacted the jaw muscles.

FIGURE 25.10. cf. Centrosaurus right dentary (TMP 2007.20.51). (A) Internal view with inset indicating area shown in (B); (B) vertical bone growth obscuring teeth. Scale bar is 1 cm in (A) and 5 cm in (B).

MANDIBLE-TEETH Osteopathy affecting the mandible of ceratopsids is known

posteriorly on either side of the main lesion, but the etiol-

from several specimens. TMP 2002.68.165 is an adult Centro-

ogy remains uncertain. This possibly represents a well-healed

saurus right dentary bearing a low, slightly swollen area just

greenstick fracture.

posterior to the predentary articulation and below the tooth

An adult right dentary (cf. Centrosaurus, TMP 2007.20.51;

rows. The lesion has several low ridges that run parallel to the

Fig. 25.10) is unusual in having a low smooth lobe of bone

long axis of the bone and have a distinctive filigree texture.

arising from the ventral edge of the tooth row, which then

This appears to represent a low-grade bone infection, active at

extends antero-dorsally and completely blocks tooth rows 6

the time of death. TMP 95.12.144, an adult Centrosaurus right

and 7. The etiology of this unusual condition, never seen be-

dentary, bears a low longitudinal ridge of bone on the ventral

fore in ceratopsids, is unknown.

edge, directly below the anteriormost tooth positions. Exter-

The coronoid normally articulates without fusion to the

nally the bone is slightly swollen at the wound site. Small,

dentary (Brown and Schlaikjer 1940). The adult left dentary,

rugose patches of bone are present internally (anterior) and

TMP 2005.12.566, differs in showing the anterior half of the

364 tanke & rothschild

FIGURE 25.11.

Pachyrhinosaurus cervical vertebra (TMP 88.55.52). (A) Posterior view showing ‘‘moth-eaten’’ appearance of ventral centrum endplate (white arrows) and prominent bone spur growth projecting ventrolaterally (black arrow); (B) proximal view showing unaffected centrum endplate and malformed right prezygapophysis (white arrows). Scale bar is 5 cm.

coronoid fully fused to the coronoid process of the dentary.

being an aid in carrying the great weight of the head. In the

However, there is no anomalous bone growth or texture to

Pipestone Creek Pachyrhinosaurus sample, normal syncervi-

explain the etiology of this condition.

cals are known but others with incomplete fusion are also

A mid-ramus fragment of an adult right dentary, TMP

represented. To date, prepared specimens include isolated first

2005.9.8, exhibits a broadly arching draining sinus internally.

cervicals (TMP 87.55.315), fused first and second cervicals

The surrounding bone and surface texture appears normal.

(TMP 86.55.217) and isolated third cervicals (TMP 86.55.53,

Externally, the bone exhibits a large, semicircular swelling

TMP 86.55.275, TMP 87.55.306). A ceratopsid from DPP (TMP

with a finely pitted texture. An elliptical opening (19 — 30 mm

91.36.263) consists of an isolated first cervical. The cause or

wide and 27 mm deep) is directed antero-ventrally and hosts a

significance of these incomplete cervical fusions is presently

prominent drainage tract. This lesion appears to represent a

unknown. The fact that some adult individuals of Pachyrhino-

case of massive osteomyelitis, active at the time of death.

saurus did not have completely fused syncervicals casts some

Dental caries (tooth decay) are unknown in any of the abundant teeth from bonebed or vertebrate microfossil assem-

doubt on the long assumed strength hypothesis ascribed to them (Campione et al. 2008).

blages. As in most extant reptiles, the teeth of ceratopsids were

Two fused cervicals (V and VI) in Centrosaurus YPM 2015

frequently and constantly replaced throughout their lifetime

were noted by Lull (1933) but were not discussed further. Two

(Edmund 1960, 1969). Thus, ceratopsid teeth were apparently

pathological cervical vertebrae also are known for Pachyrhino-

worn down and expelled before decay could set in.

saurus. TMP 89.55.978 consists of an isolated neural arch with

Malocclusion, malformed teeth, tooth fractures, and color

the right prezygapophysis misshapen and several concentric

banding on teeth are interpreted as evidence of nutritional

folds or ridges occurring external to the zygapophyses on the

stress in contemporaneous tyrannosaurids, hadrosaurs, and

dorsal side of the transverse process. The other (TMP 88.55.52;

crocodilians, but are unreported in ceratopsians.

Fig. 25.11) is similar in having a reduced and malformed zygapophysis, but also exhibits areas of deep pitting on the centrum endplate. This does not appear to be a taphonomic arti-

Postcrania

fact, as the pitting exhibits finished bone margins and the

CERVICAL VERTEBRAE

opposite endplate is unaffected (Fig. 25.11). Arising from the posterior endplate margin, a bluntly pointed spur-like growth

In all large ceratopsians, it is normal for the first three cervicals

of bone is directed latero-ventrally. These conditions appear

to be completely fused into one solid unit (Campione 2005;

to be disease-related.

Campione and Holmes 2006; Tsuihiji and Makovicky 2007;

An isolated DPP ceratopsian cervical vertebra (UALVP

Campione et al. 2008). This feature is generally accepted as

47974) of unknown position (but posterior to the syncervical)

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 365

trauma as a source; it is not a pathologic fracture occurring in response to another cause such as disease or tumor.

DORSAL VERTEBRAE Fusion of dorsal vertebrae IV-V (not cervicals IV–V as reported in Dodson 1996: 168) was noted in Styracosaurus AMNH 5372 by Brown and Schlaikjer (1937). These authors cite this and the presence of fused cervicals in Centrosaurus YPM 2015 as being ‘‘not uncommon’’ conditions in large ceratopsians, but failed to provide evidence of other fused specimens to support the claim. Fusion of dorsal vertebrae is evidently a rare phenomenon. The senior author has excavated or examined hundreds of ceratopsian trunk vertebrae from ceratopsian bonebed deposits and museum collections, but has seen only two other examples of fusion, each involving two vertebrae. Vertebrae in TMP 66.21.51 (Fig. 25.12), are thoroughly fused. Viewed from the right side one centrum is apparent, but when viewed from the left, two fused centra are indicated. These two centra are separated by a groove dorsally but fused by a smooth-textured bony mass across the ventral surface. Because there is no evidence of trauma, it is possible that this arrangement of the vertebrae represents a congenital defect. TMP 92.36.715 (Fig. 25.13) includes two fused dorsal centra. The lateral and ventral surfaces of the centra were completely fused through the ossification of lateral and ventral ligaments. Only the floor of the neural canal and areas anterior and posterior to the neural arch bases are unfused, but even here low, flat-topped bone spurs arising from the centra endplates grew towards one another. When found, the specimen was broken revealing centra endplates, both of which exhibit an unusual radiating bone texture. This condition resembles the diffuse idiopathic skeletal hyperostosis observed in sauropod tails (Rothschild and Berman 1991), but spondyloarthropathy could also be responsible. Two fused ceratopsian dorsal vertebrae (TMP 66.21.51) in left lateral view. Note large ventral bone growth (arrow). C: centrum; NS: neural spine. Scale bar is 10 cm. FIGURE 25.12.

Brown and Schlaikjer (1937) suggested presacral vertebral fusions might be related to the development of a large head, but the rarity of such fusions argues against this hypothesis. TMP 91.36.92 (Fig. 25.14) is an isolated and undescribed ceratopsian dorsal centrum. It preserves a large patch (105 — 450 mm) of rugose bone on one side. Although this may be due to infection, no draining sinuses are present.

shows compelling evidence of the individual having suffered

Some dorsal and caudal vertebrae exhibit extra growths on

and survived a broken neck. A deep crack is present on the one

and near the centrum articular faces ( Johnson and Storer 1974:

preserved centrum endplate. This feature extends from the

fig. 117). Typically, these growths are low concentric ridges

floor of the neural canal ventrolaterally, narrowing and ter-

(TMP 79.8.345) covering most of the articular face, or appear as

minating before it reaches the edge of the centrum. On the

a low hummock-like surface (TMP 81.18.38). Such growths

reverse side of the specimen, trabecular bone is exposed. The

seem to be confined to the largest and presumably oldest indi-

trabecular pattern is normal, but where the crack continues

viduals. Despite their unusual appearance, these structures are

through the body of the centrum the trabecular bone shows a

not considered to be truly pathological. Instead they probably

high degree of inter-fingering, demonstrating that the crack

represent accentuated attachment surfaces for intervertebral

is a healing fracture. A healing fracture indicates external

cartilaginous disks or are related to advanced age.

366 tanke & rothschild

bonebed account for about half of the entire paleopathology sample from this locality. Based on size, cross-sectional shape, and relative lack of curvature, two of these (TMP 85.112.52, TMP 89.55.719) are identified as being adult and mid-dorsal in position. Both show single, simple healing fractures forming a smooth callus without infection. A juvenile left dorsal rib (TMP 89.55.464) shows a smooth and slightly swollen bulge on the ventral surface of the rib head neck, suggesting a minor greenstick fracture. Pathologies to posterior dorsal ribs are common. Only a few are discussed here, and they include single non-infectious healing breaks forming a smooth callus on (Pachyrhinosaurus: TMP 88.55.191, Fig. 25.15A; TMP 89.55.389, Fig. 25.15B; TMP 89.55.1481), multiple breaks with subsequent infection and imperfect healing forming a pseudoarthrosis (Pachyrhinosaurus: TMP 87.55.90), and single non-infectious breaks imperfectly healed and forming pseudoarthroses (Pachyrhinosaurus: TMP 89.55.205, TMP 89.55.801; Centrosaurus: TMP 91.18.77, Fig. 25.15C). TMP 89.97.1, an associated Styracosaurus specimen, includes a right posterior dorsal rib in two sections, showing an imperfectly healed fracture forming a pseudoarthrosis. This is the only known Albertan ceratopsid specimen where both halves of an affected bone are preserved together with their intervening pseudoarthrosis. Curiously, where ribs are broken in two and form a pseudoarthrosis only the proximal portion has ever been found in isolated specimens. All posterior dorsal rib fractures seem to be located below the tuberculum, usually about 10–30 mm distal to this process. Many more non-pathological posterior dorsal ribs have FIGURE 25.13. Two fused dorsal vertebrae (TMP 92.36.715). (A) Lateral view; (B) end-on views. Sidedness cannot be determined. Scale bars are 10 cm.

been observed (but not quantified) in the Pipestone Creek Pachyrhinosaurus bonebed (DHT pers. obs.), indicating that although broken and healing ribs account for about half of the entire osteopathy in this element, they are still in fact uncommon. Many of the non-pathological ribs in the monodomi-

McGowan (1983) noted bony outgrowths on ‘‘many’’ of the

nant centrosaurine bonebeds were broken due to taphonomic

vertebrae in the mounted skeleton of Chasmosaurus ROM 843,

processes such as trampling (see Fiorillo 1988 and citations

but examination of a cast of this specimen on display at the

therein), either by scavenging theropod dinosaurs or other

TMP failed to disclose any such features.

dinosaurs moving through the area after skeleton disarticula-

A dorsal neural arch fragment from Styracosaurus (TMP

tion (Eberth and Getty 2005). These breaks occur in the same

90.58.1) exhibits coarse pitting on the right postzygapophy-

areas as the healing fractures in the pathological specimens.

sis, possibly related to spondyloarthropathy.

The combination of paleopathologies and trampling breakage in the same position on posterior dorsal ribs suggests an intrinsic weakness or vulnerability to breakage in this area.

SACRAL VERTEBRAE

Rib pathologies also known to occur in Albertan chasmo-

Osteopathy has not been reported or observed in this region.

saurines. In Chasmosaurus ROM 843 (cast TMP 82.52.2), left dorsal rib 10 exhibits a number of pathologies. The neck of

CERVICAL, DORSAL, AND SACRAL RIBS

the rib head is represented by an amorphous rounded mass (McGowan 1983). The rib shaft is also broken in half at mid-

Osteopathy of the cervical and sacral ribs has not been ob-

length and imperfectly healed. At the fracture site, the distal

served in ceratopsid dinosaurs. However, dorsal ribs regularly

end of the upper rib portion has extended a short distance

show injuries in the form of healing fractures. Healing rib frac-

posteriorly and terminates with a flared and roughened plate

tures in specimens from the Pipestone Creek Pachyrhinosaurus

of bone about 62 mm tall that contacts the anterior edge of

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 367

FIGURE 25.14.

Isolated dorsal centrum (TMP 91.36.92) lateral views. (A) Osteopathic side; (B) normal side. Scale bar is 10 cm.

rib 11. Rib 11 also bears a roughened patch about 70 mm in

forming a smooth callus around the neural spine. Imperfect

length along its anterior edge in response to the flared end of

healing in the second specimen repositioned the neural spine

rib 10. What became of the distal half of rib 10 is unknown; it

slightly left. Massive post-trauma infection had occurred in

may have been resorbed in life or lost premortem.

this specimen, resulting in the formation of large irregular

No anterior dorsal ribs were found to have been healed or

cavities at the base of the neural spine anteriorly, ragged bone

exhibiting healing fractures. This could be due, in part, to

replacing the normally smooth articular faces of the prezyg-

their shorter length, thicker diameters, and the covering and

apophyses, distortion of the usually round neural canal to an

protective nature of the scapulo-coracoid complex (e.g., TMP

oval shape when viewed anteriorly, and presence of finely tex-

87.55.190, noted below), plus the accompanying musculature

tured bone suggestive of a mild infection on the dorsal half of

in this region. Other contributing factors may be the mor-

the anterior centrum face.

phology of the ribs and body shape of the animal. Straight-

The third example (TMP 87.55.102; Fig. 25.16), consists of

shafted anterior dorsal ribs formed a tall, narrow chest while

four fused caudal vertebrae from close to the tip of the tail. The

posterior dorsal ribs with their greatly curved shafts protruded

vertebrae have fused in a straight line, with partial lines of

far laterally from the vertebrae and formed a barrel-shaped

separation between the vertebral bodies. The second neural

abdomen (Lehman 1989). As a result, posterior dorsal ribs

arch is present, but the others appear to have been broken off

were more likely to be impacted.

at the time of injury. The pathology is strikingly similar to a set of three fused vertebrae in a Pleistocene crocodile reported by

CAUDAL VERTEBRAE, CHEVRONS, AND TENDONS

Moodie (1926), which is an example of spondyloarthropathy (BMR pers. obs.). It is also reminiscent of pathologically fused

Osteopathy of the tail is common. This is also the case in other

distal caudal vertebral injuries in contemporaneous hadro-

contemporaneous dinosaur groups where the tail is mobile

saurs, where intraspecific trampling is suspected, although

and elongate (i.e., hadrosaurs, tyrannosaurs). Three patho-

none of the endplate ‘‘cracking’’ phenomenon is present. It is

logical caudal vertebrae specimens were recovered from the

possible that the injury in TMP 87.55.102 was the result of the

Pachyrhinosaurus bonebed. The first two, a proximal caudal

tail being stepped on by a conspecific. Farke and O’Connor

(TMP 89.55.363) and mid-caudal (TMP 89.55.287), have bro-

(2007) described a set of pathologically fused distal caudal ver-

ken and healed neural spines, broken just above the zyg-

tebrae in the Late Cretaceous abelisaur Majungasaurus. Five

apophyses. In the former, the break healed without infection,

vertebrae are affected, with the posterior three fused. The

368 tanke & rothschild

FIGURE 25.15. Centrosaurine osteopathic dorsal ribs. (A) Right posterior rib (Pachyrhinosaurus, TMP 88.55.191) in anterior view showing fracture callus (arrow); (B) left posterior rib (Pachyrhinosaurus, TMP 89.55.389) in anterior view showing fracture callus (arrow); (C) left posterior rib (Centrosaurus, TMP 91.18.77) in anterior view showing massive pseudoarthrosis (arrow). Scale bars are 10 cm.

fused vertebrae exhibit dynamic spicular bone growth and

mal caudal vertebra with destruction (osteolysis?) of the prox-

have numerous small openings that are likely draining sinuses

imal endplate. Though it is not fused to the preceding ver-

related to osteomyelitis. The tapering distal end of the pos-

tebra, the morphology and texture of the pathologic bone

teriormost vertebra in this specimen is suggestive of a trau-

is very similar to that of the two fused dorsal centra (TMP

matic amputation of the distal tail followed by serious infec-

92.36.715) noted above.

tion, which was evidently still active at the time of death. In

Both endplates of an isolated Centrosaurus subadult proxi-

TMP 87.55.102, no signs of osteomyelitis are present, and the

mal caudal centrum (TMP 95.400.111) exhibits perforations

endplate of the distalmost centrum is of normal morphology,

by up to nine small, circular holes of unknown etiology. TMP

which suggests that the remainder of the tail was still attached

80.16.1348 is a mid-caudal vertebra with the eroded remnants

and that the individual was not seriously affected by a some-

of the chevron thoroughly fused on. No associated osteopathy

what stiffened tail.

or reactive bone nearby occur to help to explain this condition.

TMP 98.93.77 (Fig. 25.17) is an isolated Centrosaurus proxi-

TMP 91.50.74 (Fig. 25.18) consists of three fused adult mid-

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 369

FIGURE 25.16. Four fused distal caudal vertebrae (Pachyrhinosaurus, TMP 87.55.102). (A) Left lateral view; (B) dorsal view. Scale bar is 5 cm.

caudal vertebrae. It is evident that the chevrons were also incorporated into this fused bone mass, even though the specimen is somewhat eroded by transport, with the distal parts of the chevrons, neural spines, and transverse processes removed by erosion. The centra are thoroughly fused by massive bone

Centrosaurus proximal caudal vertebra (TMP 98.93.77) in anterior view showing endplate destruction, and low, flat-topped osteophytes on centrum margin at the 1–3 o’clock positions. Scale bar is 5 cm.

FIGURE 25.17.

growth except where the ventral floor of the neural canal is located. Due to the lack of involvement of the neural canal floor, fusions between the vertebrae appear to be the result of

ably older) adult animals, but this individual, while approach-

ossification of the lateral ligaments. A crack runs through

ing adult size, was still immature as demonstrated by the in-

the extra bone growth resulting in the posteriormost vertebra

complete development of its parietal horn ornamentation

being separate from the preceding pair. This fortuitous break

(Sampson et al. 1997; Ryan et al. 2007). This example of fusion

allows a view of the centra endplates, which appear to be un-

may reflect spondyloarthropathy as DISH (Diffuse Idiopathic

affected by the exuberant bone growth and fusion occurring

Skeletal Hyperostosis) affects adult individuals (Rothschild

nearby. Accurate diagnosis of fused dinosaur vertebrae is some-

and Berman 1991). Other features that would help confirm

times difficult. This specimen has the appearance of diffuse

this diagnosis are the conditions of the annulus fibrosus or

idiopathic skeletal hyperostosis, which would tend to be con-

zygapophyseal articulations. The former are hidden by extra

firmed by its adult status (Rothschild and Berman 1991) but

bone growth or matrix. The zygapophyses are unaffected by

the presence of fused chevrons is suggestive of spondylo-

erosions, though the right prezygapophysis on caudal 29 was

arthropathy.

lost premortem; only a slightly roughened base remains. The

TMP 89.97.1, a partial Styracosaurus, includes a nearly com-

centra themselves appear to be unfused, but matrix fills the

plete tail. The base of the tail is obscured by matrix and over-

intervertebral spaces, obscuring critical details. This specimen

lying bones, but the first visible vertebra with a caudal rib and

will need to be X-rayed or CT-scanned for disclosure of internal

chevron appears to be the first caudal. If correctly interpreted,

aspects, which would likely yield a final diagnosis. It appears

44 vertebrae are preserved. Caudals 27–30 (Fig. 25.19) and

similar to the osteopathy in TMP 87.55.102 noted above.

their chevrons are thoroughly fused and form a swollen mass.

Several distal caudal vertebrae exhibit anomalous characters

Fused vertebrae are usually found in the largest (and presum-

involving deep infolded cracks with rounded edges on the

370 tanke & rothschild

FIGURE 25.18.

Three fused mid-caudal vertebrae (TMP 91.50.74) in ventral view; anterior to the right. Scale bar is 1 cm.

TMP 93.109.6 bears a prominent horizontal and infolded crack at mid-height of the proximal endplate. The anomaly is present across the entire width of the centrum and reaches a depth of 5 mm. The crack demonstrates no associated osteopathy, suggesting that the wound was attained and remodeled long before death. These caudal centra with their endplate ‘‘cracking’’ phenomenon are reminiscent of similar osteopathy affecting the fractured and healed cervical vertebra UALVP 47974 noted above. They are especially comparable to numerous examples of endplate cracking in hadrosaur mid to distal caudal centra, where conspecific trampling resulting in a crush or burst fracture is implicated (Tanke 1989a; Monastersky 1990). Thus, these fused caudal specimens likely represent accidentally received injuries. Many ceratopsians were probably herding animals (as inferred by the bonebed occurrences), and it is not hard to envision an individual resting on the ground and having its tail accidentally stepped on and injured by an adjacent heavy conspecific. A Centrosaurus extreme distal caudal vertebra (TMP 87.18.27) has what is interpreted as a congenital defect consisting of two intersecting infolded ‘‘cracks’’ forming a 90 degree angle on the proximal articular face of the centrum. The Four fused caudal vertebrae (#s 27–30) and chevrons (Styracosaurus, TMP 89.97.1). (A) Right lateral view; (B) dorsal view. Scale bar is 5 cm. FIGURE 25.19.

congential defect is suggested because of similar infolding in contemporaneous hadrosaur distal caudal centra as noted above, but in those specimens (even well-healed ones) an associated osteopathy such as centrum deformity (asymmetry), endplate destruction, centra fusion, offset neural arch base, or

centrum endplates. TMP 2000.12.37 consists of two fused

other anomalies are present. TMP 87.18.27 does not exhibit

mid-caudals and the fused remnants of a chevron. A promi-

any associated anomalies.

nent L-shaped crack is present on the proximal endplate of the

Fusions involving the terminal caudal vertebrae are known

larger vertebra, and a circular pit 10 mm deep is present on the

in several specimens. TMP 94.12.429 consists of two, minute,

distal endplate of the smaller one. No infection is indicated.

fused distal caudals from very near the end of the tail. An

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 371

identical occurrence is known in the partial DPP adult cen-

tion of distal articular surfaces of non-ungual phalanges and

trosaurine skeleton, TMP 82.32.1, which also has an unin-

distal (tip) loss or resorption of an ungual. Premortem loss of

jured tail. Neither of these specimens show signs of active or

some distal phalangeal elements seems likely. One hind foot

past osteopathy, suggesting that fusion of the extreme tip of

also shows similar lesions, and since all lesions were active at

the caudal series is normal in this taxon. Similarly, Gilmore

the time of death, they possibly crippled or contributed to this

(1917) noted the complete fusion of the last three caudal ver-

animal’s demise.

tebrae in an uninjured tail of ‘‘Brachyceratops,’’ an immature centrosaurine from Montana. Several examples of healing fractures of tendons are known

PELVIC GIRDLE

in purported hadrosaur specimens in TMP collections (DHT

TMP 96.176.36, a subadult centrosaurus pubis suffered a frac-

pers. obs.), but no equivalents are known among Ceratopsia.

ture at midlength through the anterior blade. A prominent finely pitted fracture callus covers the wound site and extends

SHOULDER GIRDLE Injuries to the pectoral girdle in centrosaurines are rare, with

up to 85 mm across the element on either side of the fracture. The callus was evidently loosely attached as it has flaked away from the bone in several places.

only one known in an adult Pachyrhinosaurus left scapula. Externally, this bone (TMP 87.55.190) bears two smooth, low rounded co-joined calluses on the thickened ridge (scapular

HINDLIMB

spine) that bisects the scapular blade, the first measuring 32 —

Injuries to non-pedal hindlimb elements are only known

22 mm and the larger 66 — 43 mm. The floor of the larger

from two examples. TMP 89.18.108 is the distal half of an

lesion is finely pitted and suggestive of a mild infectious pro-

adult Centrosaurus fibula, which apparently suffered massive

cess. Capasso (2005: 4) identified this specimen as having a

trauma (Fig. 25.20). The distal articular surface appears nor-

possible osteoma but admitted detailed examination was re-

mal, but a healed simple fracture with little or no evidence of

quired to confirm this diagnosis.

infection is located 150 mm above the distal end. The bone proximal to this break is normal in all regards, with a shaft

FORELIMB

width of 46 mm and a thickness of 34.5 mm. The proximal end of the fragment, however, is expanded into a massive

Moodie (1923) referred a badly fractured and poorly healed

oval-shaped club-like mass of callus. This feature is about 155

humerus from Alberta, AMNH 5207 (cast TMP 98.11.1), to

mm long, 92 mm wide and up to 61 mm thick. The callus had

Ceratopsidae, but subsequent examination shows that it be-

evidently been eroded prior to burial, so the original propor-

longs to a hadrosaur.

tions of this feature were likely somewhat larger during life.

TMP 82.18.227 is an adult Centrosaurus left ulna bearing a

Gross examination of the callus discloses a mostly smooth

prominent bi-lobed and rounded knob on its external edge,

surface and only a few small, randomly placed foramina. The

just below the proximal articulating surfaces. The knob, mea-

compact bone that covers the callus is paper-thin in several

suring 21 mm long, up to 13.5 mm wide and having a maxi-

places, and erosion and exposure have removed most of it,

mum height of 8 mm, may represent an old avulsion injury as

exposing the cancellous trabeculae. The proximal half of the

compared to diagnosed avulsion injuries observed in other

fibula was not found, probably because the two halves healed

dinosaurs.

into separate units forming a pseudoarthrosis.

McGowan (1983) identified a pathological condition on the

TMP 95.400.215 is an eroded Centrosaurus fibula shaft en-

manii of Chasmosaurus ROM 843. Here, the metacarpals ex-

cased in a massive callus. Owing to modern erosion and ex-

hibit irregular rough-surfaced outgrowths. A cast of this speci-

posure, the actual wound is now missing. The callus (where

men on display in the TMP disclosed pathological growths on

measurable) is up to 16 mm thick and was not firmly attached

right metacarpals I–III and left metacarpal I. Another cast

to the parent bone as it has broken and flaked away in many

of Chasmosaurus CMN 2280 on display at TMP also exhib-

places, exposing the bone shaft.

ited bony outgrowths on right metacarpals III–IV and left

The absence of osteopathy affecting heavy weight-bearing

metacarpal I. The left manal phalanx I-1 is similarly affected.

limb elements in these large animals is not entirely surprising

These specimens and their significance have most recently

and is probably related to the unlikely long term survivability

been examined by Rega and Holmes (2006) and Rega et al.

of animals with such serious traumas. The zoological litera-

(this volume).

ture is replete with references of small to medium-sized ex-

The DPP pachyrhinosaur specimen (TMP 2002.76.1; Ryan

tant animals fracturing major limb bones and surviving (Fig.

et al. this volume) also exhibits severe manual pathology in

25.21; e.g., Ballantine 1912; Gander 1930; Spinage 1971;

the form of stress fractures and especially premortem destruc-

McDiarmid 1975; McConnell et al. 1974; Wobeser 1992).

372 tanke & rothschild

FIGURE 25.21. Femora (posterior view) of modern Canis latrans (TMP 2007.30.29) showing advanced healing of a complete fracture of the right femur (etiology unknown) resulting in massive callus development, angulation, and overall shortening by 16%. Scale bar is 5 cm.

These animals often lead relatively normal lives despite their crippling injuries. However, although large, terrestrial quadrupedal animals can survive injuries to the limb extremities and other less critical bones (i.e., ribs), bad fractures of the major weight-bearing limb bones are usually fatal (Bulstrode et al. 1986; Rothschild and Martin 2006). Rothschild (1988) describes a stress fracture in an isolated, Distal half of Centrosaurus fibula (TMP 89.18.108) in external view with massive club-like pseudoarthrosis at proximal end. Scale bar is 10 cm. FIGURE 25.20.

adult ceratopsian phalanx from DPP. Stress-fractured phalanges are known in Centrosaurus (AMNH 5351, AMNH 5427), Styracosaurus (AMNH 5372; BMR pers. obs.), Pipestone Creek Pachyrhinosaurus, and other isolated specimens from DPP. In over 130 mostly adult Pachyrhinosaurus phalanges, only one example of a mild stress fracture was found (TMP 85.112.70). TMP 67.19.85 represents a severe form of this pathology (Fig. 25.22A). Interestingly, the stress fractures seem to affect mostly the proximal phalanges of the pes, in particular the main weight-bearing digits II–IV. Whereas all current specimens are from adult-sized animals a larger sample size is re-

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 373

FIGURE 25.22. Osteopathic centrosaurine phalanges (undetermined positions). (A) Lateral view of TMP 67.19.85 with stress fracture and prominent growth indicated by arrow; (B) distal view of TMP 91.36.98 with deep pit affecting articulating surface (arrow). Scale bars are 1 cm.

quired to test for taphonomic bias. Rothschild (1988) has sug-

cular openings would not be expected to open directly into a

gested stress-fractured phalangeal osteopathy may have been

cartilage-covered articulating surface. The appearance of the

the result of sudden acceleration, perhaps in response to pred-

lesions are unlike the shallow and elongate pits with a finely

ators. Stresses incurred during long annual migrations or ad-

textured floor (the ‘‘divots’’ of Rothschild) of osteochondrosis

vanced age should also be considered.

dessicans as is frequently seen in hadrosaur pedal phalanges

Johnson and Storer (1974: fig. 122) figured an obviously

from Alberta and Montana (Rothschild 1997: 430; Rothschild

diseased ceratopsian phalanx from Alberta. Extra irregular

and Tanke 2007). Further work including detailed internal ex-

bone growths are present on the proximal rim of the articulat-

amination will be required to resolve these curious specimens.

ing end of the element, and some proximal endplate destruc-

Finally, manual phalanx pathology resulting in severe digit

tion is evident in the form of several deep pits. Unfortunately

I angulation in Chasmosaurus (CMN 2280) has been compared

the location of this interesting specimen is unknown.

to hallux valgus in humans (Rega and Holmes 2006, Rega et al.

A number of isolated phalanges in the TMP collections

2007, this volume).

(TMP 92.36.1039; TMP 98.93.74; TMP 2007.26.11) have severely pathologic or malformed distal ends so as to suggest a serious disease process; one was apparently so serious that the

Discussion

distal portion of the affected digit was probably sloughed off

INTRASPECIFIC AGONISTIC BEHAVIOR IN

and lost in life. Three of the preserved feet of the DPP pachy-

CERATOPSIDS: PALEOPATHOLOGICAL

rhinosaur (TMP 2002.76.1; Ryan et al. this volume) show se-

EVIDENCE FROM BONEBEDS

verely pathologic conditions of the phalanges also suggestive of premortem digit loss. The etiology of this condition is cur-

Many authors have attempted to correlate osteopathy in ex-

rently unknown.

tinct or sub-fossil aquatic and terrestrial vertebrates with be-

Other unexplained etiologies affecting the distal articulat-

havior, particularly intraspecific fighting (Case 1915; Moodie

ing surfaces of phalanges include shallow to deep, circular

1923; Gilmore 1946; Macdonald 1951; Courville 1953, 1967;

to oval-shaped pits in Centrosaurus specimen, TMP 91.18.30,

Semken et al. 1964; Wells 1964; Brothwell 1965; Kerley and

and isolated specimens TMP 66.32.11, TMP 91.36.26, TMP

Bass 1967; Nelson and Madsen 1978; Buffetaut 1983; Evans

91.36.98, and TMP 92.36.226 (Fig. 25.22B). It is curious that

1983; Monastersky 1989; Boucot 1990: 425; Ruben 1990;

only the distal ends are affected, and presently we can provide

Larsen 1997; Marshall et al. 1998; Tanke and Currie 1998; Mc-

no explanation for this phenomenon. The deep pits are mor-

Whinney et al. 2001; Rothschild et al. 2001; Zollikofer et al.

phologically similar to presumed vascular openings in cen-

2002; McCall et al. 2003; Avilla et al. 2004; Katsura 2004;

trosaurine cranial, vertebral, and limb bone elements, but vas-

Lingham-Soliar 2004; Schulp et al. 2004; Carpenter et al.

374 tanke & rothschild

2005; Everhart 2005; Rothschild and Molnar 2005; Gerholdt

don’t provide much support for an hypothesis of frequent,

and Godfrey 2006; McCammon 2006; Roach and Brinkman

intraspecific aggressive behaviors in these animals. Super-

2007; Longrich and Tanke in prep.).

numerary holes in ceratopsian skulls, often ascribed to goring

Lull (1933); Russell (1935); Davitashvili (1961); Farlow

injuries from conspecifics, have been recently addressed as

and Dodson (1975); Hopson (1977); Molnar (1977); Spassov

non-traumatic in origin (Tanke and Farke 2007) reducing such

(1979); Ostrom (1986); Alexander (1989); Dodson and Currie

‘‘evidence’’ even further. Ontogenetically mature centrosau-

(1990); Sampson (1997, 2001); and Farke (2004) have all ad-

rines have a propensity to resorb parts of their skulls, espe-

dressed the topic of intraspecific fighting or sparring in the

cially the orbital horncore tips or orbital bosses (Sampson

larger horned dinosaurs. Most of the earlier authors suggest

et al. 1997; Tanke and Farke 2003, 2007; Currie et al. 2008).

that aggressive intraspecific behavior in ceratopsians causing

Other affected regions include nasal horncore tips and parts of

injuries was a regular occurrence. Some of the more recent

the frill (Tanke and Farke 2007), although the timing and bio-

papers and the current general consensus in the paleontologi-

logical processes involved are not entirely understood. How

cal community follow Farlow and Dodson (1975) in conclud-

much further resorption would go with continuing ontogeny

ing that intraspecific centrosaurine behavior is common,

is unknown, but it could potentially result in large modifica-

probably highly aggressive, and the cause of frequent sub-

tions and bone loss in an individual, which could mimic mas-

lethal injuries, especially to the head. Furthermore, Farlow

sive pathology as in the case of the pachyrhinosaur nasal boss,

and Dodson (1975) have suggested that, due to the large me-

TMP 89.55.899 (Fig. 25.2).

dian nasal horn, centrosaurines were possibly more solitary in

Severe fractures in Centrosaurus and Pachyrhinosaurus pari-

habit. They argued that, because of the difficulty of blocking

etals could have been incurred during intraspecific fight-

nasal horn strikes from an adversary, injuries in these animals

ing between adults. Some, however, could have occurred and

would be more frequent than in their chasmosaurine cousins.

healed much earlier in life, as exemplified by the subadult

The excavation of centrosaurine bonebeds (and ultimately

Pachyrhinosaurus parietal, TMP 89.55.125, which shows pseu-

other dinosaur bonebeds), versus the more traditional empha-

doarthrosis across the posterior parietal bar. In this instance,

sis on isolated articulated or associated individuals, provides a

one would expect that one or both of the lateral parietal rami

unique opportunity to study the types and relative frequen-

would stabilize the fracture site and allow for proper repair,

cies of pathological conditions present within a single popu-

but instead, a pseudoarthrosis was formed. This is surprising

lation or group of dinosaurs. Thus, monodominant or low

given the way the frill is structured, and suggests that one or

diversity bonebeds (especially those in mudstone deposits,

both of the lateral rami were also fractured and never healed

which may have suffered less transport damage and winnow-

properly. Possibly part or all of the back of the frill was only

ing) and the osteopathies contained therein provide an excel-

loosely attached.

lent basis for paleopathological interpretation and indicate

How a subadult individual could break its frill is unknown.

the ‘‘health’’ and well being of a single population. The speci-

Although the frill would have been thin and vulnerable to

mens from the bonebeds can be used as a test of hypotheses re-

damage (Dodson 1984, 1986; Dodson and Currie 1988; Samp-

garding ceratopsian behavior, particularly intraspecific fight-

son et al. 1997), it is unlikely that immature individuals

ing resulting in fractures. Rothschild and Martin (2006: 123)

would be engaging in intraspecific territorial or dominance

were correct in stating that bone fractures can so thor-

behaviors that could result in serious damage. However, Byers

oughlyheal as to be undetectable to the paleopathologist.

(1987) noted play behavior and resultant injuries in juvenile

However, in Albertan dinosaur material in the TMP pale-

ibex (Capra sp.) and briefly speculated on play behavior in

opathology collections, virtually all of the fracture callus spec-

juvenile dinosaurs. Alternatively, as adults these injuries may

imens were undergoing active repair at the time of death and

have been sustained by any number of accidental means (e.g.,

therefore are easily observed. Only a few ‘‘completely healed’’

hooking their frills on trees or branches). Improper healing

fractures were present, but these were readily identifiable due

due to continued movement of the affected individual may

to permanent angulation of the distal portion of the element

have ensued and, possibly, would have been further enhanced

or by prominent pseudoarthroses. It is curious that most fract-

when the individual reached adult size and rapidly developed

ure callus specimens in TMP were in an active or well ad-

frill ornamentations (Sampson et al. 1997; Goodwin et al.

vanced condition at the time of death of the animal. Perhaps

2006; Tanke and Farke 2007). In this scenario, some frill pa-

the fracture(s) and related soft tissue trauma coupled with in-

thologies seen in adults may simply reflect injuries incurred

fection or predation contributed to the demise of these indi-

much earlier in life. Osteopathy deep inside the body, including fusions of neck

viduals. Paleopathologies in centrosaurines, while regularly seen,

and dorsal vertebrae, bone growths, and simple zygapophy-

occur in low numbers in large bonebed samples and, thus,

seal malformations, are also unlikely to be related to aggres-

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 375

sive intraspecific behavior, but may instead represent simple

discarded if none were found. Thus, only 29 paleopathologies

day-to-day wear and tear and age-related phenomena. When

in 12,000+ uncovered and examined specimens (0.24%) can

osteopathies similar to these are discounted from the aggres-

be estimated. In the Centrosaurus bonebed (BB 43), a conserva-

sive behavior debate, there is only a small sample of injuries

tive estimate of 15,000+ scrap specimens were uncovered dur-

that could be the result of intraspecific competition.

ing over a decade of excavations. Addition of the catalogued

In over 775 specimens collected from the Centrosaurus bone-

material results in only nine confirmed paleopathologies in

bed (BB 43; Quarry 143) at DPP, only 9 paleopathologies of all

nearly 16,000 specimens (0.056%), only four of which appear

types were located (Table 25.2), or 1.16%. Of these, 4 (0.52%)

related to trauma (0.025%).

appear directly related to trauma.

One could argue that taphonomic biases selectively remove

Similarly, the Styracosaurus bonebed from DPP (BB 42)

pathologies from the fossil record, but this is unlikely. The

yielded several hundred elements, among which only two pa-

thickened and dense callus of bone that often accompanies

thologies were observed: (1) an indeterminate waterworn limb

wound sites was evidently resistant to predepositional erosion

bone shaft (TMP 94.12.853) covered with finely textured

and therefore more likely to be buried and fossilized. Once the

newly deposited bone suggestive of a fracture callus; and (2) a

fossil is exposed at the surface, the pathological region of the

neural arch fragment with deep pitting of the right postzyg-

bone is still resistant to destruction. Extensively worn and

apophysis (TMP 90.57.16). Only the limb bone shaft osteopa-

weathered specimens have been found in southern Alberta

thy may have traumatic origins. DPP Centrosaurus bonebeds

that consist of the pathology only, the rest of the bone hav-

30, 41a, 91, 138, and 168 all yielded numerous bones, yet in

ing eroded either during the Cretaceous or after recent ex-

these sites, osteopathy occurrences are low. The occurrences

posure. Furthermore, the unusual appearance of a pathology

were mostly represented by healing dorsal rib fractures from

will often attract the attention of fieldworkers, resulting in its

the posterior series, which often exhibit pseudoarthroses a

collection. With all of these points favoring the preservation

short distance below the rib head. Many specimens from these

and collection of paleopathologies, they should be well rep-

sites were unavailable for inclusion in this study.

resented in museum collections. Thus, in our view, their low

From the bone-dense Pipestone Creek Pachyrhinosaurus

frequency of occurrence in monodominant bonebeds and

bonebed (Currie et al. 2008; Ralrick and Tanke 2008), four

museum collections correctly reflects their low frequency in

summers of work produced over 2,300 catalogued specimens.

nature (Table 25.4). Accordingly, osteopathy in Albertan cen-

Even in this field locality, where the senior author kept a close

trosaurine bonebeds are simply not present in sufficient num-

watch for any examples of osteopathy, such specimens were

bers to support the hypothesis that these animals regularly

rare, with only 34 examples documented (1.48%; Table 25.3).

engaged in aggressive intraspecific behavior that resulted in

Five of these may represent non-pathological bone resorption

bone injuries.

or other conditions. Of the remaining sample, perhaps 29

With ‘‘horn thrust injuries’’ in chasmosaurines being rein-

(1.26%) could be interpreted as evidence of trauma, possibly

terpreted as likely non-pathologic in nature (Tanke and Farke

incurred from intraspecific strife. The rest of the collection

2007), the ‘‘evidence’’ for fighting behavior in this subfamily is

contains some elements that reflects developmental abnor-

also minimal, although bonebed evidence is lacking. As for

malities, secondary infections, individual variation, or other

the possibility of more injuries occurring in centrosaurines

conditions related to advanced age.

than in chasmosaurines, this hypothesis remains untested

Moving farther afield, no definite pathologies are reported

due to a lack of good comparative chasmosaurine material.

from two Einiosaurus bonebeds in Montana that have pro-

However, if the extra fontanelles in chasmosaurines are rein-

duced more than 750 bones (S. Sampson pers. com.), nor were

terpreted as a non-pathological feature (Tanke and Farke

any found in the excavation of a Centrosaurus bonebed in

2007), osteopathy is likely rare in this subfamily as well.

west-central Saskatchewan that produced 50 specimens (T. Tokaryk pers. com.). In some ways these numbers are deceiving. It must be remembered that while a strong bias in favor of the collection of

POSSIBLE BEHAVIORAL SIGNIFICANCE OF RIB PALEOPATHOLOGIES IN ALBERTAN CERATOPSIDS

pathological specimens exists, there is also strong bias against

Paleopathological evidence does not lend much support to

the collection of fragmentary bones in these bonebeds. For

the idea that centrosaurines and possibly chasmosaurines en-

every specimen collected from a bonebed, many more incom-

gaged in frequent intraspecific fighting as previously thought.

plete bones and fragments are uncovered, examined, and dis-

Because the focus of osteopathy in Albertan ceratopsids ap-

carded. For example, in the Pachyrhinosaurus bonebed, an es-

pears to be along the dorsal rib series, this pattern may provide

timated 10,000+ scrap specimens have been examined for

some insight into the nature of ceratopsid social behaviors.

pathologies and other features (toothmarks, etc.) and then

The fact that at least some centrosaurines and chasmosaurines

376 tanke & rothschild

Table 25.1. Horned Dinosaur Bonebeds in Alberta That Have Been Systematically Excavated by the Royal Tyrrell Museum Since 1979 (Tanke 2007; Tanke this volume). Genus/species

TMP bonebed

Field locality

Centrosaurus apertus

30

Dinosaur Provincial Park, AB

Centrosaurus apertus

41a

Dinosaur Provincial Park, AB

Centrosaurus apertus

43

Dinosaur Provincial Park, AB

Centrosaurus apertus

91

Dinosaur Provincial Park, AB

Centrosaurus apertus

168

Dinosaur Provincial Park, AB

Centrosaurus cf. apertus

Hilda

Hilda, AB

C. brinkmani

138

Dinosaur Provincial Park, AB

Pachyrhinosaurus n. sp.

Pipestone Creek

Grande Prairie, AB

Styracosaurus sp.

42

Dinosaur Provincial Park, AB

traveled in large groups or herds for part (if not all) of the time is now well established from bonebed evidence (Eberth 1996, 1998; Sampson 2001; Eberth and Getty 2005; Tanke 2007; Eberth et al. this volume). Animals living in groups or herds and interacting with others of their own kind have a higher

Table 25.2. List of Centrosaurus Specimens Demonstrating Osteopathy Recovered from Bonebed 43 (quarry 143), Dinosaur Provincial Park, Alberta. Specimen no.

Osteopathy type

potential for injuries inflicted by conspecifics than animals

TMP 79.11.9

Diseased indeterminate skull bone.

leading a more solitary existence. As adults, all ceratopsids

TMP 82.18.227

Ulna with rounded exostosis on shaft.

were large and powerful animals with a big head bristling with

TMP 87.18.27

Distal caudal vertebra; ?congential defect.

dangerous horns and spikes of various sizes. Direct head-to-

TMP 89.18.108

Fibula, fracture with pseudoarthrosis

TMP 90.18.1

Rib with thin callus.

TMP 91.18.18

Left squamosal with ‘‘punched-out’’ lesion (Tanke

TMP 91.18.30

Phalanx with pitting on distal articulating end.

that could result in bone fractures. Rather, acts of bluff such as

TMP 91.18.31

Rib fragment with fracture callus.

uncompleted charges, vocalizations, and display by raising

TMP 91.18.77

Posterior dorsal rib with massive pseudoarthrosis

head combat may have been avoided or minimized to avoid serious injuries. It would seem more likely that during dominance or territorial disputes, centrosaurines (and probably chasmosaurines) would not engage in aggressive behavior

the highly decorated frill to a near vertical position to appear

(Fig. 25.20).

and Farke 2007: 327).

(Fig. 25.15C).

larger (sensu Farlow and Dodson 1975) seem more likely behaviors, settling their differences by highly ritualized head displays or mutual non-fatal head to flank butting (see Sampson 2001: fig. 19.1).

FUTURE DIRECTIONS

Flank butting in Pachyrhinosaurus and other centrosaurines

A simple histological approach, examining the margins of the

is suggested from the preponderance of broken and healing

extra holes that appear in the skull, especially the ones that

posterior ribs. In the extant American buffalo (Bison bison),

occur in the squamosals could confirm their pathological or

a form known to engage in flank or side-butting during ago-

non-pathological origins. Study of markedly asymmetrical

nistic behavior (Fuller 1960), broken ribs in adult bulls are

posterior parietals in some centrosaurines would be of inter-

common (McHugh 1958). In one scenario, antagonists may

est, particularly if tension from the pulling of soft tissues is

have stood end-to-end and head-butted one another’s flanks

involved. Many of the specimens described herein, especially

with the nasal boss or upper surfaces of the frill spikes. Bit-

those from DPP, are exquisitely preserved both internally and

ing with their sharp parrot-like beaks may also have been

externally, and no doubt more could be learned from MRI or

employed. An alternative flank butting scenario would see

CT analysis of these specimens. Straight et al. (2005, in press)

two animals assuming a ‘‘T’’ configuration, with the unhin-

successfully undertook isotopic analyses on a DPP hadrosaur

dered animal actively butting or ramming into the others side.

proximal caudal neural spine bearing two healing fractures.

Flank-butting behavior may have also been adopted by chas-

Analysis of bone powder samples from fracture callus and

mosaurines, but the presence of a wide flaring frill in this

non-injured sites along the same bone demonstrated an ele-

group combined with the barrel-shaped abdomen present in

vated body temperature at the wound sites. Such isotopic

all large ceratopsians might make end to end body positioning

work might be expanded to include contemporaneous cera-

in this subfamily difficult.

topsian material as well as ceratopsian taxa separated spatially

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 377

Table 25.3. Osteopathic or Unusual Pachyrhinosaurus n. sp. Specimens from the Pipestone Creek Bonebed (Wapiti Formation; Late Campanian) near Grande Prairie, Alberta (specimens are adult unless noted; does not include resorption pitting of postorbital bosses. J: juvenile; S: subadult; *: non-pathologic?) TMP no.

Osteopathy type

TMP 85.112.39

Left posterior dorsal rib with pseudoarthrosis.

TMP 85.112.52

Mid-dorsal rib shaft with fracture callus.

TMP 85.112.70

Phalanx with mild stress fracture.

TMP 85.112.86

Small rib fragment of indeterminate position (posterior?) with pseudoarthrosis.

TMP 86.55.111

Anomalous smooth lobe of bone of undetermined etiology arising from underside of left squamosal; avulsion?

TMP 86.55.304*

Jugal/epijugal fragment with 20 mm deep circular hole in epijugal = extreme non-pathologic bone resorption (Fig.

TMP 87.55.90

Right posterior dorsal rib with 2 healing fractures; distal-most forms a pseudoarthrosis.

TMP 87.55.101

Left quadrate. Distal articulating end concave instead of convex and medial side bearing a bluntly pointed bone spur;

TMP 87.55.102

Four fused distal caudal vertebrae (Fig. 25.16A, B).

25.5)?

above medial condyle (anterior side) a deep sub-circular pit (20—14—8 mm).

TMP 87.55.190

Left scapula with two low raised oval shaped lesions externally, mid-length on scapular spine: 32—22 mm; 66—43 mm.

TMP 87.55.210

Fracture resulting in severe asymmetry of posterior parietal; premortem loss of left frill spike (P3; Fig. 25.9a).

TMP 88.55.52

Cervical vertebra with malformed prezygapophysis; ‘‘moth-eaten’’ appearance on posterior endplate; bone spur on

TMP 88.55.90

Parietal bar (S) with bioerosion on dorsal surface (= osteomyelitis?).

TMP 88.55.191

Right posterior dorsal rib with healing fracture (Fig. 25.15A).

TMP 89.55.63

Dorsal rib ( J) with healing fracture of neck.

TMP 89.55.125

Posterior parietal (S) with transverse fracture of parietal bar forming a pseudoarthrosis.

TMP 89.55.205

Left posterior dorsal rib with pseudoarthrosis.

TMP 89.55.188*

Semicircular notch missing from left ventral edge of rostrum; non-pathological? (Fig. 25.1 right).

TMP 89.55.269

Left postorbital (S) with deep infolding of cortical bone (58 mm long, up to 23 mm wide, up to 13 mm deep) posterior

TMP 89.55.287

Mid-caudal vertebra with neural spine fracture and localized bone malformation or erosion.

TMP 89.55.363

Proximal caudal vertebra with neural spine fracture.

TMP 89.55.389

Left posterior dorsal rib with healing fracture (Fig. 25.15B).

TMP 89.55.464

Left dorsal rib (S) with swelling on ventral neck.

TMP 89.55.719

Rib (dorsal?) fragment with fracture callus.

TMP 89.55.883

Parietal bar with small rounded exostosis on proximal end near dorsal midline.

TMP 89.55.899*

Portion of ?right nasal boss with anomalous shape and extensive bone remodelling/resorption ventrally and internally.

centrum (Fig. 25.11A, B).

and medial to orbital horncore.

Abnormal shape makes orientation and interpretation difficult (non-pathological?) (Fig. 25.2B, C). TMP 89.55.978

Unfused adult-sized cervical neural arch with the right prezygapophysis outline misshapen and slightly swollen with

TMP 89.55.1072

Right quadrate with deep transverse crack between distal articulating condyles; medial condyle with circular depressed

several concentric folds or ridges externally. pit on articular surface. TMP 89.55.1085

Asymmetrical posterior parietal with dorsal (vs. lateral) curvature of right P3 frill spike and asymmetry (Fig. 25.9B).

TMP 89.55.1091*

Jugal/epijugal with small depressed pit centered on epijugal exposing ?trabecular bone due to erosion or non-pathologic event?

TMP 89.55.1234

Skull with large hole ventro-anterior to right orbit; low circular growth on left maxillary ramus; ventral edge of left squamosal resorbed; ‘‘punched-out’’ lesions on internal left squamosal; See Rothschild and Tanke (1997); Tanke and Farke (2003, 2007) for descriptions and figures.

TMP 89.55.1300

Left posterior dorsal rib ( J) with small bone ‘‘lump’’ on ventral neck.

TMP 89.55.1503*

Posterior parietal with right P3 horn larger, different shape and angulation than opposite one (individual variation; Fig. 25.9C).

TMP 89.55.1541

Small rounded exostosis on dorsal proximal parietal bar.

378 tanke & rothschild

Table 25.4. Categorized Osteopathies and Anomalies from 34 Specimens Recovered from Pipestone Creek Pachyrhinosaurus Bonebed. % Represented Category

Occurrences

(=[Occurrences/34]—100)

Juvenile (all regions/types)

2

5.9

Subadult (all regions/types)

3

8.8

Adult (all regions/types)

29

85.3

Nonpathologic? (*)

5

14.7

Skull (all regions/types)

16

47

Rostrum (*)

1

2.9

Nasal (*)

1

2.9

Maxilla (TMP89.55.1234)

1

2.9

Postorbital

1

2.9

Jugal/epijugal

2 (2*)

5.9 (0)

Quadrate

2

5.9

Squamosal (incl. TMP

2

5.9

Parietal

6

17.6

Parietal subadult

2

5.9

Parietal adult

5 (1*)

14.7 (11.8*)

Side of face (TMP 89.55.1234)

1

2.9

89.55.1234)

Postcrania (all)

18

53

Cervical vertebrae

2

5.9

Caudal vertebrae**

3

8.8

Ribs (all; N = 11)

11

32.4

Ribs: mid-dorsal

3

8.8

Ribs: posterior dorsal

6

17.6

Ribs: position undetermined

2

5.9

Scapula

1

2.9

Phalanx

1

2.9

Notes: TMP 89.55.1234 exhibits multiple osteopathies, but each pathology is considered separately here. * Non-pathologic? ** Four fused vertebrae counted as one specimen.

and stratigraphically. It would be desirable to examine osteop-

son, Tracy Ford, for donation of the specimens in Fig. 25.21.

athy frequency and body locations in chasmosaurine bone-

The authors thank Andy Farke, Brenda-Chinnery-Allgeier,

beds to see if they correlate with the centrosaurine bonebed

and David Eberth for reviews and editorial assistance. Most of

data presented herein. Firm quantification of all specimens

the specimen photographs are by Sue Sabrowski (TMP), to

(collected or discarded) from future ceratopsian bonebed ex-

whom the authors are grateful.

cavations would better elucidate epidemiological aspects of References Cited

osteopathy in the group. Acknowledgments

This paper is the result of a long-standing interest in dinosaurian paleopathology and had a lengthy gestation period extending back to 1989. We thank the following individuals for their data and assistance: Kevin Aulenback, Clive Coy, Philip J. Currie, Jim Gardner, Tetsuto Miyashita, Patty E. Ralrick, Michael J. Ryan, Scott Sampson, Brandon Strilisky, Tim Tokaryk, and Jackie Wilke. Valuable editorial comments on a much earlier version of this paper were provided by Julie L. Cormack. The authors thank the late Kenneth K. Ford and his

Alexander, R. 1989. Dynamics of Dinosaurs and Other Extinct Giants. New York: Columbia University Press. Anonymous. 1934. Glands may have caused evolution of freak dinosaurs. Science News Letter 25: 182. Avilla, L. S., R. Fernandes, and D. F. B. Ramos. 2004. Bite marks on a crocodylomorph from the upper Cretaceous of Brazil: Evidence for social behavior? Journal of Vertebrate Paleontology 24: 971–973. Baker, J., and D. Brothwell. 1980. Animal Diseases in Archaeology. London: Academic Press. Ballantine, W. J. H. 1912. Recovery of animals from injuries. Journal of the Bombay Society of Natural History 21: 1069.

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 379

Barghusen, H. R. 1975. A review of fighting adaptations in dinocephalians (Reptilia, Therapsida). Paleobiology 1: 295–311. Boucot, A. J. 1990. Sexual behavior. In A. J. Boucot. ed., Evolutionary Paleobiology of Behavior and Coevolution, pp. 381–429. Amsterdam: Elsevier Science Publishers. Brothwell, D. R. 1965. Digging Up Bones: The Excavation, Treatment and Study of Human Skeletal Remains. London: Trustees of the British Museum (Natural History). Brown, B., and E. M. Schlaikjer. 1937. The skeleton of Styracosaurus with the description of a new species. American Museum Novitates 955: 1–12. ———. 1940. A new element in the ceratopsian jaw with additional notes on the mandible. American Museum Novitates 1092: 1–13. Buffetaut, E. 1983. Wounds on the jaw of an Eocene mesosuchian crocodilian as possible evidence for the antiquity of crocodilian intraspecific fighting behavior. Paläontologische Zeitschrift 57: 143–145. Bulstrode, C., J. King, and B. Roper. 1986. What happens to wild animals with broken bones? Lancet 1: 29–31. Byers, J. A. 1987. Where the deer and antelope play. Natural History 96: 54–60. Campione, N. E. 2005. Features and homologies of the ceratopsid syncervical: A comparative analysis with basal tetrapods. In D. R. Braman, F. Therrien, E. B. Koppelhus, and W. Taylor, eds., Dinosaur Park Symposium: Short Papers, Abstracts, and Program, pp. 12–14. Drumheller: Royal Tyrrell Museum of Palaeontology. Campione, N. E., D. C. Evans, and D. H. Tanke. 2008. The neoceratopsian syncervical: Variation and function. In H. Allen, ed., Alberta Palaeontological Society, Twelfth Annual Symposium, Abstracts, pp. 19–21. Calgary: Mount Royal College. Campione, N. E., and R. Holmes. 2006. The anatomy and homologies of the ceratopsid syncervical. Journal of Vertebrate Paleontology 26(4): 1014–1017. Capasso, L. L. 2005. Antiquity of cancer. International Journal of Cancer 113: 2–13. Carpenter, K. 1997. Agonistic behavior in pachycephalosaurs (Ornithischia: Dinosauria): A new look at head-butting behavior. Contributions to Geology, University of Wyoming 32: 19–25. Carpenter, K., F. Sanders, L. A. McWhinney, and L. Wood. 2005. Evidence for predator-prey relationships: Examples for Allosaurus and Stegosaurus. In K. Carpenter, ed., The Carnivorous Dinosaurs, pp. 325–350. Bloomington: Indiana University Press. Case, E. C. 1915. On a nearly complete skull of Symbos cavifrons Leidy from Michigan. Occasional Paper of the Museum of Zoology, University of Michigan 13: 1–3. Chure, D. J., A. R. Fiorillo, and A. Jacobsen. 1998. Prey bone utilization by predatory dinosaurs in the late Jurassic of North America, with comments on prey bone use by dinosaurs throughout the Mesozoic. Gaia15: 227–232. Colbert, E. H. 1981. Dinosaurs with horns. Terra 19: 23–29. Courville, C. B. 1953. Cranial injuries in prehistoric animals— with special notes on a healed wound of the skull in the Dire Wolf (Canis [Aenocyon] dirus [Leidy]) and a mortal wound in

380 tanke & rothschild

the California Black Bear (Ursus americanus). Bulletin of the Los Angeles Neurological Society 18: 117–126. ———. 1967. Cranial injuries in prehistoric man. In D. Brothwell and A. T. Sandison, eds., Diseases in Antiquity: A Survey of the Diseases, Injuries and Surgery of Early Populations, pp. 606–622. Springfield: Charles C. Thomas. Currey, J. 1984. The Mechanical Adaptations of Bones. Princeton: Princeton University Press. Currie, P. J., W. Langston Jr., and D. H. Tanke. 2008. A new species of Pachyrhinosaurus (Dinosauria, Ceratopsidae) from the Upper Cretaceous of Alberta. In P. J. Currie, W. Langston, Jr., and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. 1–108. Ottawa: NRC Research Press. Davitashvili, L. S. 1961. Teoriya polovogo otbora. [The theory of sexual selection]. Moscow: Izdatel’stvo Akademii Nauk [Academy of Science Press]. Dixon, D., B. Cox, R. J. G. Savage, and B. Gardiner. 1988. The Macmillan Illustrated Encyclopedia of Dinosaurs and Prehistoric Animals: A Visual Who’s Who of Prehistoric Life. New York: Macmillan. Dodson, P. 1984. Small Judithian ceratopsians, Montana and Alberta. In W.-E. Reif and F. Westphal, eds., Third Symposium on Mesozoic Terrestrial Ecosystems, Short Papers, pp. 73–78. Tubingen: Attempto Verlag. ———. 1986. Avaceratops lammersi: A new ceratopsid from the Judith River Formation of Montana. Proceedings of the Academy of Natural Sciences of Philadelphia 138: 305–317. ———. 1996. The Horned Dinosaur: A Natural History. Princeton: Princeton University Press. Dodson, P., and P. J. Currie. 1988. The smallest ceratopsid skull: Judith River Formation of Alberta. Canadian Journal of Earth Sciences 25: 926–930. ———. 1990. Neoceratopsia. In D. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 593–618. Berkeley: University of California Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D.B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A. 1996. Ceratopsian bonebeds in the Dinosaur Park Formation (Campanian) of southern Alberta: Bigger than we thought? Journal of Vertebrate Paleontology 16(3, Suppl.): 32A. ———. 1998. Clustered ceratopsian bonebeds, southern Alberta, Canada: Primary evidence for the size of ceratopsian-herd death assemblages. In D. L. Wolberg, K. Gittis, S. Miller, L. Carey, and A. Raynor, eds., The Dinofest Symposium, p. 13. Philadelphia: Academy of Natural Sciences. Eberth, D. A., D. B. Brinkman, and V. Barkas. 2010. A centrosaurine mega-bonebed from the Upper Cretaceous of southern Alberta: Implications for behavior and death events. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 495–508. Bloomington: Indiana University Press.

Eberth, D. A., and M. A. Getty. 2005. Ceratopsian bonebeds: Occurrence, origins, and significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 501–536. Bloomington: Indiana University Press. Edmund, A. G. 1960. Tooth replacement phenomena in the lower vertebrates. Royal Ontario Museum Life Sciences Contribution 52: 1–190. ———. 1969. Dentition. In C. Gans, ed., Biology of the Reptilia. Vol. 1: Morphology A, pp. 117–200. London: Academic Press. Erickson, G. M., and K. H. Olson. 1996. Bite marks attributable to Tyrannosaurus rex: Preliminary descriptions and implications. Journal of Vertebrate Paleontology 16: 175–178. Evans, S. E. 1983. Mandibular fracture and inferred behavior in a fossil reptile. Copeia 1983: 845–847. Everhart, M. 2005. Mosasaur pathology. Online at: www.oceansofkansas.com/mosapath.html Farke, A. A. 2004. Horn use in Triceratops (Dinosauria: Ceratopsidae): Testing behavioral hypotheses using scale models. Palaeontologia Electronica 7: 1–10. Farke, A. A., and W. Alley. 2006. An abnormal squamosal of Triceratops (Ornithischia: Ceratopsidae) from the upper Cretaceous Hell Creek Formation of South Dakota. New Mexico Museum of Natural History and Science Bulletin 35: 371–372. Farke, A. A., and P. M. O’Connor. 2007. Pathology of Majungasaurus crenatissimus (Theropoda, Abelisauridae) from the Late Cretaceous of Madagascar. Journal of Vertebrate Paleontology 27: 180–184. Farlow, J. O., and P. Dodson. 1975. The behavioral significance of frill and horn morphology in ceratopsian dinosaurs. Evolution 29: 353–361. Fiorillo, A. R. 1988. Taphonomy of Hazard Homestead Quarry (Ogallala Group), Hitchcock County, Nebraska. Contributions to Geology, University of Wyoming 26: 57–97. Fuller, W. A. 1960. Behavior and social organization of the wild bison of Wood Buffalo National Park, Canada. Arctic 13: 3–19. Galton, P. M. 1970. Pachycephalosaurids: Dinosaurian battering rams. Discovery 6: 23–32. Gander, F. F. 1930. Fractured leg not fatal to rabbit. Journal of Mammalogy 11: 240. Gerholdt, J. M., and S. J. Godfrey. 2006. Hard knock life: Adversity and survival in the fossil record. Bugeye Times (Calvert Marine Museum) 31: 1–6, 7. Gilmore, C. W. 1917. Brachyceratops: A ceratopsian dinosaur from the Two Medicine Formation of Montana, with some notes on associated fossil reptiles. U.S. Geological Survey Professional Paper 103: 1–45. ———. 1946. Notes on recently mounted reptile fossil skeletons in the United States National Museum. Proceeding of the U.S. National Museum 96: 195–203. Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian. 2006. The smallest known Triceratops skull: New observations on ceratopsid cranial anatomy and ontogeny. Journal of Vertebrate Paleontology 26: 103–112.

Haas, G. 1955. The jaw musculature in Protoceratops and in other ceratopsians. American Museum Novitates 1729: 1–24. Halperin, E. C. 2004. Paleo-oncology: The role of ancient remains in the study of cancer. Perspectives on Biology and Medicine 47: 1–14. Happ, J. W. 2008. An analysis of predatory-prey behavior in a head-to-head encounter between Tyrannosaurus rex and Triceratops. In P. Larson and K. Carpenter, eds., Tyrannosaurus rex, the Tyrant King, pp. 354–368. Bloomington: Indiana University Press. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monographs 49: 1–300. Holmes, R., M. J. Ryan, and D. Lloyd. 2007. Restoration of the pathological parietal ornamentation of the holotype skull of Styracosaurus albertensis (CMN 344) with reference to new undistorted specimens. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 78–79. Drumheller: Royal Tyrrell Museum of Palaeontology. Hopson, J. A. 1977. Relative brain size and behavior in archosaurian reptiles. Annual Review of Ecology and Systematics 8: 429–448. Horner, J. R., and M. B. Goodwin. 2008. Ontogeny of cranial epiossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Jacobsen, A. R. 1998. Feeding behavior of carnivorous dinosaurs as determined by tooth marks on dinosaur bones. Historical Biology 13: 17–26. Johnson, H., and J. E. Storer 1974. A Guide to Alberta Vertebrate Fossils from the Age of Dinosaurs. Edmonton: Provincial Museum and Archives of Alberta, Publication Number 4. Katsura, Y. 2004. Paleopathology of Toyotamaphimeia machikanensis (Diapsida, Crocodylia) from the middle Pleistocene of central Japan. Historical Biology: A Journal of Paleobiology 16: 93–97. Kerley, E. R., and W. M. Bass. 1967. Paleopathology: Meeting ground for many disciplines. Science 157: 638–644. Krauss, D. A., A. Pezon, P. Nguyen, and I. Salame. 2007. Horn morphology determines frill morphology in chasmosaurine ceratopsians. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Program, pp. 96–104. Drumheller: Royal Tyrrell Museum of Palaeontology. Krauss, D. A., A. Pezon, P. Nguyen, I. Salame, and S. B. Rywkin. 2010. Evolutionary interactions between horn and frill morphology in chasmosaurine ceratopsians. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 282–292. Bloomington: Indiana University Press. Langston, W., Jr. 1967. The thick-headed ceratopsian dinosaur Pachyrhinosaurus (Reptilia: Ornithischia), from the Edmonton Formation near Drumheller, Canada. Canadian Journal of Earth Sciences 4: 171–186. ———. 1975. The ceratopsian dinosaurs and associated lower vertebrates from the St. Mary River Formation (Maastrichtian) at Scabby Butte, Southern Alberta. Canadian Journal of Earth Sciences 12: 1576–1608.

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 381

Larsen, C. S. 1997. Bioarchaeology: Interpreting Behavior from the Human Skeleton. Cambridge: Cambridge University Press. Lehman, T. M. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. Lingham-Soliar, T. 2004. Paleopathology and injury in the extinct mosasaur (Lepidosauromorpha, Squamata) and implications for modern reptiles. Lethaia 37: 255–262. Lull, R. S. 1908. The cranial musculature and the origin of the frill in the ceratopsian dinosaurs. American Journal of Science 25: 387–398. ———. 1933. A revision of the Ceratopsia or horned dinosaurs. Peabody Museum of Natural History Bulletin 3: 1–175. Macdonald, J. R. 1951. Pathological vertebrates from South Dakota. Bulletin of the Geological Society of America 62: 1539. Marshall, C., D. Brinkman, R. Lau, and K. Bowman. 1998. Fracture and osteomyelitis in PII of the second pedal digit of Deinonychus antirrhopus (Ostrom) an Early Cretaceous ‘‘raptor’’ dinosaur. Palaeontology Newsletter 39: 16. McCall, S., V. Naples, and L. Martin. 2003. Assessing behavior in extinct animals: Was Smilodon social? Brain, Behavior and Evolution 61: 159–164. McCammon, S. 2006. A mammoth death match preserved for the ages. On line at http://www.npr.org/templates/story/ story.php?storyId=5173078 McConnell, E. E., P. A. Basson, V. De Vos, B. J. Myers, and R. E. Kunz. 1974. A survey of diseases among 100 free-ranging baboons (Papio ursinus) from Kruger National Park. Onderstepoort Journal of Veterinary Research 41: 97–168. McDiarmid, A. 1975. Some disorders of wild deer in the United Kingdom. Veterinary Record 97: 6–9. McGowan, C. 1983. The Successful Dragons: A Natural History of Extinct Reptiles. Toronto: Samuel Stevens. McHugh, T. 1958. Social behavior of the American Bison (Bison bison bison). Zoologica 43: 1–40. McWhinney, L. A., B. M. Rothschild, and K. Carpenter. 2001. Post-traumatic chronic osteomyelitis in Stegosaurus tail spikes. In K. Carpenter and J. I. Kirkland, eds., The Armored Dinosaurs, pp. 141–156. Bloomington: Indiana University Press. Molnar, R. E. 1977. Analogies in the evolution of combat and display structures in ornithopods and ungulates. Evolutionary Theory 3: 165–190. Monastersky, R. 1989. A nose for combat. Science News 136: 318. ———. 1990. Reopening old wounds: Physicians and paleontologists learn new lessons from ancient ailments. Science News 137: 40–42. Moodie, R. L. 1923. Paleopathology: An Introduction to the Study of Ancient Evidences of Disease. Urbana: University of Illinois Press. ———. 1926. Studies in paleopathology XVI. Spondylitis deformans in a crocodile from the Pleistocene of Cuba. Annals of Medical History 8: 78–82. Nelson, M. E., and J. H. Madsen, Jr. 1978. Late Pleistocene musk oxen from Utah. Transactions of the Kansas Academy of Science 81: 277–295.

382 tanke & rothschild

Norman, D. 1985. The Illustrated Encyclopedia of Dinosaurs. London: Salamander Books. Ostrom, J. H. 1964. A functional analysis of jaw mechanics in the dinosaur Triceratops. Postilla 88: 1–35. ———. 1986. Social and unsocial behavior in dinosaurs. In M. H. Nitecki and J. A. Kitchell, eds., Evolution of Animal Behavior: Paleontological and Field Approaches, pp. 41–61. New York: Oxford University Press. Parks, W. A. 1921. The head and forelimb of a specimen of Centrosaurus apertus. Transactions of the Royal Society of Canada 15, Sect. IV: 53–64. Ralrick, P. E., and D. H. Tanke. 2008. Comments on the quarry map and preliminary taphonomic observations of the Pachyrhinosaurus (Dinosauria: Ceratopsidae) bone bed at Pipestone Creek, Alberta, Canada. In P. J. Currie, W. Langston Jr., and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. 109–116. Ottawa: NRC Research Press. Rega, E., and R. Holmes. 2006. Manual pathology indicative of locomotor behavior in two chasmosaurine ceratopsid dinosaurs. Journal of Vertebrate Paleontology 26(3, Suppl.): 114A. Rega, E., R. Holmes, and A. Tirabasso. 2007. Animation of locomotor behavior based on manual pathology in two chasmosaurine ceratopsid dinosaurs. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, p. 123. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2010. Habitual locomotor behavior inferred from manual pathology in two Late Cretaceous chasmosaurine ceratopsid dinosaurs, Chasmosaurus irvinensis (CMN 41357) and Chasmosaurus belli (ROM 843). In M. J. Ryan, B. J. ChinneryAllgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 340–354. Bloomington: Indiana University Press. Roach, B. T., and D. L. Brinkman. 2007. A reevaluation of cooperative pack hunting and gregariousness in Deinonychus antirrhopus and other nonavian theropod dinosaurs. Bulletin of the Peabody Museum of Natural History 48: 103–138. Rothschild, B. M. 1988. Stress fracture in a ceratopsian phalanx. Journal of Paleontology 62: 302–303. ———. 1997. Dinosaurian Paleopathology. In J. O. Farlow and M. K. Brett-Surman, eds., The Complete Dinosaur, pp. 426–448. Bloomington: Indiana University Press. Rothschild, B. M., and D. Berman. 1991. Fusion of caudal vertebrae in Late Jurassic sauropods. Journal of Vertebrate Paleontology 11: 29–36. Rothschild, B. M., and L. D. Martin. 1993. Paleopathology: Disease in the Fossil Record. London: CRC Press. ———. 2006. Skeletal impact of disease. New Mexic Museum of Natural History and Science Bulletin 33: 1–226. Rothschild, B. M., and R. E. Molnar. 2005. Sauropod stress fractures as clues to activity. In V. Tidwell and K. Carpenter, eds., Thunder-Lizards, pp. 381–392. Bloomington: Indiana University Press. Rothschild, B. M., and D. H. Tanke. 1992. Paleopathology: Insights to lifestyle and health in the geological record. Geoscience Canada 19: 73–82.

———. 1997. Thunder in the Cretaceous: Interspecies conflict as evidence for ceratopsian migration? In D. L. Wolberg, E. Stump, and G. D. Rosenberg, eds., Dinofest International, pp. 77–82. Philadelphia: Academy of Natural Sciences. ———. 2007. Osteochondrosis in Late Cretaceous hadrosaurs: A manifestation of ontologic failure. In K. Carpenter, ed., Horns and Beaks: Ceratopsians and Ornithopod Dinosaurs, pp. 171–183. Bloomington: Indiana University Press. Rothschild, B. M., D. H. Tanke, and T. Ford. 2001. Theropod stress fractures and avulsions as a clue to activity. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 331– 336. Bloomington: Indiana University Press. Rothschild, B. M., D. H. Tanke, M. Helbling, and L. D. Martin. 2003. Epidemiologic study of tumors in dinosaurs. Naturwissenschaften 90: 495–500. Rothschild, B. M., B. J. Witzke, and I. Hershkovitz. 1999. Metastic cancer in the Jurassic. The Lancet 354: 398. Ruben, J. A. 1990. Evidence of convergent behavioral patterns in male crocodilians and phytosaurs. In A. J. Boucot, ed., Evolutionary Paleobiology of Behavior and Coevolution, pp. 427–428. Amsterdam: Elsevier Science Publishers. Russell, L. S. 1935. Musculature and functions in the Ceratopsia. National Museum of Canada Bulletin 77: 39–48. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., D. A. Eberth, D. B. Brinkman, P. J. Currie, and D. H. Tanke. 2010. A new Pachyrhinosaurus-like ceratopsid from the Upper Dinosaur Park Formation (Late Campanian) of Southern Alberta, Canada. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 141–155. Bloomington: Indiana University Press. Ryan, M. J., H. Holmes, and A. P. Russell. 2007. A revision of the late Campanian centrosaurine ceratopsid genus Styracosaurus from the western interior of North America. Journal of Vertebrate Paleontology 27: 944–962. Sampson, S. D. 1997. Dinosaur Courtship and Combat. In J. O. Farlow and M. K. Brett-Surman, eds., The Complete Dinosaur, pp. 383–393. Bloomington: Indiana University Press. ———. 2001. Speculations on socioecology of ceratopsid dinosaurs (Ornithischia: Neoceratopsia). In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 263–276. Bloomington: Indiana University Press. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Schulp, A. S., G. H. I. M. Walenkamp, P. A. M. Hofman, B. M. Rothschild, and J. W. M. Jagt. 2004. Rib fracture in Prognathodon saturator (Mosasauridae: Late Cretaceous). Netherlands Journal of Geosciences 83: 251–254. Semken, H. A., B. B. Miller, and J. B. Stevens. 1964. Late Wiscon-

sin woodland musk oxen in association with pollen and invertebrates from Michigan. Journal of Paleontology 38: 823–835. Spassov, N. B. 1979. Sexual selection and the evolution of hornlike structures of ceratopsian dinosaurs. Paleontologiya, Stratigrafiya i Litologiya 11: 37–48. Spinage, C. A. 1971. Two records of pathological conditions in the impala (Aepyceros melampus). Journal of Zoology, London 164: 269–270. Sternberg, C. M. 1940. Ceratopsidae from Alberta. Journal of Paleontology 14: 468–480. ———. 1950. Pachyrhinosaurus canadensis: Representing a new family of the Ceratopsia from southern Alberta. National Museum of Canada Bulletin 118: 109–120. Straight, W. H., G. L. Davis, H. C. W. Skinner, A. Haims, B. L. McClennan, and D. H. Tanke. In press. Computed tomographic imaging of bone lesions in hadrosaurs: Paleohistological and stable-isotope analysis. Journal of Vertebrate Paleontology. Straight, W. H., J. D. Karr, H. N. Woodward, R. E. Barrick, D. H. Tanke, and G. S. Dwyer. 2005. Dinosaur fever: Isotopic evidence for locally elevated temperature surrounding healing injuries in hadrosaur bone. Journal of Vertebrate Paleontology 25(3, Suppl.): 119A. Tanke, D. H. 1988. Ontogeny and dimorphism in Pachyrhinosaurus (Reptilia, Ceratopsidae), Pipestone Creek, N.W. Alberta, Canada. Journal of Vertebrate Paleontology 8(3, Suppl.): 27A. Tanke, D. H. 1989a. Paleopathologies in Late Cretaceous hadrosaurs (Reptilia: Ornithischia) from Alberta, Canada. Journal of Vertebrate Paleontology 9(3, Suppl.): 41A. ———. 1989b. K/U Centrosaurine (Ornithischia: Ceratopsidae) paleopathologies and behavioral implications. Journal of Vertebrate Paleontology 9(3, Suppl.): 41A. ———. 2007. History and census of ceratopsian discoveries and work in Alberta, Canada. In D. R. Braman (complier), Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 147–148. (CD-ROM.) Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2010. Lost in plain sight: Rediscovery of William E. Cutler’s missing Eoceratops. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 541–550. Bloomington: Indiana University Press. Tanke, D. H., and P. J. Currie. 1998. Head-biting in theropods: Paleopathological evidence. Gaia 15: 167–184. Tanke, D. H., and A. A. Farke. 2003. Cranial abnormalities in horned dinosaurs: Disease and normal biological processes— not combat wounds. In H. Allen, ed., Alberta Palaeontological Society, Seventh Annual Symposium, Abstracts Volume, pp. 78–81. Calgary: Alberta Palaeontological Society. ———. 2007. Bone resorption, bone lesions and extra cranial fenestrae in ceratopsid dinosaurs: A preliminary assessment. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 319–347. Bloomington: Indiana University Press.

Paleopathologies in Albertan Ceratopsids and Their Behavioral Significance 383

Tanke, D. H., and B. M. Rothschild. 1997. Paleopathology. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, pp. 525– 530. London: Academic Press. ———. 2002. DINOSORES: An annotated bibliography of dinosaur paleopathology and related topics—1838–2001. New Mexico Museum of Natural History and Science Bulletin 20. ———. 2007. Examples of paleopathology among Albertan Ceratopsia. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 149–153. Drumheller: Royal Tyrrell Museum of Palaeontology. Thompson, S., and R. Holmes. 2007. Forelimb stance and step cycle in Chasmosaurus irvinensis (Dinosauria: Neoceratopsia). Palaeontologia Electronica 10: 1–17. Tsuihiji, T. and P. J. Makovicky. 2007. Homology of the neoceratopsian cervical bar elements. Journal of Paleontology 81: 1132–1138.

384 tanke & rothschild

Tyson, H. 1977. Functional craniology of the Ceratopsia (Reptilia: Ornithischia) with special reference to Eoceratops. M.Sc. thesis. University of Alberta, Edmonton. Wells, C. 1964. Bones, Bodies and Disease. London: Thames and Hudson. Wobeser, G. 1992. Traumatic, degenerative, and developmental lesions in wolves and coyotes from Saskatchewan. Journal of Wildlife Diseases 28: 268–275. Wu, X-C., D. B. Brinkman, D. A. Eberth, and D. R. Braman. 2007. A new ceratopsid dinosaur (Ornithischia) from the uppermost Horseshoe Canyon Formation (upper Maastrichtian), Alberta, Canada. Canadian Journal of Earth Sciences 44: 1243–1265. Zollikofer, C. P. E., M. S. Ponce de Leon, B. Vandersmeersch, and F. Lévêque. 2002. Evidence for interpersonal violence in the St. Césaire Neanderthal. Proceedings of the National Academy of Science for the United States of America 99: 6444–6448.

PART FOUR HORNED DINOSAURS IN TIME AND SPACE Paleobiogeography, Taphonomy, and Paleoecology

26 An Update on the Paleobiogeography of Ceratopsian Dinosaurs B R E N D A J . C H I N N E R Y- A L L G E I E R A N D J A M E S I . K I R K L A N D

the paleobiogeography of ceratopsian (‘‘horned’’)

Several dispersal events are required to reconcile the

dinosaurs has rarely been analyzed, and usually only in

distribution of, and relationships among, ceratopsian

the context of general dinosaur paleobiogeography. In

dinosaurs. One or two dispersal events brought cera-

light of new ceratopsian discoveries that have expanded

topsians from their origins in Asia to North America, one

both the temporal and physical ranges of higher-level

perhaps leading to the leptoceratopsid lineage and the

clades, we review previous work and present new infor-

other to the ceratopsid lineage. A dispersal event back to

mation about the possible dispersal events that must

Asia may explain the occurrence of Turanoceratops, and

have occurred in this group of dinosaurs.

another the occurrence of Udanoceratops. Dispersal vec-

The earliest basalmost ceratopsians, Chaoyangsaurus

tors are more difficult to discern because the animals

and Yinlong, are known from Middle–Upper Jurassic sedi-

could have traveled between (1) eastern Asia, Siberia, and

ments in Asia. Both Psittacosauridae and their sister

Alaska, and the west coast of North America; (2) the east

group, the ‘‘frilled’’ neoceratopsians (represented by

coast of North America and Europe; or (3) both. Geologic

Archaeoceratops, Auroraceratops, Liaoceratops), occur in the

evidence, plus the Aptian occurrence in North America,

Early Cretaceous (Barremian-Albian) of Asia; roughly

supports the hypothesis that ceratopsians must have ini-

contemporaneous, unnamed specimens are known from

tially dispersed to North America from Asia via Europe

the Arundel, Wayan, and Cloverly formations in North

prior to the first North American/Asian immigration

America and, questionably, Australia. Members of the

event across the Bering Land Bridge near the end of the

clade (Bagaceratops, Cerasinops, Graciliceratops, Lepto-

Albian, nearly 15 million years later.

ceratops, Magnirostris, Montanoceratops, Prenoceratops, and Protoceratops), as well as Yamaceratops (Eberth et al. 2009) and the poorly understood (but paleobiogeographically

Introduction

important) Udanoceratops, persisted into the Late Creta-

The term ‘‘paleobiogeography’’ means, in essence, the study

ceous in Asia and North America. Ceratopsids are known

of the distribution patterns of extinct organisms (whether or

exclusively from the Late Cretaceous, as is Zuniceratops,

not these organisms have extant relatives). On one level, pa-

the sister taxon of the clade. The similarly aged

leobiogeographic patterns of extinct organisms are useful in

Turanoceratops is the only Asian representative of the

studies of past climate changes, extinctions, and paleogeog-

clade and is regarded in this discussion as a ceratopsid.

raphy—reconstructing past continental positions and coast-

387

lines. For paleontologists, paleobiogeography is an important

always create gaps in the geologic record, generally affecting

part of the study of any extinct organism because it can re-

continent shapes, coastline positions, and the existence and

veal aspects of its evolutionary history not readily ascertained

positions of islands. So, although most geologists agree on

from the fossils themselves.

many aspects of paleogeography, many differences of opin-

The study of dinosaur paleobiogeography usually includes,

ion remain based on the particular bodies of evidence used

but is not limited to, geographic origins of clades, patterns

to render the reconstructions. While seemingly minor, these

of distribution, and the final distribution of clades at global

differences can, in fact, significantly alter interpretations of

and/or local scales. Ceratopsian dinosaurs have classically

movements of a fossil clade, as exemplified here with the cur-

been interpreted as exhibiting a relatively simple paleobio-

rent interpretation of ceratopsian dispersal patterns.

geographic pattern because they were distributed in only two

As an alternative to the hypothesis of dispersal between Asia

principal areas, Asia and North America (Sereno 1999). As

and North America solely by the Bering Land Bridge, we sug-

understood presently, the clade (Fig. 26.1) originated in Asia,

gest the possibility of dispersal of ceratopsians (and other

as evidenced by the presence there of the oldest, as well as

organisms) across what is now Europe during the Late Jurassic

many of the most basal, members of the clade. Later, some-

and Early Cretaceous. We do not dispute dispersal via the Ber-

what more derived members of the clade are also present in

ing Land Bridge after the mid-Cretaceous, but the presence

North America in large numbers, so one or more dispersal

of the clade in North America suggests the existence of a dis-

events from Asia to North America are understood to have

persal route before this land bridge was available. In this anal-

occurred. Evolution continued on both continents until the

ysis, Mesozoic paleogeography will be discussed, cladistic

extinction of the Ceratopsia at the end of the Cretaceous.

analyses compared, and other information gathered (pri-

The overwhelming preponderance of ceratopsian fossils are

marily paleobiogeographic information on organisms liv-

from one of these two locations, which have been the focus of

ing contemporaneously with ceratopsians); together, these

past paleobiogeographic studies. Fragmentary fossils found

provide a more complete understanding of ceratopsian paleo-

elsewhere in the world have generally not been considered

biogeography.

seriously in light of this simple, bimodal pattern. This study is a critical reappraisal of classic ceratopsian pa-

Institutional Abbreviation. RM PZ: Swedish Museum of Natural History, Department of Palaeozoology, Stockholm.

leobiogeography, including two primary topics—paleogeography itself and new (and previously underappreciated) fossil

Previous Work

evidence. The classic interpretations of ceratopsian paleobiogeography include one or more dispersal events from Asia to

Dinosaur paleobiogeography in general has been the subject

western North America over the Bering Land Bridge. This hy-

of, or included in, numerous publications throughout the last

pothesis was concise until recently, because nearly all North

century (see Russell 1993, Upchurch et al. 2002, and Holtz et

American ceratopsian fossils were known from the western

al. 2004 for comprehensive discussions and references on the

half of the continent, and all were known from no earlier than

subject). By including most recognized dinosaur taxa, these

the mid-Cretaceous, which is the most widely accepted (albeit

analyses are complex, but provide valuable information on

generalized) time of the formation of the Bering Land Bridge.

community structure and change through time and across the

Cladistic analyses of the clade also did not dispute this hy-

globe. However, not all taxa—particularly those represented

pothesis (i.e., Sereno 1999, 2000). However, in the last 10–15

only by fragmentary specimens—from each clade are included

years, many new ceratopsian fossils that complicate this rela-

because they can induce loss of resolution, and thus finer pat-

tively simple scenario have been discovered in both Asia and

terns within clades can be unclear. The following is a summary

North America, and more recent cladistic analyses similarly

of several of the more recent analyses of dinosaur paleobio-

contrast with the original hypotheses. Interestingly, ceratop-

geography

sian fossils have been discovered in several new areas across

hypotheses.

emphasizing

ceratopsian

paleobiogeographic

the world (Fig. 26.2A, D, E), calling into serious doubt the

Sereno (1999) studied vicariance and dispersal scenarios in

previous assumption that ceratopsian dinosaur dispersal did

dinosaur paleobiogeography in response to previous, sub-

not occur before the mid-Cretaceous, and then only in one

jective discussions of the topic (notably Russell 1993) and

direction—out of Asia east to North America.

used area cladograms to discern patterns. Area cladograms are

Paleontologists generally rely on published paleogeographic

cladograms that graphically arrange locations (i.e., conti-

studies and reconstructions to help determine dispersal and

nents) based on data from the fossils found in each area. The

vicariance patterns within a clade. Although geologists now

primary data used for the analyses were earlier cladistic analy-

have vast amounts of information with which to reconstruct

ses conducted by the author (i.e., Sereno 1986). Fossil locali-

the ancient globe at various time periods, lost data from inevi-

ties are plotted on the cladogram in order to discern patterns

table geologic processes, including erosion of sediment, will

of dispersal and/or vicariance, with nodes representing more

388 chinnery-allgeier & kirkland

recently shared centers of origin. Problems with this method

and an Asian-American assemblage of taxa after the Albian-

include the exclusion of landmasses (particularly ones that

Cenomanian boundary when the Bering Land Bridge became

may have played critical paleobiogeographic roles) that do

accessible (Holtz et al. 2004).

not preserve sufficient fossil dinosaur material, and the possi-

In all but the last of the above analyses, as well as others not

bilities of errors in the cladistic analyses. In addition, time is

discussed here, conclusions were influenced primarily by the

not factored in, so origins of clades, ghost lineages, and other

authors’ opinions of the relationships within the Dinosauria.

temporal data have to be inferred. In the analysis of Sereno

These data were then compared to current paleogeographical

(1999), as in the classic analyses, ceratopsians were described

information to see how they compared. In all cases, the only

as limited in distribution to Asia and western North America,

fossil information included was in the form of described and

providing a clear example of dispersal across the Bering Land

named taxa.

Bridge during the Late Cretaceous. At the time of this analysis

Although the paleobiogeography of the Ceratopsia as a

the early ceratopsians Yinlong and Chaoyangsaurus were as yet

clade has been mentioned by several recent authors (Chin-

undiscovered, and geologic evidence supported this perspec-

nery et al. 1998a; Sereno 2000; Lindgren et al 2007; Godefroit

tive (Worrall 1991).

and Lambert 2007), often the topic has been mentioned in

Upchurch et al. (2002) used Tree Reconciliation Analysis

passing or not at all, except at the lower clade level. Paleobio-

(TRA) to search for the most probable geologic connections

geography of the Ceratopsidae (excluding Turanoceratops) has

based on the area relationships in a cladogram. This method is

been an ongoing investigation for many years (see Sampson

again based on existing cladograms, with the additions of the

and Loewen this volume), while a paleobiogeographic analy-

geological and temporal position of each taxon included as

sis incorporating Asian basal ceratopsians has not yet been

data. As with area cladograms, problems with this method

attempted to any great degree. In addition to the information

include inaccurate cladograms, lack of data at any geological

gleaned from the aforementioned analyses, in this discus-

locality, and, uniquely, failure of the method if dispersal or

sion we combine cladistic relationships within the Ceratopsia,

extinction is the primary reason for the location of a taxon

geologic evidence, and paleobiogeographic information from

(the authors were looking specifically for signals of vicariance;

other (non-ceratopsian) taxa to form a concise, inclusive

Upchurch et al. 2002). Results from the TRA analyses were

hypothesis for the paleobiogeographic patterns within the

then used to reconstruct paleogeographic maps of different

Ceratopsia.

time periods. Holtz et al. (2004) used a phenetic method to build a distance matrix with localities on one axis and taxa on the

Ceratopsian Distribution

other, which was then analyzed using the clustering technique

Ceratopsians are currently known to occur in Eurasia, North

Unweighted Pair-Group Method with Arithmetic Averages

America (including Mexico), Europe, and possibly Australia

(UPGMA). Several problems are associated with this method.

(Table 26.1; Rich and Vickers-Rich 2003; You and Dodson

Holtz et al. (2004) included all taxa as equal entities, even

2004; Dodson et al. 2004; Godefroit and Lambert 2007; Lind-

though some taxa may be more important signals than others.

gren et al. 2007). The earliest basal ceratopsian Yinlong is

Some taxa were necessarily excluded in order to see certain

known from the middle to Late Jurassic (Xu et al. 2006) allow-

trends because when all were included a loss of resolution

ing for hypotheses of dispersal of ceratopsians from Asia

occurred. Only taxa that were found in at least two localities

before the mid Cretaceous. Psittacosauridae, the sister group

were included, and each locality had to have at least two diag-

to Neoceratopsia, first occurs during the early Cretaceous in

nosable taxa. Finally, a time bias occurred toward later periods,

Asia (Barremian-Albian). The oldest neoceratopsians known

since they tend to produce larger quantities of fossils. Results

from Asia, Chaoyangsaurus (possibly not a member of Neocera-

of this method are shown as cluster diagrams of localities,

topsia, but an early ceratopsian), Liaoceratops, Archaeoceratops,

which are then clustered together and discussed in terms of

and Auroraceratops also existed during the Early Cretaceous

the taxa that occur within. Significant results pertaining to

(Barremian through Aptian/Albian). Other basal neoceratop-

this paper include the clustering of Early Cretaceous North

sians from Asia include Graciliceratops, Bagaceratops, Magni-

American and European faunas (Clusters D and E; Holtz et al.

rostris, Protoceratops, and Yamaceratops, all of which are known

2004), which constituted the foundation for a hypothetical

from the Late Cretaceous. Udanoceratops, a poorly understood

connection between these two areas. However, subunits of

basal taxon, is known only from the Late Cretaceous of Asia.

Cluster L, which contain later Early Cretaceous and early

Finally, the one Asian ceratopsid genus, Turanoceratops, is in-

Late Cretaceous taxa from western North America and Asia,

teresting not only by its occurrence in Asia, but also because it

somewhat contradicted Clusters D and E. The authors inter-

occurred during the Turonian age of the early Late Cretaceous.

preted their results as supporting possible connections be-

Basal neoceratopsians in North America now include five

tween Europe and North America during the Early Cretaceous,

taxa, the oldest of which is an unnamed taxon from the Clov-

An Update on the Paleobiogeography of Ceratopsian Dinosaurs 389

Table 26.1. Locations and Chronostratigraphy of Ceratopsians (Exclusive of Ceratopsidae) Chronostratigraphy (in order)

Location—Eurasia

Location—N. America

Location—Others

Taxon

Middle–Late Jurassic

China (nw)

Yinlong

latest Jurassic or earliest

China (ne)

Chaoyangsaurus

?Valanginian to Albian

Russia (se)

Psittacosauridae

Berriasian-Albian

China (nw, nc, ne, c)

Psittacosauridae

Barremian

China (ne)

Liaoceratops

Cretaceous

early Aptian

Australia

Aptian

United States (ne)

Serendipaceratops Neoceratopsia indet. teeth

Aptian-Albian

Mongolia (c, se)

Psittacosauridae

Albian

China (c, nc)

Archaeoceratops

Albian

late Early Cretaceous

United States (nw:

Neoceratopsia indet.

Idaho)

partial skeleton

China (c, nc)

Cenomanian

Auroraceratops United States (c, w)

Neoceratopsia indet. teeth

Cenomanian-Santonian

Mongolia (se)

Yamaceratops

Cenomanian-Santonian

Mongolia (sc)

Graciliceratops

Turonian

China

Turanoceratops

Turonian

United States (sw)

Coniacian-Santonian

Zuniceratops Belgium (c)

cf. Craspedodon (Neoceratopsia teeth)

late Santonian–early

China (n, ne, c, nc)

Protoceratops

Mongolia (sc)

Protoceratops

Mongolia (sc)

Udanoceratops

Campanian late Santonian–early Campanian ?late Santonian–early Campanian earliest Campanian–

Sweden (s)

middle Campanian

teeth

Early Campanian middle Campanian

Neoceratopsia indet.

United States (nw) Mongolia (sc)

Cerasinops Bagaceratops

Late Campanian

United States (nw)

Prenoceratops

early Maastrichtian

United States (nw)

Montanoceratops

late Maastrichtian

Canada (sw; Alberta)

Leptoceratops

late Maastrichtian

United States (nw)

Leptoceratops

Campanian-

Mexico (ne); United

Ceratopsidae

Maastrichtian

States (w, includes Alaska); Canada (w)

Abbreviations: c: central; e: east; nc: northcentral; ne: northeast; nw: northwest; s: south; sc: southcentral; se: southeast; sw: southwest; w: west.

390 chinnery-allgeier & kirkland

erly Formation (Albian; Maxwell pers. com.; Cifelli pers.

graphic reconstructions of ceratopsians. Basal neoceratopsian

com.). The other four taxa are Cerasinops, Prenoceratops, Mon-

remains are now known from as early as the Albian in western

tanoceratops, and Leptoceratops, which occur during the Early

North America, but again the specimens are fragmentary and

Campanian, Late Campanian, Early Maastrichtian, and Late

thus have not been included in biogeographic discussions of

Maastrichtian, respectively. Other evidence of basal forms

the clade. The fossils known to date consist of one undescribed

includes partial skeletons and teeth from as early as the Aptian

partial skull (Maxwell pers. com.), two partial skeletons (Fig.

(Fig. 26.2; Chinnery et al. 1998b; Krumenacker 2005). These

26.2F; Weishampel et al. 2002), and teeth from the Ceno-

fragmentary fossils have been largely ignored by researchers

manian (Fig. 26.2B; Chinnery et al. 1998b). Again, although

(e.g., Weishampel 1990), but provide direct evidence of an

not informative cladistically at this time, these fossils are defi-

early appearance of neoceratopsian dinosaurs in both the east

nitely important biogeographically.

(Maryland, Aptian) and west (Utah and Idaho, Albian) of the

The Asian Udanoceratops tschizhovi consistently nests within

continent. North American ceratopsids are all known from

Leptoceratopsidae, a family that is otherwise distributed only

the Late Cretaceous only, but the sister group to Ceratopsidae,

in North America (Dodson et al. 2004; Chinnery and Horner

Zuniceratops, is Turonian in age and is thus another early form

2007). Udanoceratops is represented by an incomplete skull

(Wolfe and Kirkland 1998; Dodson et al. 2004).

lacking the dorsum, braincase, and frill, as well as undescribed

The geographic range of ceratopsians has increased recently

postcranial elements (Kurzanov 1992). Characteristics that

with the discovery of teeth and vertebrae in Sweden (Fig.

place Udanoceratops among the North American leptocera-

26.2D; Lindgren et al 2007) and teeth in Belgium (Fig. 26.2A;

topsids include shape of the dentary, curvature of the tooth

cf. Craspedodon lonzeensis, Godefroit and Lambert 2007). Two

row, and tooth occlusion pattern, among others. However, as

suspect ulnae have been found in the Early Cretaceous of Aus-

many important skull characters are unknown due to lack of

tralia that have been referred to basal Neoceratopsia (Fig.

material, caution must be used with cladistic placement of

26.2E), so the possibility exists of the occurrence of this clade

this taxon. Although the characters above show a close rela-

in Australia during the Early Cretaceous (Rich and Vickers-

tionship between Udanoceratops and Leptoceratops, the taxa

Rich 2003; but see Carpenter and Kirkland 1998).

are in actuality extremely different in overall size; shape of

Only the fossils and taxa pertinent to the present discussion

the rostral, premaxillary, and dentary bones; relative size of

will be considered here; for a comprehensive discussion of all

the nares; and shape and orientation of the jugal (Kurzanov

ceratopsian taxa known up to 2004 the reader is referred to

1992). As further specimens are discovered hypotheses of rela-

Dodson et al. (2004).

tionships might change, but for now the presence of Udano-

The taxon Chaoyangsaurus youngi (Zhao et al. 1999) is known from either the late Jurassic or earliest Cretaceous (Xu et al.

ceratops in Asia requires a dispersal event from North America, as all other and earlier leptoceratopsids are found there.

2006). This taxon is a basal ceratopsian or basal neoceratop-

Turanoceratops from Asia is the only member of Ceratopsidae

sian, positioned variously one step before or one step after

known outside of North America (Nessov et al. 1989). Origi-

Psittacosauridae in cladistic analyses (Chinnery 2004; Xu et

nally described from very fragmentary material, some of which

al. 2006; Makovicky and Norell 2006; Chinnery and Horner

has been subsequently misplaced, the taxon has been declared

2007). The Asian taxon Yinlong downsi is diagnosed as the ear-

a nomen dubium by several authors (Dodson and Currie 1990;

liest ceratopsian presently known, and is conclusively Jurassic

You and Dodson 2004; Dodson et al. 2004; Makovicky and

in age (Xu et al. 2006). These two early taxa are important

Norell 2006). The original specimens are intriguing, however,

additions to our knowledge of the temporal range of cera-

including two-rooted teeth and postorbital horncores, traits

topsian dinosaurs and therefore are important biogeographi-

found only in ceratopsids (but see discussion of Zuniceratops

cally. Their presence allows for the possibility of one or more

below; Nessov et al. 1989). We hope to learn more about

dispersal events of ceratopsians out of Asia before the middle

Turanoceratops in the near future, as apparently more material

Cretaceous, when the Bering Land Bridge was in place.

has been found recently (Sues pers. com.). The presence of a

Basal neoceratopsians were present in eastern North Amer-

ceratopsid in Asia as early as the Turonian calls into question

ica as early as the Aptian (Chinnery et al. 1998b). Although the

the geographic origin of Ceratopsidae. In the majority of the

teeth from the Arundel Formation of Maryland (Fig. 26.2C)

recent cladistic analyses (Fig. 26.1), the Asian Protoceratop-

have been known to exist for some time (Krantz 1996), the

sidae is the sister group to Zuniceratops (see below) plus Cera-

fragmentary nature of the specimens has ensured their omis-

topsidae (see also Kirkland and DeBlieux this volume). De-

sion from cladistic analyses and biogeographic interpreta-

pending of course on the relationship of Turanoceratops and

tions of the clade. Unfortunately, the Arundel Formation is

Zuniceratops, this clade is just as likely to have evolved in Asia as

unlikely to yield any more substantial evidence of ceratopsian

in North America, contrary to the accepted hypothesis of Cera-

dinosaurs (Lipka pers. com.), but even so, these teeth, along

topsidae originating in North America (Dodson et al. 2004).

with the following specimens, need to be included in biogeo-

One of the most significant recent discoveries in North

An Update on the Paleobiogeography of Ceratopsian Dinosaurs 391

FIGURE 26.1.

Cladogram of relationships among Ceratopsia, modified from Chinnery and Horner 2007.

America was that of Zuniceratops christopheri, the sister taxon

may provide more confusion than help in biogeographic

to Ceratopsidae (Wolfe and Kirkland 1998; Wolfe et al. this

reconstructions.

volume). Zuniceratops appears to be transitional between basal neoceratopsians and Ceratopsidae, exhibiting characteristics from each group. Well-developed brow horns, among other

Paleogeography

characters possessed by Zuniceratops, are found only in cera-

Paleogeography of continents has often been studied using

topsids, while lack of a nasal horn, single rooted teeth, and

geologic evidence, including plate tectonics, continent recon-

other characters exclude Zuniceratops from Ceratopsidae. The

structions, and coastline reconstructions (Smith et al. 1994;

early appearance of this transitional form (Turonian) is impor-

Scotese 2002; Blakey 2005; and references within). Different

tant biogeographically, as evolution within Neoceratopsia is

scenarios are depicted for land masses during the Late Jurassic

now known to have occurred earlier than once thought.

and Cretaceous, based on interpretation of available data. The

New discoveries of basal neoceratopsians from Europe add

following is a short review of a few of the pertinent, recent

intriguing new data to be considered. Lindgren et al. (2007)

paleogeographic studies chosen to show how interpretations

described teeth and vertebrae from the earliest and latest

of continents and coastlines are prepared.

Early Campanian of southern Sweden (Fig. 26.2D). Teeth

Geologic data for the maps of Blakey (2005) were obtained

which appear to be ceratopsian have also been described from

from a variety of sources including most of the references in

the middle Late Cretaceous of Belgium (Fig. 26.2A; Godefroit

Scotese and Sager (1989), as well as from the Paleomap Project

and Lambert 2007). These teeth are similar to those of basal

by Scotese (2002), and the data were plotted on a modern

neoceratopsians, but occur relatively later in time, and so

rectangular map that was ‘‘wrapped’’ around a globe.

392 chinnery-allgeier & kirkland

FIGURE 26.2. Representative basal neoceratopsian fossils important for questions of paleobiogeography of Ceratopsia. (A) Neoceratopsia indet. teeth from Sweden, modified from Lindgren et al. 2007; (B) Neoceratopsia indet. teeth from Utah, modified from Chinnery et al. 1998b; (C) Neoceratopsia indet. teeth from Maryland, modified from Chinnery et al. 1998b; (D) Serendipaceratops ulna; (E) cf. Craspedodon teeth, modified from Godefroit and Lambert 2007; (F) Neoceratopsia indet. partial skeleton, modified from Weishampel et al. 2002.

Scotese (2002) provides more background for his recon-

from paleoclimate studies (i.e., Fawcett et al. 1994), to ocean

structions as part of the Paleomap Project. Scotese indicates

margins (Dalziel 1991; Smith et al. 1994; Cocks et al. 1997 and

that the past positions of the continents can be determined

others) to other paleogeography studies (Wang 1985; Scotese

using evidence from paleomagnetism, linear magnetic anom-

et al. 1988; Scotese and Langford 1995).

alies, paleobiogeography, paleoclimatology, and geologic his-

Russell (1993) used the dinosaur distribution data from

tory. The references used in the Paleomap Project range widely

Weishampel (1990) along with paleogeographic maps con-

An Update on the Paleobiogeography of Ceratopsian Dinosaurs 393

structed by Scotese and Golonka (1992) to make his area map

Europe may have existed during the Jurassic and Early Creta-

diagrams.

ceous (Smith et al 1994; Scotese 2002; Blakey 2005; Godefroit

On a more regional level ratios of Ar isotopes have been

and Lambert 2007). Some reconstructions do not, however,

used to pinpoint the age of the Cedar Mountain Formation

show a possible European connection between North America

and its associated fauna, providing concrete evidence of the

and Asia (Russell 1993: figs. 2 and 3), and some organisms

presence of certain organisms in the Utah area by the Albian

seem to challenge the possibility of a connection (i.e., cryp-

(Cifelli et al. 1997; Upchurch et al. 2002; Grandstein et al.

tobranchoid salamanders; Culver and Rawson 2000).

2004; Kirkland 2005). As the fossils in the Cedar Mountain

Russell (1993) suggested the possibility that North America

Formation are closely related to taxa from the Albian of

and Europe may have been intermittently connected during

Europe and Asia, this evidence either constrains the formation

the Middle and Late Jurassic but that the Turgai Sea separated

of the Bering Land Bridge or provides evidence of a different

Asia from Europe during this time. Asian dinosaur and mam-

means of dispersal between Asia and North America.

mal taxa that appear to be endemic to Asia are cited as possible

The above studies along with others were compared to reevaluate the following questions:

evidence of isolation of the Asian landmass during the Middle and Late Jurassic, but paleogeographical information is as yet unconfirmed (Russell 1993). If this were the case, ceratopsian

1. When did the Bering Land Bridge form and thus allow

dinosaurs must have dispersed from Asia prior to the Middle

dispersal from eastern Asia to western North America?

Jurassic, originated in Europe or North America instead of

2. During which time periods was dispersal possible

Asia, or not dispersed until the Bering Land Bridge was in

from Asia to Europe and from Europe to North

place. The discovery of a ceratopsian taxon from the middle to

America (or North America to Europe)?

Late Jurassic of Asia (Yinlong; Xu et al. 2006) provides some

3. When did the inland sea bisect North America and

corroboration for the former suggestion.

prohibit movement from the west coast to the east coast (or vice versa)?

EARLY CRETACEOUS

4. How do the answers to these questions compare with other lines of evidence for the paleobiogeography of

Fig. 26.4 shows several reconstructions of land masses during

ceratopsian dinosaurs?

the Early Cretaceous (140–120 Ma). All depict the Bering Land Bridge being in existence no earlier than approximately 120

To answer these questions, we examined reconstructions of

Ma (Fig. 26.4D, North America at 140 Ma). In all reconstruc-

the globe for the Late Jurassic, Early, Middle, and early Late

tions Europe is still a group of islands in a shallow sea including

Cretaceous, focusing on three geographic areas: the Bering

the Fenno-Scandinavian Shield (Russell 1993). Russell (1993)

Land Bridge (between eastern Asia and western North Amer-

depicts a land connection between both Europe and North

ica), Europe (connections with both Asia and North America,

America and Europe and Asia (Fig. 26.4B), and in the Smith et

along with presence and size of the Atlantic Ocean and the

al. (1994) reconstruction Europe is connected with Asia (Fig.

Turgai Sea), and the interior of North America.

26.4A). Scotese does not have a reconstruction for this time period, but on his 3-dimensional globes (www.scotese.com)

LATE JURASSIC

the Bering Land Bridge is in place by 140 Ma, breaks apart around 120 Ma, and then reappears again during the Late

Fig. 26.3 depicts various interpretations of continent posi-

Cretaceous (Scotese 2002). Despite the differences in the paleo-

tions and shorelines during the Late Jurassic (150–153 million

geographic reconstructions for this time period there is abun-

years ago). In Fig. 26.3A, no land connections are present

dant evidence linking Early Cretaceous North American and

between western North America and eastern Asia, and Europe

European taxa (Winkler et al. 1988; Weishampel and Bjork

is depicted as a group of large islands in a shallow sea. The

1989; Weishampel et al. 1991; Jerzykiewicz and Russell 1991;

other reconstructions of the Late Jurassic northern hemi-

Russell 1993).

sphere are very similar; the only difference among them are

During the late Early Cretaceous a marine transgression

the shapes and locations of the European islands (Fig. 26.3B–

resulted in the separation of eastern and western North Amer-

D). Two reconstructions of this time period are shown by

ica by a midcontinental sea. Interpretations of the extent of

Blakey (2005), as the North Atlantic series of reconstructions

this sea at this time differ in the reconstructions presented

is separate from the North American series. A transient route

here (Fig. 26.4A, C, D). In the Blakey reconstructions (Fig.

between Asia and eastern North America is not unfeasible

26.4C, D) the interior seaway was not in evidence 140 Ma (Fig.

with these scenarios, especially with the reconstruction of the

26.4D) but does extend through the majority of the continent

North Atlantic region during the Late Jurassic (Fig. 26.3C),

by 120 Ma (Fig. 26.4C). In contrast to these, Fig. 26.4A shows

and many scientists agree that transient routes through

only partial separation of North America by 120 Ma. Timing

394 chinnery-allgeier & kirkland

FIGURE 26.3. Reconstructions of the globe or regions during the Late Jurassic. (A) Reconstruction of the globe at 152 Ma (modified from Scotese 2002); (B) reconstruction of the globe at 153 Ma (modified from Smith et al. 1994); (C) reconstruction of the North Atlantic region at 150 Ma (modified from Blakey 2005); (D) reconstruction of North America at 150 Ma (modified from Blakey 2005).

of the separation of eastern and western North America is an

the European islands are still separated from North America

important consideration in the study of ceratopsian paleo-

and Asia. Contrary to these maps, Smith et al. (1994) depict

biogeography, because of the early appearance of the clade

the interior seaway as completely dividing the eastern and

in the northeastern part of the continent (Chinnery et al.

western parts of North America by 105 Ma (Fig. 26.5A illus-

1998b). Basal neoceratopsian teeth are present in Maryland

trates the globe at 95 Ma). Also, at 105 Ma, Europe and Asia are

from the Aptian, around 120 Ma. As discussed below, the Ber-

still connected in the Smith et al. (1994) and Russell (1993)

ing Land Bridge most likely did not exist before 125 Ma at the

reconstructions. The Bering Land Bridge is firmly in place by

earliest. If ceratopsian dinosaurs did disperse from Asia to

the Aptian according to Smith et al. (1994; 120 Ma, although

North America solely by this route, only five million years was

this area of the globe is unclear) and Russell (1993), but the

available for them to reach the east coast.

Blakey reconstructions do not show a definitive land bridge occurring until around 90 Ma (Fig. 26.6C, D).

MIDDLE CRETACEOUS During the Middle Cretaceous (Fig. 26.5), the North American

EARLY LATE CRETACEOUS

interior seaway is depicted as increasing in width and length

Fig. 26.6 shows reconstructions of the globe or portions of the

at 115 Ma (Fig. 26.5C) and then decreasing in both dimen-

globe during the early Late Cretaceous (94–85 Ma). All re-

sions by 100 Ma (Fig. 26.5D). Through these two time periods

constructions include the Bering Land Bridge, and all show

An Update on the Paleobiogeography of Ceratopsian Dinosaurs 395

FIGURE 26.4. Reconstructions and diagrams of the globe or regions during the Early Cretaceous. (A) Reconstruction of the globe, 120 Ma (modified from Smith et al. 1994); (B) diagrammatic representation of regional connections, late Early Cretaceous (modified from Russell 1993); (C) reconstruction of the North Atlantic region, 120 Ma (modified from Blakey 2005); (D) reconstruction of North America, 140 Ma (modified from Blakey 2005).

slightly more separation of the European land masses and

ing any specific time periods, the evidence does not preclude

North America. From this time period on, all evidence sug-

the possibility of such connections occurring. Finally, North

gests that dispersal events within Ceratopsia and other dino-

America was at least partly divided into eastern and western

saur and mammal clades occurred across the Bering Land

land masses by the late Early Cretaceous. This is significant

Bridge (Smith et al. 1994; You and Dodson 2004; Dodson et al.

because the earliest ceratopsian fossils known currently from

2004).

North America are from the Aptian of the northeastern part of

Information from the various reconstructions of Mesozoic

the continent.

geography can provide some putative answers for our original questions. First, most current geographic scenarios support the hypothesis that the Bering Land Bridge was not in place until approximately the end of the Albian (Cifelli et al. 1997;

Mesozoic Dispersal Patterns in Non-Ceratopsians

Upchurch et al. 2002; Gradstein et al. 2004; Kirkland 2005).

One of the more compelling lines of evidence for a possible

Second, although the geology of the European region is not

dispersal route of ceratopsian dinosaurs through Europe is the

adequately known at this time to provide concrete evidence

dinosaur assemblage of the Cedar Mountain Formation of

of connections among Europe, North America, and Asia dur-

Utah, which includes neoceratopsian teeth (Chinnery et al.

396 chinnery-allgeier & kirkland

FIGURE 26.5.

Reconstructions and diagrams of the globe or regions during the Middle Cretaceous. (A) Reconstruction of the globe, 95 Ma (modified from Smith et al. 1994); (B) diagrammatic representation of regional connections, Middle Cretaceous (modified from Russell 1993); (C) reconstruction of North America, 115 Ma (modified from Blakey 2005); (D) reconstruction of North America, 100 Ma (modified from Blakey 2005).

FIGURE 26.6.

Reconstructions of the globe or regions during the Late Cretaceous. (A) Reconstruction of the globe at 94 Ma (modified from Scotese 2002); (B) reconstruction of the globe at 90 Ma (modified from Smith et al. 1994); (C) reconstruction of the North Atlantic region, 90 Ma (modified from Blakey 2005); (D) North America at 85 Ma (modified from Blakey 2005).

An Update on the Paleobiogeography of Ceratopsian Dinosaurs 397

1998b). Other dinosaurs found in this formation are the anky-

for the phylogenetic relationships to be upheld (i.e., Sereno

losaur Gastonia burgei (Kirkland 2005), Cedarosaurus weiskopfae

2000). Problems with using this approach include differences

(Tidwell et al. 1999) among other sauropods including titano-

among cladograms based on the characters chosen and the

saurids and possibly camarosaurids, and advanced iguan-

character states interpreted by each researcher, lack of pale-

odonts (including ‘‘Iguanodon’’ ottingeri; Kirkland 2005). Also

ogeographic congruence, and the exclusion of fossils that are

present are theropods including the coelurosaur Nedcolbertia

too fragmentary to be included in a cladistic analysis.

justinhoffmani (Kirkland et al. 1998), the therizinosauroid Fal-

Of the many cladistic analyses provided in recent publica-

carius utahensis (Kirkland et al. 2005), and a large carnosaur.

tions (Chinnery and Weishampel 1998; Zhao et al. 1999;

Many of the Cedar Mountain Formation dinosaurs have affini-

Sereno 2000; Makovicky 2001; Xu et al. 2002, 2006; You and

ties with European taxa. Cedarosaurus and Venenosaurus are

Dodson 2003; Chinnery 2004; Makovicky and Norell 2006;

both related to Pelorosaurus, a sauropod of similar age from En-

Chinnery and Horner 2007), differences occur due to variation

gland (Carpenter and Tidwell 2005). The advanced iguan-

in the characters used and how the characters are coded by

odonts bear teeth similar to those found in the Albian-

each researcher. For the most part the cladograms are similar

Cenomanian of England (‘‘Iguanodon’’ hilli and ‘‘Trachodon’’

due to the recycling of many characters (most character

cantabrigiensis), and therizinosauroids have until recently

matrices are based on earlier analyses but include additions

been known only from Asia. Gastonia is very similar to Pola-

and corrections). However, one major issue for pale-

canthus from western Europe (Blows 1996; Carpenter 2001),

obiogeographic analyses has come to the forefront: the relative

and Nedcolbertia resembles European coelurosaurs from the

positions of Leptoceratopsidae and Protoceratopsidae to

same time period (Kirkland and Madsen 2007). The European

Ceratopsidae.

affinities of these and other dinosaurs from the Cedar Moun-

Coronosauria is the clade that includes the most recent com-

tain Formation in Utah support a connection between North

mon ancestor of Protoceratops andrewsi and Triceratops horridus,

America and Europe 120–125 million years ago, before the

and all of its descendents (Sereno 1998). In some analyses

Bering Land Bridge was established (Plafker and Berg 1994;

(Sereno 2000; You and Dodson 2003; Dodson et al. 2004),

Kirkland and Madsen 2007).

Coronosauria includes Leptoceratopsidae, or one or more taxa

Presently, all members of Pachycephalosauridae (the sister

included in this clade (taxa included in this clade are Mon-

group to Ceratopsia) are known from the Late Cretaceous of

tanoceratops, Prenoceratops, Udanoceratops, Leptoceratops, and

Asia or North America. However, two possible problematic

Cerasinops; Fig. 26.1). This is especially interesting when

pachycephalosaurids, Yaverlandia bitholu from the Wealden

viewed from a paleobiogeographic aspect because if any mem-

Formation (Barremian) of England and Stenopelix valdensis

bers of the current Leptoceratopsidae are more closely related

from the Barremian of Germany, may indicate a more global

to Ceratopsidae than any other taxa are, it would mean that

distribution of this clade (but see Sullivan 2006).

Ceratopsidae evolved from the same common ancestor as that

Other organisms have similar geographic ranges and may indicate dispersal from Asia to Europe during or before the

leptoceratopsid taxon. Only one initial dispersal event from Asia to North America is required for this scenario.

early Cretaceous. For example, gobiconodontid mammals are

However, in other analyses Protoceratops or one or more

now known from Europe and England, which is of interest as

members of Protoceratopsidae are more closely related to Tri-

all other members of this clade are known from Asia and

ceratops than are any leptoceratopsids, thus excluding Lep-

North America (Sweetman 2006). The oldest members of this

toceratopsidae from Coronosauria (taxa in Protoceratopsidae

mammal group are from the Hauterivian of Asia (Cuenca-

are Graciliceratops, Protoceratops, and Bagaceratops; Makovicky

Bescos and Canudo 2003). Dispersal of Asian forms to Europe

2001; Xu et al. 2002; Chinnery 2004; Makovicky and Norell

probably occurred during the Barremian, and Cuenca-Bescos

2006; Chinnery and Horner 2007). If this second cladistic

and Canudo (2003) theorize that a second dispersal event

hypothesis is correct, Ceratopsidae must have evolved from a

brought gobiconodontids to North America across the Bering

common ancestor shared with Protoceratopsidae. More cera-

Land Bridge during the Aptian or Albian. However, the evi-

topsian dispersal events from Asia would have necessarily

dence can also support a dispersal event from Europe to North

occurred, one leading to Leptoceratopsidae and one leading to

America.

Ceratopsidae. Using the most recently published cladogram (Chinnery and Horner 2007), the latter hypothesis is more

Ceratopsian Relationships

parsimonious. A cladogram can only provide partial information on the

One approach to determining the paleobiogeography of Cera-

paleobiogeography of ceratopsians, as many important, usu-

topsia is by looking at the cladogram produced by a phylo-

ally fragmentary, fossils are not typically included in analyses.

genetic analysis and, based solely on this information, con-

These will now be discussed in terms of their affinities and

cluding what dispersal patterns must have occurred in order

possible relationships to other, better known ceratopsians.

398 chinnery-allgeier & kirkland

The teeth from the Arundel Formation in Maryland and the Cedar Mountain Formation in Utah have been positively

pronounced cingulum set back from the primary ridge otherwise found only in leptoceratopsids.

identified as neoceratopsian based on characteristics includ-

Finally, teeth from Belgium previously described as the

ing an oval shape of the crown in lateral view, a pronounced

iguanodont Craspedodon lonzeensis (Dollo 1883) have been

cingulum, indentations on either side of the primary ridge,

recently redescribed as neoceratopsian (Godefroit and Lam-

and especially the convex labial sides of the teeth (Fig. 26.2B,

bert 2007). These teeth appear to be most similar to neo-

C; Chinnery et al. 1998b; Smith and Dodson 2003). The wear

ceratopsian teeth (Fig. 26.2A; Godefroit and Lambert 2007:

patterns of the Arundel teeth are the same ‘‘vertical-notch’’

figs. 2 and 3), with similar enamel distribution, shape in mes-

wear patterns found on the dentary teeth of all described

ial or distal view, and well-developed secondary ridges. The

North American basal neoceratopsians (as well as Udanocera-

oblique shear wear pattern is similar to that seen in Asian

tops from Asia; Fig. 26.2C; Sternberg 1951; Kurzanov 1992;

neoceratopsian teeth and in maxillary teeth of leptoceratop-

Chinnery and Weishampel 1998; Chinnery 2004; Chinnery

sids, and the presence of a well-developed cingulum set apart

and Horner 2007). The wear patterns of the Cedar Mountain

from the primary ridge on these teeth place them closer to

teeth, in contrast, are of oblique sheer, which could mean that

members of Leptoceratopsidae.

they are maxillary teeth or that the animal to which these

Based on the characteristics discussed above, the teeth

teeth pertain has a different ancestry from the other basal

described above have been placed in a hypothetical (not statis-

forms in North America (Fig. 26.2B).

tical or formally cladistic) phylogeny plotted against geologi-

The newly described teeth from Sweden (Fig. 26.2D; Lind-

cal time (Fig. 26.7). Craspedodon teeth have been placed as the

gren et al. 2007) exhibit some of the aforementioned charac-

basalmost leptoceratopsid and the Swedish tooth has been

ters but with differences (see Lindgren et al. 2007: fig. 3). The

placed between the Psittacosauridae and Liaoceratops. The

teeth have enamel on both sides of the crowns, a characteristic

other two tooth taxa—the Aptian teeth from Maryland and

of ornithopod teeth but also of very basal ceratopsian teeth

the Albian teeth from Utah—have been placed as unresolved

(Chaoyangsaurus and members of Psittacosauridae retain this

taxa one step below Leptoceratopsidae on the phylogeny.

plesiomorphic trait, whereas all other neoceratopsians have

Though it is not standard (or recommended) practice to place

enamel only on the non-occluding surface; Chinnery and

taxa on a cladogram a priori, we do so here only to permit the

Horner 2007). The enamel on the Swedish teeth is thicker on

reader to explore the possible relationships of these fragmen-

the lingual side than on the labial side, which could represent

tary fossils and what their existence might mean for the geo-

an intermediate character state. The teeth are wide in a labio-

graphical and temporal distribution of the Ceratopsia.

lingual direction as are other basal neoceratopsian teeth, but all are slightly concave on the labial side, which is a characteristic of ornithopod teeth (Chinnery and Weishampel

Discussion

1998). They all have prominent primary ridges and a cingu-

As reported in 1998 (Chinnery et al. 1998a), several dispersal

lum, although the cingulum is only pronounced on one speci-

events must have occurred in order to reconcile the distribu-

men, RM PZ R1833. In all specimens except for one (RM PZ

tions of and relationships among ceratopsian dinosaurs (aside

R1833), the primary ridge is confluent with the cingulum, a

from the Australian possible ceratopsian Serendipaceratops;

characteristic found in all ceratopsians (and other ornithischi-

Chinnery et al. 1998a). Up to two dispersal events brought

ans) except for leptoceratopsids. The wear pattern of four of

ceratopsians from their point of origin in Asia to North Amer-

the five teeth shown is of a vertical-notch or oblique-notch

ica; (1) the ancestor of the Leptoceratopsidae lineage, and

pattern, but is quite different from the wear pattern found in

(2) the ancestor of the Ceratopsidae lineage. If Turanoceratops

leptoceratopsids. The wear pattern of RM PZ R1833 is oblique

is the sister taxon of Ceratopsidae, then the ceratopsid lineage

as in maxillary teeth of North American taxa and all teeth of

may have originated in Asia despite the preponderance of

Asian basal forms. The locality of these specimens is quite

occurrences in North America. If not, then a dispersal event

distant from all but one other neoceratopsian locality (see

back to Asia led to Turanoceratops, and possibly another to

below), and if the taxon is ceratopsian, autapomorphies

Udanoceratops. The presence of ceratopsians in Europe is most

might therefore be expected in a distinct population. How-

likely due to a dispersal event from Asia based on the primitive

ever, it is our opinion that although RM PZ R1833 appears to

status of the specimens.

be neoceratopsian, the other teeth are not demonstrably neo-

How these dispersal events took place is more difficult to

ceratopsian. The vertebrae described by Lindgren et al. (2007)

discern, however, because the animals could have traveled

are not diagnostic. The Swedish fossil will be included in this

between eastern Asia and Siberia through Alaska and to the

analysis and is tentatively placed at the base of the Neocera-

west coast of North America, or they could have moved

topsia due to the possession of enamel on both sides of the

between the east coast of North America and Europe. Most

tooth (Fig. 26.7). It should be noted that RM PZ R1833 has a

likely, both routes were used (see Paleogeography section

An Update on the Paleobiogeography of Ceratopsian Dinosaurs 399

FIGURE 26.7.

Relationships among ceratopsian dinosaurs, with time included. Based on the cladogram in Fig. 26.1. Taxa followed with an ‘‘*’’ have not been included in any cladistic analyses yet, and their positions relative to other ceratopsians are tentative.

400 chinnery-allgeier & kirkland

above). The first dispersal events most likely occurred during

Conclusion

or prior to the Aptian because the oldest ceratopsian fossils currently known from outside Asia (teeth from the Arundel

In summary, we suggest the following steps for ceratopsian

Formation, Maryland; Chinnery et al. 1998b) are known from

paleobiogeography:

this time. In order for neoceratopsians to be present at this time and location, dispersal across the Bering Land Bridge

1. The origin and initial radiation of the clade in Asia,

would have to have taken place during the early Aptian, with

evidenced by the occurrences there of the earliest and

a subsequent immediate movement of these dinosaurs across

most basal members of the clade as well as its sister

the entire North American continent. Geologic evidence is

groups.

becoming fairly conclusive that the Bering Land Bridge was

2. A dispersal event of basal neoceratopsians from

not in place at this time, and possibly not for another 15 mil-

western Asia through Europe to eastern North

lion years (Cifelli et al. 1997; Upchurch et al. 2002; Gradstein

America, and then a spread across North America

et al. 2004). Thus, the most parsimonious explanation for the

of the ancestor of the known Leptoceratopsidae.

early presence of neoceratopsian fossils in North America is an

This dispersal event may have resulted in some

alternate dispersal route, most likely across Europe during

populations staying in Europe and establishing

the Late Jurassic or Early Cretaceous. Terrestrial deposits and

separate endemic groups (as evidenced by the unusual

fossils of these ages on the European land mass are not well

European teeth), but the lack of early neoceratopsian

known, but based on the majority of paleogeographic and

fossils in otherwise fossiliferous Early Cretaceous

geologic information available, the epieric Turgai Sea between Asia and Europe was spanned by transitory land bridges, as

European localities suggests otherwise. 3. A radiation of basal neoceratopsians in Asia, possibly

was the emerging Atlantic Ocean between Europe and North

including the origin of the ancestral ceratopsoid

America.

(Turanoceratops or a close relative) and dispersal of one

Neoceratopsian dinosaurs are not the only organisms to

or more taxa to North America across the Bering Land

appear in North America earlier than possible via the Bering

Bridge when it became accessible in or around the

Land Bridge. Evidence is accumulating for taxa with inferred

Turonian. This dispersal event brought ancestral

Asian origins appearing in North America earlier than previously thought (i.e., therizinosaurs) In addition, close affini-

ceratopsids to North America. 4. A major radiation of ceratopsids in North America,

ties of North American taxa and Western European taxa are

and possibly one or more dispersal events in which a

numerous (see above), arguing for movement of organisms in

ceratopsid emigrated back to Asia (if Turanoceratops is

one or both directions between these two continents during

not an ancestor to Ceratopsidae). Another dispersal

this time frame.

event of a leptoceratopsid back to Asia may have

Ceratopsians are now known from Europe, but are younger in age (mid- and middle Late Cretaceous; Lindgren et al. 2007;

occurred, explaining the presence of Udanoceratops, but this is unconfirmed at this time.

Godefroit and Lambert 2007) than some from North America.

5. One or more dispersal events from eastern North

Although both Lindgren et al. (2007) and Godefroit and Lam-

America to Europe during the mid-Cretaceous.

bert (2007) argue that the presence of neoceratopsians in Europe indicates dispersal events (one or more) from Asia dur-

References Cited

ing the Early Cretaceous, it may be possible that dispersal

Blakey, R. C. 2005. http://jan.ucc.nau.edu/approximately rcb7/RCB.html. Blows, W. T. 1996. A new species of Polacanthus (Ornithischia: Ankylosauria) from the Lower Cretaceous of Sussex, England. Geological Magazine 133: 671–682. Carpenter, K. 2001. Skull of the polacanthid ankylosaur Hylaeosaurus armatus Mantell, 1883, from the Lower Cretaceous of England. In K. Carpenter, K. F. Hirsch, and J. R. Horner, eds., The Armored Dinosaurs, pp. 169–172. Bloomington: Indiana University Press. Carpenter, K., and J. I. Kirkland. 1998. Review of Lower and Middle Cretaceous ankylosaurs from North America. In S. G. Lucas, J. I. Kirkland, and J. W. Estep, eds., Lower and Middle Cretaceous Terrestrial Ecosystems, pp. 249–270. New Mexico Museum of Natural History and Science Bulletin 14.

occurred from North America, due to the presence of basal neoceratopsian teeth in Maryland during the Aptian. Finally, we want to stress that the absence of fossil material in cladistic analyses does not mean that relevant material does not exist or that it is unimportant, only that it is not useful phylogenetically at the time of the analysis. In this case, fragmentary fossil material that is of limited phylogenetic utility nevertheless clearly has paleobiogeographic importance, exemplifying a more paleogeographically and temporally extensive distribution of ceratopsian dinosaurs than can be ascertained from cladistic analyses alone. Ghost lineages imply that abundant fossils of early ceratopsians have yet to be discovered, and future finds will hopefully settle these issues.

An Update on the Paleobiogeography of Ceratopsian Dinosaurs 401

Carpenter, K., and V. Tidwell. 2005. Reassessment of the Early Cretaceous sauropod Astrodon johnsoni Leidy 1865 (Titanosauriformes). In V. Tidwell and K. Carpenter, eds., Thunder-Lizards: The Sauropodomorph Dinosaurs, pp. 78–114. Bloomington: Indiana University Press. Chinnery, B. J. 2004. Description of Prenoceratops pieganensis gen. et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Chinnery, B. J., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Chinnery, B. J., T. R. Lipka, J. I. Kirkland, J. M. Parrish, and M. K. Brett-Surman. 1998b. Neoceratopsian teeth from the lower to middle Cretaceous of North America. In S. G. Lucas, J. I. Kirkland, and J. W. Estep, eds., Lower and Middle Cretaceous Terrestrial Ecosystems, pp. 297–302. New Mexico Museum of Natural History and Science Bulletin 14. Chinnery, B. J., and D. B. Weishampel. 1998. Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) and relationships among basal neoceratopsians. Journal of Vertebrate Paleontology 18: 569–585. Chinnery, B. J., D. G. Wolfe, and J. I. Kirkland. 1998a. A new consideration of neoceratopsian biogeography. In D. L. Wolberg, K. Gittis, S. Miller, L. Carey, and A. Raynor, eds., The DinoFest Symposium, p. 7. Philadelphia: Academy of Natural Sciences. Cifelli, R. L., J. I. Kirkland, A. Weil, A. L. Deino, and B. J. Kowallis. 1997. High-precision 40Ar/ 39Ar geochronology and the advent of North America’s Late Cretaceous terrestrial fauna. Proceedings of the National Academy of Sciences 94: 11163–11167. Cocks, L. R. M., W. S. McKerrow, and C. R. van Staal. 1997. The margins of Avalonia. Geological Magazine 134: 627–634. Cuenca-Bescós, C., and J. I. Canudo. 2003. A new gobiconodontid mammal from the Early Cretaceous of Spain and its palaeogeographic implications. Acta Paleontologica Polonica 48: 575–582. Culver, S. J., and P. F. Rawson. 2000. Biotic Response to Global Change: The Last 145 Million Years. Cambridge: Cambridge University Press. Dalziel, I. W. D. 1991. Pacific margins of Laurentia and East Antarctica-Australia as a conjugate rift pair: Evidence and implications for an Eocambrian supercontinent. Geology 19: 598–601. Dodson, P., and P. J. Currie. 1990. Basal Neoceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 593–618. Berkeley: University of California Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–516. Berkeley: University of California Press. Dollo, L. 1883. Note sur les restes de dinosauriens rencontrés dans le Crétacé supérieur de la Belgique. Bulletin de L’Institut Royal des Sciences Naturelles de Belgique 2: 205–221. Eberth, D. A., Y. Kobayashi, Y.-N. Lee, O. Mateus, F. Therrien, D. K. Zelenitsky, and M. A. Norell. 2009. Assignment of Yamaceratops dorngobiensis and associated redbeds at Shine Us

402 chinnery-allgeier & kirkland

Khudag (Eastern Gobi, Dorngobi Province, Mongolia) to the redescribed Javkhlant Formation (Upper Cretaceous). Journal of Vertebrate Paleontology 29: 295–302. Fawcett, P. J., E. J. Barron, V. D. Robinson, and B. J. Katz. 1994. The climatic evolution of India and Australia from the Late Permian to mid-Jurassic: A comparison of climate model results with the geologic record. In G. D. Klein, ed., Pangea: Paleoclimate, Tectonics, and Sedimentation during Accretion, Zenith, and Breakup of a Supercontinent, pp. 139–157. Geological Society of America Special Paper 288. Godefroit, P., and O. Lambert. 2007. A re-appraisal of Craspedodon lonzeensis Dollo, 1883 from the Upper Cretaceous of Belgium: The first record of a neoceratopsian dinosaur in Europe? Bulletin de L’Institut Royal des Sciences Naturelles de Belgique 77: 83–93. Gradstein, F. M., J. G. Ogg, and A. G. Smith. 2004. A Geologic Time Scale 2004. Cambridge: Cambridge University Press. Holtz, T. R., R. E. Chapman, and M. C. Lamanna. 2004. Mesozoic biogeography of Dinosauria. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 627–642. Berkeley: University of California Press. Jerzykiewicz, T., and D. A. Russell. 1991. Late Mesozoic stratigraphy and vertebrates of the Gobi Basin. Cretaceous Research 12: 345–377. Kirkland, J. I. 2005. Utah’s newly recognized dinosaur record from the Early Cretaceous Cedar Mountain Formation. Utah Geological Survey, Survey Notes 37: 1–6. Kirkland, J. I., B. B. Britt, C. H. Whittle, S. K. Madsen, and D. L. Burge. 1998. A small coelurosaurian theropod from the Yellow Cat Member of the Cedar Mountain Formation (Lower Cretaceous, Barremian) of eastern Utah. In S. G. Lucas, J. I. Kirkland, and J. W. Estep, eds., Lower to Middle Cretaceous Non-marine Cretaceous Faunas, pp. 239–248. New Mexico Museum of Natural History and Science Bulletin 14. Kirkland, J. I., and D. D. DeBlieux. 2010. New basal centrosaurine ceratopsian skulls from the Wahweap Formation (middle Campanian), Grand Staircase–Escalante National Monument, southern Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 117–140. Bloomington: Indiana University Press. Kirkland, J. I., and S. K. Madsen. 2007. The Lower Cretaceous Cedar Mountain Formation, eastern Utah: The view up an always interesting learning curve. Geological Society of America Rocky Mountain Section Annual Meeting, St. George, Utah, May 4– 6, 2007. Kirkland, J. I., L. E. Zanno, S. D. Sampson, J. M. Clark, and D. D. DeBlieux. 2005. A primitive therizinosauroid dinosaur from the Early Cretaceous of Utah. Nature 435: 84–87. Krantz, P. M. 1996. Notes of the sedimentary iron ores of Maryland and their dinosaurian fauna. Maryland Geological Survey Special Paper 3: 87–115. Krumenacker, L. J. 2005. Preliminary report on new vertebrates from the upper Gannett Group (Aptian) and Wayan Formation (Albian) of East Idaho. Paludicola 5: 55–64. Kurzanov, C. M. 1992. A gigantic protoceratopsid from

the Upper Cretaceous of Mongolia. Paleontological Journal 26: 103–116. Lindgren, J., P. J. Currie, M. Siverson, J. Rees, P. Cederstrom, and F. Lindgren. 2007. The first neoceratopsian dinosaur remains from Europe. Palaeontology 50: 929–937. Makovicky, P. J. 2001. A Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) braincase from the Horseshoe Canyon Formation of Alberta. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 243–262. Bloomington: Indiana University Press. Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngobiensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Nessov, L. A., L. F. Kaznyshkina, and G. O. Cherepanov. 1989. Mesozoic ceratopsian dinosaurs and crocodilians of central Asia. In T. N. Bogdanova and L. I. Khozatsky, eds., Theoretical and Applied Aspects of Modern Paleontology, pp. 144–154. Moscow. [In Russian.] Plafker, G., and H. C. Berg. 1994. The geology of Alaska. Geological Society of America, Decade of North American Geology, Vol. G-1. Rich, T. H., and P. Vickers-Rich. 2003. Protoceratopsian? ulnae from Australia. Records of the Queen Victoria Museum 113. Russell, D. A. 1993. The role of Central Asia in dinosaurian biogeography. Canadian Journal of Earth Sciences 30: 2002–2012. Sampson, S. D., and M. A. Loewen. 2010. Unraveling a radiation: A review of the diversity, stratigraphic distribution, biogeography, and evolution of horned dinosaurs (Ornithischia: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 405–427. Bloomington: Indiana University Press. Scotese, C. R. 2002. http://www.scotese.com (Paleomap website). Scotese, C. R., L. M. Gahagan, and R. L. Larson. 1988. Plate tectonic reconstructions of the Cretaceous and Cenozoic ocean basins. In C. R. Scotese and W. W. Sager, eds., Mesozoic and Cenozoic Plate Reconstructions, pp. 27–48. Tectonophysics 155. Scotese, C. R., and J. Golonka. 1992. Paleogeographic Atlas. Paleomap Progress Report 20-0692, Department of Geology, University of Texas at Arlington. Scotese, C. R., and R. P. Langford. 1995. Pangea and the Paleogeography of the Permian. In P. A. Scholle, T. M. Peryt, and D. S. Ulmer-Scholle, eds., The Permian of Northern Pangea, V.1, Paleogeography, Paleoclimates, and Stratigraphy, pp. 3–19. Berlin: Springer-Verlag. Scotese, C. R., and W. W. Sager, eds. 1989. Mesozoic and Cenozoic Plate Reconstructions. Amsterdam: Elsevier. Sereno, P. C. 1986. Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). National Geographic Research 2: 234–256. ———. 1998. A rationale for Phylogenetic definitions with application to the higher-level taxonomy of Dinosauria. Neues Fahrbuch für Geologie and Paläontologie Abhandlungen 210: 41–83. ———. 1999. Dinosaurian biogeography: Vicariance, dispersal, and

regional extinction. In Y. Tomida, T. H. Rich, and P. VickersRich, eds., Proceedings of the Second Gondwanan Dinosaur Symposium, pp. 249–257. Tokyo: National Science Museum Monographs No. 15. ———. 2000. The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia. In M. J. Benton, M. A. Shishkin, D. M. Unwin, and E. N. Kurochkin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. 480–516. Cambridge: Cambridge University Press. Smith, A. G., D. G. Smith, and B. M. Funnell. 1994. Atlas of Mesozoic and Cenozoic Coastlines. Cambridge: Cambridge University Press. Smith, J. B., and P. Dodson. 2003. A proposal for a standard terminology of anatomical notation and orientation in fossil vertebrate dentitions. Journal of Vertebrate Paleontology 23: 1–12. Sternberg, C. M. 1951. Complete skeleton of Leptoceratops gracilis Brown from the Upper Edmonton member on the Red Deer River, Alberta. Bulletin of the National Museum of Canada 123: 225–255. Sullivan, R. M. 2006. A taxonomic review of the Pachycephalosauridae (Dinosauria: Ornithischia). In S. G. Lucas and R. M. Sullivan, eds., Late Cretaceous Vertebrates from the Western Interior, pp. 347–365. New Mexico Museum of Natural History and Science Bulletin 35. Sweetman, S. C. 2006. A gobiconodontid (Mammalia, Eutriconodonta) from the Early Cretaceous (Barremian) Wessex Formation of the Isle of Wight, southern Britain. Palaeontology 49: 889–897. Tidwell, V., K. Carpenter, and W. Brooks. 1999. New sauropod from the Lower Cretaceous of Utah, USA. Oryctos 2: 21–37. Upchurch, P., C. A. Hunn, and D. B. Norman. 2002. An analysis of dinosaurian biogeography: Evidence for the existence of vicariance and dispersal patterns caused by geologic events. Proceedings of the Royal Society of London 269: 613–621. Wang, H. 1985. Atlas of the Paleogeography of China. Beijing: Cartographic Publishing House, Institute of Geology, Chinese Academy of Sciences, Wuhan College of Geology. Weishampel, D. B. 1990. Dinosaurian distribution. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 63–140. Berkeley: University of California Press. Weishampel, D. B., and P. R. Bjork. 1989. The first indisputable remains of Iguanodon (Ornithischia: Ornithopoda) from North America: Iguanodon lakotensis, sp. nov. Journal of Vertebrate Paleontology 9: 56–66. Weishampel, D. B., D. Grigorescu, and D. B. Norman. 1991. The dinosaurs of Transylvania: Island biogeography in the Late Cretaceous. National Geographic Research and Exploration 7: 196–215. Weishampel, D. B., M. B. Meers, W. A. Akersten, and A. D. McCrady. 2002. New Early Cretaceous dinosaur remains, including possible ceratopsians, from the Wayan Formation of eastern Idaho. In W. A. Akersten, M. E. Thompson, D. J. Meldrum, R. A. Rapp, and H. G. McDonald, eds., And Whereas . . . Papers on the Vertebrate Paleontology of Idaho Honoring John A. White, Vol. 2, pp. 5–17. Idaho Museum of Natural History Occasional Paper 37.

An Update on the Paleobiogeography of Ceratopsian Dinosaurs 403

Winkler, D. A., L. L. Jacobs, J. R. Branch, P. A. Murray, W. R. Downs, and P. Trudel. 1988. The Proctor Lake dinosaur locality, Lower Cretaceous of Texas. Hunteria 2: 1–8. Wolfe, D. G., and J. I. Kirkland. 1998. Zuniceratops christopheri n. gen. & n. sp., a ceratopsian dinosaur from the Moreno Hill Formation (Cretaceous, Turonian) of west-central New Mexico. New Mexico Museum of Natural History and Science Bulletin 14: 303–317. Wolfe, D. G., J. I. Kirkland, D. Smith, K. Poole, B. J. ChinneryAllgeier, and A. McDonald. 2010. Zuniceratops christopheri: The North American Ceratopsid sister taxon reconstructed on the basis of new data. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 91–98. Bloomington: Indiana University Press. Worrall, D. M. 1991. Tectonic history of the Bering Sea and the evolution of Tertiary strike-slip basins of the Bering Shelf. Geological Society of America Special Paper 257: 1–106.

404 chinnery-allgeier & kirkland

Xu, X., C. A. Forster, J. M. Clark, and J. Mo. 2006. A basal ceratopsian with transitional features from the Late Jurassic of northwestern China. Proceedings of the Royal Society B 273: 2135–2140. Xu, X., P. J. Makovicky, X.-L. Wang, M. A. Norell, and H.-L. You. 2002. A ceratopsian dinosaur from China and the early evolution of Ceratopsia. Nature 416: 314–317. You, H., and P. Dodson. 2003. Redescription of neoceratopsian dinosaur Archaeoceratops and early evolution of Neoceratopsia. Acta Paleontologica Polonica 48: 261–272. ———. 2004. Basal Ceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press. Zhao, X., Z. Cheng, and X. Xu. 1999. The earliest ceratopsian from the Tuchengzi Formation of Liaoning, China. Journal of Vertebrate Paleontology 19: 681–691.

27 Unraveling a Radiation: A Review of the Diversity, Stratigraphic Distribution, Biogeography, and Evolution of Horned Dinosaurs (Ornithischia: Ceratopsidae) SCOTT D. SAMPSON AND MARK A. LOEWEN

recent discoveries have greatly expanded our knowl-

relatively low diversity at any one time, the clade appears

edge of the horned dinosaur clade Ceratopsidae. Taxo-

to have experienced relatively rapid turnover of species,

nomically, ceratopsids are currently represented by about

with species durations averaging considerably less than

32 species—including 15 species of centrosaurines and

1 million years. Biogeographically, ceratopsids appear

17 of chasmosaurines—an approximate doubling of

to have had surprisingly small species ranges, with the

known diversity in less than 5 years. Phylogenetically,

great majority of taxa known from single geologic for-

general agreement exists as to the basic structural ele-

mations. Although a portion of this endemicity may well

ments of the ceratopsid tree. Within Chasmosaurine,

be due to inadequate sampling, there is strong evidence

Chasmosaurus and Pentaceratops are regarded as basal

of north-south endemicity within the Western Interior

branches, Anchiceratops and Arrhinoceratops as inter-

Basin during the late Campanian. A corollary of this

mediate branches, and Triceratops and Torosaurus as

finding is that, despite their large-to-giant body sizes,

among the most nested members of the clade. Within

most ceratopsids likely did not engage in long-distance

Centrosaurinae, Albertaceratops and other forms with

migrations. Instead, the prevalent pattern of lati-

long supraorbital horncores are most basal, Centrosaurus

tudinally disjunct biogeographic distributions provides

and Styracosaurus are united in a distinct clade, and

strong support for the provincialism hypothesis pre-

Einiosaurus is allied with a series of more nested,

viously postulated for Late Cretaceous terrestrial floras

‘‘pachyrhinosaur’’-like forms bearing nasal and supra-

and faunas inhabiting the Western Interior Basin. Such

orbital bosses. Stratigraphically, most ceratopsid taxa can

diminutive species ranges are problematic for under-

now be placed with considerable resolution into strati-

standing the paleobiology of ceratopsids, suggestive

graphic context, with a temporal range of approximately

of low-to-intermediate metabolic rates or abundant

79–65 Ma. In contrast to the conclusions of some pre-

food supplies, or both. Given that virtually all cera-

vious studies, it appears that ceratopsid diversity within

topsid species are currently known from small geo-

single ecosystems at any one time was generally low, typ-

graphic ranges, that the current record includes several

ically consisting of a single species each of centrosaurine

temporal gaps, and that most taxa are known from a

and chasmosaurine. Taking into account the high overall

restricted region in the northern region of the Western

species diversity of ceratopsids (minimally 24 taxa dur-

Interior Basin, it is likely that many additional species

ing the final 8 million years of the Campanian) and the

await discovery.

405

Introduction Ceratopsid dinosaurs were a diverse group of Late Cretaceous, large-bodied (4–8 m long; 1–4 ton), quadrupedal herbivores that include some of the most remarkable vertebrates known. Their exceptionally derived skulls include edentulous, parrotlike beaks, robust dental batteries with unique shearing dentitions, hypertrophied narial regions, and a broad array of signature ornamentations—from nasal and supraorbital horns to expansive, elaborately adorned parietosquamosal frills (Figs. 27.1, 27.2). The longest of these skulls achieved lengths in excess of 3 m, the largest known for any terrestrial vertebrate (Colbert and Bump 1947; Lehman 1998). Ceratopsids are subdivided into two subclades (‘‘subfamilies’’): Centrosaurinae, typically with subcircular narial regions and relatively short, highly adorned frills (e.g., Styracosaurus albertensis); and Chasmosaurinae, typically with elongate narial regions and more elongate, less adorned frills (e.g., Triceratops horridus). Ceratopsidae are well suited for an analysis of vertebrate evolutionary radiation based on multiple features: (1) large and massive skulls that are relatively abundant in the fossil record (i.e., majority of taxa represented by mostly complete skull materials); (2) putative species-specific signaling structures preserved as bony outgrowths of the skull; (3) relatively brief temporal Late Cretaceous (Campanian and Maastrichtian) distribution (currently approximately14 million years); and (4) a limited geographic distribution restricted to the Western Interior Basin (WIB) of North America (Dodson et al. 2004). (The lone possible exception is Turanoceratops tardabilis, a poorly known taxon erected on fragmentary elements —including double-rooted teeth—from the Late Cretaceous of Uzbekistan [Nessov et al. 1989]). As a result, ceratopsids arguably provide greater potential than any other major clade within Dinosauria to investigate the tempo and mode of evolution. This unique status is heightened by abundant recent discoveries that greatly increase the material basis of the clade. Although the first ceratopsid dinosaur was described well over a century ago (Marsh 1891), the great diversity of the

FIGURE 27.1. Representative ceratopsid skulls within Centrosaurinae, depicted in left lateral and dorsal views. Skulls in both left and right columns are arranged in order of stratigraphic occurrence from (A) youngest to ( J) oldest. Included taxa as follows. (A) Wapiti new taxon A (Currie et al. 2008); (B) Pachyrhinosaurus canadensis; (C) Achelousaurus horneri; (D) Einiosaurus procurvicornis; (E) Two Medicine new taxon (McDonald and Horner this volume); (F) Styracosaurus albertensis; (G) Centrosaurus apertus; (H) Centrosaurus brinkmani; (I) Albertaceratops nesmoi; ( J) Wahweap new taxon (Kirkland and DeBlieux this volume). Based largely on firsthand observations and modified after Sampson et al. (1997), Ryan and Russell (2005), and Ryan (2007).

406 sampson & loewen

FIGURE 27.2.

Representative ceratopsid skulls within Chasmosaurinae, depicted in left lateral and dorsal views. Skulls in both left and right columns are arranged in order of stratigraphic occurrence, from (A) youngest to (L) oldest. Included taxa as follows. (A) Triceratops horridus; (B) Diceratops hatcheri; (C) Torosaurus latus; (D) Eotriceratops xerinsularis; (E) Arrhinoceratops brachyops; (F) Anchiceratops ornatus; (G) Agujaceratops mariscalensis; (H) Cerro del Pueblo new taxon (Loewen et al. this volume); (I) Pentaceratops sternbergi; ( J) Chasmosaurus irvinensis; (K) Chasmosaurus belli; (L) Chasmosaurus russelli. Based largely on firsthand observations and modified after Forster et al. (1993), Holmes et al. (2001), and Wu et al. (2007).

group has not been appreciated until very recently. The recent

of the large-scale patterns and processes associated with the

surge in ceratopsid discoveries documented in this volume is

ceratopsid radiation.

due in part to the fact that more paleontological crews than ever before are working in Late Cretaceous–aged deposits within the WIB of North America. Moreover, investigators are

Diversity and Taxonomy

exploring regions that have thus far received minimal at-

A recent review of ceratopsid dinosaurs (Dodson et al. 2004)

tention. In particular, many of the new-found specimens have

postulated 6 valid monospecific genera of centrosaurines

been recovered from the otherwise poorly known south-

(with 2 questionable additions) and 10 species of chasmo-

ern region of the WIB, providing an important new window

saurines arrayed within 7 genera, for a total of 16 species.

into the evolution of horned dinosaurs (Sampson et al. 2004;

Since that time, a number of key specimens have been un-

Loewen et al. this volume; Sullivan and Lucas this volume).

earthed and/or published for the first time (Figs. 27.1–27.4).

These new genera and species greatly expand our knowledge

The present tally of known ceratopsid taxa (including genera

of the evolution of ceratopsids. However, before a rigorous

and species that have yet to be described in detail) is 32 species

analysis of the ceratopsid radiation can be undertaken, the

—with 15 species of centrosaurines and 17 of chasmosaurines.

known sample of taxa and specimens must be placed into a

Remarkably, these additions, many of which are addressed in

high-resolution taxonomic, phylogenetic, stratigraphic, pa-

this symposium volume, amount to a doubling in the known

leoenvironmental, and geographic framework. Headway has

diversity of ceratopsids in less than 5 years!

been made in each of these areas, yet additional data col-

Within Centrosaurinae, the list of new taxa includes at least

lection and synthesis remains to be done. The primary pur-

five species, three of which are thought to represent new gen-

pose of this paper is to provide an updated overview of rele-

era (Ryan and Russell 2005; Ryan et al. 2006; Ryan 2007; Kirk-

vant work to date, in particular highlighting the importance

land and DeBlieux this volume; McDonald and Horner this

of recent finds. In addition, we integrate evidence emerg-

volume; Sampson unpublished data). One of these is Alberta-

ing from this range of data sources to speculate upon some

ceratops nesmoi from the Campanian Oldman Formation of

Unraveling a Radiation 407

southern Alberta and northern Montana (Ryan 2007). In con-

and interorbital region, has been recovered from the late

trast to all previously known centrosaurines, Albertaceratops

Campanian–early Maastrichtian Prince Creek Formation of

possesses elongate, rostrodorsally directed supraorbital horn-

Alaska (Fiorillo and Gangloff 2003). Finally, Ryan et al. (2006)

cores. Also from the Oldman Formation of Alberta (although

postulated that a centrosaurine specimen (TMP 2002.76.1)

from higher in section, near the top of the unit) is a new spe-

with nasal and supraorbital bosses—known from a skull and

cies of Centrosaurus, C. brinkmani (Ryan and Russell 2005);

postcranium recovered from the Lethbridge Coal Zone at top

C. brinkmani is notable for possessing relatively small, later-

of the Dinosaur Park Formation—probably represents yet an-

ally projecting supraorbital horncores and elaborate bony

other new species; however, Ryan et al. (this volume) indi-

growths on the caudal margin of the parietal. Farther south, in

cate that, without the typically diagnostic parietal, the speci-

Grand Staircase-Escalante National Monument (GSENM),

men forms an unresolved trichotomy with Achelousaurus and

southern Utah, another putative centrosaurine closely allied

Pachyrhinosaurus. If TMP 2002.76.1 can eventually be referred

with Albertaceratops has recently been unearthed (Kirkland

to Achelousaurus horneri this taxon would be one of the only

and DeBlieux this volume). Recovered from the early Campa-

Campanian ceratopsids to be documented from more than

nian Wahweap Formation and referred to here as ‘‘Wahweap

one formation. Interestingly, both A. horneri and the Dinosaur

new taxon,’’ this animal appears to be absolutely smaller than

Park Formation specimen occur immediately below marine

other ceratopsids and possesses a pair of curved and highly

shales of the Bearpaw Formation (Sampson 1995; Ryan and

elongate processes on the caudal margin of the parietal. To-

Evans 2005; Ryan et al. 2006); however, due to longitudinal

gether, Albertaceratops and Wahweap new taxon comprise the

differences and the progressive nature of the Bearpaw trans-

oldest and basalmost members of Centrosaurinae, elucidating

gression, the latter is estimated to be on the order of 500,000

the early evolution of the group. Indeed, Wahweap new taxon

years older than the A. horneri materials from the Two Medi-

is currently the oldest known ceratopsid, dating to the early

cine Formation (Fig. 27.3). The considerable increase in the

Campanian, about 79 Ma. A third centrosaurine taxon with

diversity of centrosaurines with boss-like skull roof ornamen-

elongate supraorbital horncores (Kaiparowits new taxon A)

tations is notable. Until recently, Pachyrhinosaurus was re-

was recently recovered from the Kaiparowits Formation of

garded as an oddity within an already bizarre group; today the

GSENM, southern Utah (Sampson unpublished data). In a re-

known diversity of pachyrhinosaur-like forms (i.e., with nasal

cent review of Styracosaurus, Ryan et al. (2007) concluded that

and supraorbital bosses) has grown to at least five taxa (with a

S. ovatus, known only from the Two Medicine Formation of

possible sixth taxon from the Dinosaur Park Formation), com-

Montana, represents a valid taxon distinct from S. albertensis.

prising a substantial portion of the total diversity of Cen-

McDonald and Horner (this volume) report on previously un-

trosaurinae (Fig. 27.3).

described materials of this taxon, and document additional

Within Chasmosaurinae, the list of new ceratopsid taxa in-

support for the validity of the species. In addition, however,

cludes seven new genera and species since 2007, several of

the latter authors provide evidence of a sister taxon relation-

which are described in this volume for the first time (Fig. 27.4;

ship with Einiosaurus rather than Styracosaurus, and establish a

Sampson and Loewen 2007; Sullivan and Lucas 2007; Wu

new genus to accommodate this species.

et al. 2007; Loewen et al. this volume; Ott and Larson this

In the remaining, most nested portion of the clade Centro-

volume; Ryan et al. this volume). Wu et al. (2007) described

saurinae are a growing number of Pachyrhinosaurus-like forms

Eotriceratops xerinsularis—a large animal with long supra-

from the latest Campanian and early Maastrichtian of Mon-

orbital horns and elongate, spindle-shaped epoccipitals—

tana, Alberta and Alaska, all of which replace nasal and su-

from the Maastrichtian Horseshoe Canyon Formation of Al-

praorbital horncores in adults with species-specific pachy-

berta, positing this taxon to be nested within a derived

ostotic ‘‘bosses’’ (Figs. 27.1, 27.3; Sampson 1995; Sampson et

subclade that includes Triceratops and Torosaurus. Much far-

al. 1997; Fiorillo and Gangloff 2003; Ryan et al. 2006; Fanti

ther south, the latest Campanian Cerro del Pueblo Formation

and Currie 2007; Currie et al. 2008; P. Currie pers. com. 2007).

of Coahuila, Mexico, has yielded a new chasmosaurine taxon

Two of these taxa are known from abundant remains pre-

bearing hyper-robust supraorbital horncores that is thought

served within monodominant bonebeds in the late Cam-

to be closely related to Anchiceratops (Loewen et al. this vol-

panian Wapiti Formation (Currie et al. 2008; Currie pers.

ume). Over the past few years, the late Campanian Kaiparo-

com.). Like Pachyrhinosaurus canadensis, the Wapiti Formation

wits Formation of southern Utah has yielded remains of two

taxa possess large boss-like structures atop the narial and or-

new, as yet undescribed chasmosaurines (in addition to the

bital regions; unlike P. canadensis, at least the older of the

centrosaurine noted above) that are currently under study by

two Wapiti taxa exhibits variable, often elaborate ornamen-

the authors. The first (Kaiparowits new taxon B) is very large

tations on the dorsum of the median parietal bar (Sampson et

(skull length approximately 2.5 m), with abbreviated, laterally

al. 1997; Currie et al. 2008). Another Pachyrhinosaurus-like

projecting supraorbital horncores and an elongate frill bear-

form, with a continuous boss extending between the nasals

ing a pronounced median embayment on the caudal parietal

408 sampson & loewen

FIGURE 27.3. Stratigraphic ranges and evolutionary relationships of centrosaurine ceratopsids. In instances where the phylogenetic relationships of a taxon have not been assessed through cladistic analysis, the taxon is placed conservatively within the tree, usually as part of an unresolved polytomy. Dark bars depict known stratigraphic ranges, with representative fossils recovered from the top and bottom of this range. White bars depict uncertain stratigraphic ranges (i.e., species age and duration poorly established). Stratigraphic data derived from Goodwin and Deino (1989), Eberth and Hamblin (1993), Rogers et al. (1993), Fassett and Steiner (1997), and Roberts et al. (2005). Phylogenetic data and stratigraphic occurrences derived from Wolfe and Kirkland (1998), Ryan and Russell (2005), Kirkland and DeBlieux (2006), Ryan and Evans (2005), Ryan (2007), and McDonald and Horner (this volume).

Unraveling a Radiation 409

FIGURE 27.4. Stratigraphic ranges and evolutionary relationships of chasmosaurine ceratopsids. In instances where the phylogenetic relationships of a taxon have not been assessed through cladistic analysis, the taxon is placed conservatively within the tree, usually as part of an unresolved polytomy. Dark bars depict known stratigraphic ranges, with representative fossils recovered from the top and bottom of this range. White bars depict uncertain stratigraphic ranges (i.e., species age and duration poorly established). Stratigraphic data derived from Goodwin and Deino (1989), Eberth and Hamblin (1993), Rogers et al. (1993), Fassett and Steiner (1997), and Roberts et al. (2005). Phylogenetic data derived Wolfe and Kirkland (1998), Holmes et al. (2001), and Wu et al. 2007; Loewen et al. (this volume); Ott and Larson (this volume); and Ryan et al. (this volume).

410 sampson & loewen

margin (Smith et al. 2004; Sampson and Loewen 2007). The

tops are nested in an intermediate position, and Triceratops

second (Kaiparowits new taxon C) is a chasmosaurine with an

and Torosaurus are among the most nested members of the

abbreviated, transversely broad parietosquamosal frill bearing

clade. Similarly, within Centrosaurinae, there is currently

10 elongate processes along the caudal margin. Ryan et al.

agreement that Albertaceratops and other forms with long su-

(this volume) describe a new, highly ornamented chasmo-

praorbital horncores are basal taxa; Centrosaurus and Styraco-

saurine from the Judith River Formation of Montana, and Ott

saurus are united in a distinct clade; and Einiosaurus forms

and Larson (this volume) erect a new small-bodied taxon from

a clade with Achelousaurus and all other pachyrhinosaur-like

the Hell Creek Formation of South Dakota. Sullivan and Lucas

forms. It is worth noting, however, that the most compre-

(this volume) document yet another new genus and species

hensive analysis to date of Ceratopsidae within-group rela-

bearing a distinctive squamosal from the Naashoibito mem-

tionships (Dodson et al. 2004) recovered a polytomy within

ber (lower Maastrichtian) of the Ojo Alamo Formation of New

Centrosaurinae consisting of Einiosaurus + (Achelousaurus +

Mexico. Finally, Lucas et al. (2006) argued for removal of Chas-

Pachyrhinosaurus) + (Styracosaurus + Centrosaurus). Figs. 27.3

mosaurus mariscalensis from the genus Chasmosaurus, and

and 27.4 summarize these findings, and place recently de-

erected a new genus, Agujaceratops, to accommodate this

scribed taxa conservatively, generally as part of unresolved

taxon.

polytomies. All previous investigations of historical relationships within

Phylogeny

Ceratopsidae have been hampered by two factors. First, character selection has been biased overwhelmingly toward the

The monophyly of Ceratopsidae has not been challenged

skull, and particularly the adorned skull roof (Dodson et al.

since the advent of modern phylogenetic (cladistic) analysis

2004). This bias is understandable given the relatively conser-

(Dodson and Currie 1990; Dodson et al. 2004). (It is worth

vative nature of the ceratopsid postcranium in contrast to the

noting, however, that a detailed analysis of historical rela-

highly variable skull (Dodson et al. 2004). Nevertheless, some

tionships with a large number of ceratopsids and basal neo-

studies (Adams 1988; Chinnery 2004) indicate the presence of

ceratopsians has yet to be conducted.) Similarly, all recent

relevant characters in the ceratopsid postcranium that have

phylogenetic analyses support monophyly of the subclades

not been incorporated into phylogenetic analyses. Second,

Centrosaurinae and Chasmosaurinae (e.g., Lehman 1990,

previous analyses have included only a small subset of the

1998, Chasmosaurinae; Dodson and Currie 1990, Dodson et

total number of species, typically focusing on resolving rela-

al. 2004, Chasmosaurinae and Centrosaurinae; Forster 1990,

tionships within either Chasmosaurinae or Centrosaurinae.

Chasmosaurinae; Godfrey and Holmes 1995, Chasmosau-

In the few instances where the scope has been broadened to

rinae; Sampson 1995, Centrosaurinae; Penkalski and Dodson

include all ceratopsids (Dodson and Currie 1990; Dodson et

1999, Centrosaurinae; Holmes et al. 2001, Chasmosaurinae;

al. 2004; Ryan 2007), analyses have included genera only. In

Ryan 2007, Centrosaurinae and Chasmosaurinae). However,

other words, to date there has not been a comprehensive,

not surprisingly, there has been some disagreement as to the

species-level phylogenetic analysis of ceratopsid relation-

placement of taxa within these clades. For example, Lehman’s

ships. Such an analysis, which is prerequisite to any meaning-

(1996) analysis of chasmosaurine relationships produced a

ful investigation of the horned dinosaur radiation, is currently

simple, pectinate arrangement of taxa branching off a single

in preparation (Forster et al. in prep.).

main stem. Conversely, other analyses of the group (e.g., Forster et al. 1993; Holmes et al. 2001) have recovered a separate

Stratigraphy

clade for Pentaceratops and Chasmosaurus, whereas Sampson et al. (2004) recovered strong support for Chasmosaurus as a dis-

Another pivotal component in the emerging ceratopsid story

tinct clade, with Pentaceratops more closely allied to other,

has been a major increase in stratigraphic resolution within

more nested members of Chasmosaurinae. Within Centrosau-

key formations, enabling inter-formational comparisons that

rinae, Penkalski and Dodson (1999) argued for the removal of

were previously impossible (Figs. 27.3–27.5). To cite a recent

Avaceratops from Centrosaurinae, positing this taxon as the

example, laser-fusion

sister to all other ceratopsids. Differing topologies may be due

horizons intercalated throughout the upper Campanian Kai-

in part to character selection, but taxonomic scope is another

parowits Formation, southern Utah, have yielded absolute age

likely factor, since most of the analyses cited above included

dates ranging from approximately 76.1 to 74.0 Ma (Roberts et

relatively few taxa.

al. 2005). These radiometric dates, along with previously pub-

Ar/ 39Ar analyses of four volcanic ash

40

Despite expected incongruencies, general agreement exists

lished dates for other formations (Goodwin and Deino 1989;

as to the overall structure of the ceratopsid tree (Figs. 27.3,

Rogers et al. 1993; Eberth and Hamblin 1993; Fassett and

27.4). Within Chasmosaurine, Chasmosaurus and Pentacera-

Steiner 1997), illustrate that the Kaiparowits Formation in

tops are regarded as basal taxa, Anchiceratops and Arrhinocera-

southern Utah is contemporaneous or penecontemporaneous

Unraveling a Radiation 411

FIGURE 27.5. Ceratopsid occurrences and chronostratigraphic relationships of key geologic formations used in this study, derived from data in Goodwin and Deino (1989), Eberth and Hamblin (1993), Rogers et al. (1993), Fassett and Steiner (1997), Wolfe and Kirkland (1998), Holmes et al. (2001), Roberts et al. (2005), Ryan and Russell (2005), Kirkland and DeBlieux (2006), Ryan (2007), and Wu et al. (2007). Indicated taxa include 1: Wahweap new taxon; 2: Kaiparowits new taxon C; 3: Albertaceratops nesmoi; 3?: Judith River new taxon; 4: c.f. Avaceratops lammersi; 5: Centrosaurus brinkmani; 6: Centrosaurus apertus; 7: Styracosaurus albertensis; 8: Two Medicine new taxon; 9: Einiosaurus procurvicornis; 10: Achelousaurus horneri; 11: Dinosaur Park new taxon(?); 12: Pachyrhinosaurus canadensis; 13: Wapiti new taxon A; 14: Wapiti new taxon B; 15: Prince Creek new taxon; 16: Kaiparowits new taxon A; 17: Chasmosaurus russelli; 18: Chasmosaurus belli; 19: Chasmosaurus irvinensis; 20: Kaiparowits new taxon B; 21: Agujaceratops mariscalensis; 22: Pentaceratops sternbergi; 23: Cerro del Pueblo new taxon; 24: Anchiceratops ornatus; 25: Arrhinoceratops brachyops; 26: Torosaurus latus and Torosaurus utahensis; 27: Ojo Alamo new taxon; 28: Eotriceratops xerinsularis; 29: Diceratops hatcheri; 30: Hell Creek new taxon; 31: Triceratops porosis; 32: Triceratops horridus.

with several dinosaur-rich formations to the north (Dinosaur

dates published for the Wahweap formation ( Jinnah et al.

Park Formation, Alberta; upper portions of the Judith River

2007), as well as the stratigraphic position of the holotype

and Two Medicine formations, Montana) and (to a lesser ex-

specimen (Kirkland and DeBlieux this volume), it is estimated

tent) to the east (Fruitland Formation and lower portion of the

that this taxon dates to about 79 Ma. The youngest known

Kirtland Formation, New Mexico). The ceratopsid-bearing

ceratopsid is Triceratops horridus, a widely distributed taxon

portion of the Aguja Formation of southwest Texas (Upper

that occurs virtually up to the Cretaceous-Paleogene bound-

Shale member, including the Kritosaurus zone of Lehman

ary. Of this 14 million year duration, by far the best sampled

1997) was previously regarded as penecontemporaneous with

interval is the final 5 million years of the Campanian (ap-

the above-named formations (Lehman 1997); however, recent

proximately 76–71 Ma). Of the 32 taxa noted in Figs. 27.3 and

study indicates that the Aguja Formation is significantly

27.4, the bulk of these (21 taxa, or 66%) are restricted to this 5

younger in age, with the most fossiliferous zone likely span-

million year interval. Particularly significant is the high reso-

ning the latest Campanian and early Maastrichtian (Atchley

lution stratigraphic placement of taxa within these late Cam-

et al. 2004; see Sullivan and Lucas 2003 for a contrary view).

panian geologic formations (Figs. 27.3–27.5). For example,

Together with radiometric dates for key formations have

Holmes et al. (2001) and Ryan and Evans (2005) report the

come critical new data pertaining to the stratigraphic ranges

successive occurrence of three species of the genus Chasmo-

of many ceratopsid species. These data are providing the foun-

saurus—first C. russelli, then C. belli, and finally C. irvinensis—

dation for assessing species durations, as well as for testing

within the late Campanian Dinosaur Park Formation. Ryan

hypotheses of coexistence and species turnover. Following a

and Evans (2005) document a coincident succession of cen-

review of the relevant literature, we conducted the first place-

trosaurines within the same unit—Centrosaurus apertus, Sty-

ment of all ceratopsid taxa into a comprehensive stratigraphic

racosaurus albertensis, and an undescribed pachyrhinosaur

framework (Figs. 27.3–27.5). A few caveats must be noted.

(Fig. 27.3). A parallel sequence of three centrosaurine taxa—

Absolute ages and durations of species are estimated on

‘‘Styracosaurus’’ ovatus, Einiosaurus procurvicornis, and Achelou-

the basis of relative position of these taxa, in most cases

saurus horneri—has also been documented in younger sedi-

within radiometrically dated units. However, these assess-

ments of the Two Medicine Formation (Horner et al. 1992;

ments should be regarded as first approximations, since they

Sampson 1995; McDonald and Horner this volume). Specifi-

are based in part on the assumption of constant sediment

cally, the stratigraphic sequence of ceratopsids within the Two

accumulation rates, an assumption did not apply in all cases

Medicine Formation occurs between about 74.9–74.0 Ma,

(Rogers 1998). Another limiting factor that strongly impacts

whereas the Dinosaur Park Formation taxa occur between

calculation of species durations—one that applies to macro-

about 76.4–74.8 Ma (Figs. 27.3–27.5). A caveat here is that the

vertebrates generally—is small sample sizes. Many taxa are

Two Medicine centrosaurines are known from very few lo-

known from one or a few specimens and from very narrow

calities, so any conclusions as to stratigraphic ranges must be

stratigraphic intervals. Nevertheless, some ceratopsid species

regarded cautiously.

are known from well over a dozen specimens and a substantial

Although there appears to be little or no temporal overlap

stratigraphic span. For example, Centrosaurus apertus is rep-

of ceratopsid faunas in the Two Medicine and Dinosaur

resented by more than 20 skulls and skeletons, as well as ma-

Park formations, overlap does occur between the Dinosaur

terials from 18 bonebeds, spanning well over 500,000 years

Park and Kaiparowits formations. Kaiparowits new taxon A, B,

(Ryan and Evans 2005; Ryan et al. 2007), and Triceratops hor-

and C are all known from the lower portion of the middle

ridus is known from several dozen skulls. In a growing number

member, which dates to about 75.3 Ma. Consequently, the

of instances, it is also possible to make an argument for faunal

new long-horned centrosaurine from Utah (Kaiparowits new

turnover based on the occurrence in close relatives in strati-

taxon A) appears to have been coeval with Centrosaurus apertus

graphically adjacent facies (see below). Particularly for well

in the north, whereas the two new chasmosaurines from Utah

sampled units like the Dinosaur Park Formation, it seems rea-

(Kaiparowits new taxon B and C) appear to have overlapped

sonable to establish minimum species durations for several

in time with Chasmosaurus belli (Figs. 27.3–27.5). Within

taxa (Figs. 27.3, 27.4). Therefore, although these findings

Chasmosaurinae, other examples of temporal overlap of taxa

should be regarded tentatively, they represent current best es-

include Torosaurus latus, Diceratops hatcheri, and Triceratops

timates of the absolute age and known temporal duration of

horridus in the late Maastrichtian, as well as Agujaceratops mari-

all ceratopsid taxa, thereby elucidating the overall pattern of

scalensis with either Anchiceratops ornatus or Arrhinoceratops

the ceratopsid radiation.

brachyops in the latest Campanian (Fig. 27.4). Remarkably, the

The earliest known ceratopsid is a purported centrosaurine

stratigraphic occurrences of Kaiparowits new taxon A with

(denoted herein as Wahweap new taxon) from the early Cam-

Centrosaurus apertus appear to represent the only known clearly

panian Wahweap Formation (Wahweap new taxon; Kirkland

documented example of temporal co-occurrence within Cen-

and DeBlieux this volume). Based on the first radiometric

trosaurinae (Fig. 27.3). Note that these two taxa do not over-

Unraveling a Radiation 413

lap geographically, with C. apertus thus far known only from

Paleoenvironment

Alberta and Kaiparowits new taxon A restricted to Utah. Ryan and Evans (2005) reported the possible overlap of Styraco-

Another key parameter of the ceratopsid radiation is paleo-

saurus albertensis and Centrosaurus apertus within the Dinosaur

environment. Species habitats and ranges are contingent

Park Formation, based upon a single fragmentary parietal

upon a variety of paleoenvironmental factors—for example,

spike recovered from the uppermost portion of the Centro-

climate, vegetation distributions, landmass size and physical

saurus stratigraphic zone. However, Ryan et al. (2007) re-

barriers. Moreover, evolution is often driven by environmen-

examined this outlier specimen and concluded that it most

tal perturbations that can produce a variety of biotic effects,

closely resembles the process 1 hook on the caudal parietal

including habitat tracking, extinction, and the formation of

margin of Centrosaurus apertus (see Sampson et al. 1997 for

population isolates (e.g., Eldredge 1999; Gould 2002). Thus,

a review of the numbering system of parietal marginal pro-

paleoenvironmental associations comprise a critical source of

cesses). The only other potential example is the co-occurrence

information for understanding the biology and radiation of

of Avaceratops lammersi with either Centrosaurus apertus or C.

any clade.

brinkmani. However, the absolute age of Avaceratops is in ques-

The history of ceratopsids is intimately linked to the Late

tion (Fig. 27.3), as is its taxonomic position within Centro-

Cretaceous paleogeography of North America, in particular

saurinae (Penkalski and Dodson 1999; Dodson et al. 2004).

the waxing and waning of the Cretaceous Western Interior

Thus, despite considerable diversity (about 13 taxa in 7 mil-

Seaway (KWIS; Fig. 27.6). Approximately 100 Ma, this epieric

lion years), there is currently no solid evidence of temporally

sea flooded the central portion of North America, forming

coincident centrosaurine species within the northern (or

semi-isolated eastern and western landmasses (Laramidia and

southern) portion of the WIB.

Appalachia, respectively; Scotese 2001). Bordering the KWIS

The longest surviving ceratopsid species reported in the lit-

to the west was the Sevier thrust belt. Over the next 35 mil-

erature is Pentaceratops sternbergi, thought to occur over a 2.5

lion years, the seaway underwent a series of transgressive-

million year period from about 75.0–72.5 Ma (Lucas et al.

regressive cycles, dramatically altering the size of the WIB and

2006). Triceratops horridus is a distant second in this category,

the nature of its paleoenvironments. Several fossiliferous for-

with an estimated duration of 1.5 million years (approxi-

mations deposited east of the Sevier orogenic belt and west of

mately 67.0–65.5 Ma). However, the purported lengthy strati-

the KWIS represent a variety of Late Cretaceous terrestrial

graphic ranges for these two taxa are in need of critical eval-

and nearshore marine paleoenvironments, including near-

uation, particularly since all other species have significantly

shore coastal plain, more distal alluvial plain, and (at least

shorter documented durations. As noted above, however, the

during the Maastrichtian), upland, intermontane basins. The

majority of ceratopsid taxa are known from a small number

last major marine incursion into the north of Laramidia was

([5) of specimens, and some taxa known from greater sample

the Bearpaw transgression, which entailed a 200-mile west-

sizes are limited to very restricted stratigraphic intervals (e.g.,

ward migration of the strandline over a period of approxi-

Einiosaurus procurvicornis). The best documented case of strati-

mately 3.5 million years during the late Campanian, with

graphic ranges and species turnover within Ceratopsidae per-

minor fluctuations (i.e., regressive episodes) occurring during

tains to Centrosaurus apertus and Styracosaurus albertensis, both

this interval (Gill and Cobban 1973; Rogers 1998). During the

of which are represented by numerous (]10) specimens re-

Maastrichtian, the KWIS retreated northeastward, ultimately

covered from a substantial stratigraphic interval within an in-

reestablishing subaerial connections between eastern and

tensively sampled unit (the Dinosaur Park Formation). The

western North America. A remnant of the KWIS, the Cannon-

documented durations of these two taxa are approximately

ball Sea, persisted over much of north-central North America

700,000 and 500,000 years, respectively, and both are brack-

through the latest Maastrichtian and well into the Paleogene

eted by closely related centrosaurines, increasing confidence

(Scotese 2001; Fig. 27.6A).

that the observed durations may be reasonable approxima-

The sister taxon to Ceratopsidae, Zuniceratops, dates to

tions of the actual lifespans of these species. Given these

about 90 Ma (Wolfe and Kirkland 1998), well after emplace-

data, supported by additional sequences of short-lived, non-

ment of the KWIS, and the earliest known ceratopsid occurs at

overlapping species chasmosaurines (Dinosaur Park Forma-

about 79 Ma (Kirkland and DeBlieux this volume). These find-

tion: Figs. 27.4, 27.5) and centrosaurines (Two Medicine For-

ings, together with the lack of ceratopsid remains from Ap-

mation; Figs. 27.3, 27.5) hint at a general pattern of rapid

palachia, indicate that this advanced clade of neoceratopsians

replacement that requires further testing. If this finding of

originated subsequent to isolation of Laramidia in the mid

relatively brief durations (i.e., substantially less than 1 million

Cretaceous. The last known ceratopsid, Triceratops, occurs in

years) is substantiated by further evidence, the 2.5 million

conjunction with the receding Cannonball Sea at the close of

year span of Pentaceratops sternbergi is particularly anomalous

the Cretaceous. With the exception of the latest Maastrichtian

and perhaps deserving of reconsideration.

forms (Triceratops, Torosaurus, and Diceratops), which occu-

414 sampson & loewen

pied the WIB after retreat of the seaway and reconnection of Laramidia and Appalachia, all ceratopsids were restricted to the narrow, fluctuating band of alluvial and coastal plain sediments in the WIB between the KWIS and the Sevier orogenic belt. Several claims have been made about the specific paleoenvironmental associations of ceratopsids. In the wake of the dinosaur renaissance of the 1970s and 80s, these large horned dinosaurs increasingly became viewed as rhinoceros- or antelope-like animals living in mixed-sex herds within dry, upland settings (Currie and Dodson 1984; Currie 1989). Lehman (1987) challenged this view, citing evidence that at least one taxon—the late Maastrichtian Triceratops—inhabited humid coastal lowlands. The coastal hypothesis was supported by several subsequent studies that addressed earlier, Campanianaged ceratopsids (Brinkman 1990; Eberth and Brinkman 1997; Brinkman et al. 1998). Campanian ceratopsids in the northern portion of the WIB occur predominantly in sediments deposited during transgression of the KWIS. The clastic wedge of dinosaur-bearing rocks deposited in the northern portion of the KWIS includes the Judith River, Dinosaur Park, and Two Medicine formations, all of which are capped by marine shales of the Bearpaw Formation. Thus, localities placed higher in section within this clastic unit represent increasingly more coastal settings. In a study of microvertebrate localities, Brinkman (1990) found that the frequency of ceratopsid remains increased upward through the Dinosaur Park Formation, suggesting that these horned dinosaurs were relatively more abundant in nearshore habitats. A subsequent study by Brinkman et al. (1998) documented further evidence of the same stratigraphic pattern. These authors also examined ceratopsid occurrences longitudinally, comparing chronostratigraphically equivalent horizons in southern Alberta at three levels on an east-west gradient. Based upon the results of earlier studies, they predicted that the frequency of ceratopsid remains would increase eastward, since more eastern sites would be progressively closer to the KWIS shoreline. The study’s results supported this prediction, with the abundance of ceratopsids from both macrovertebrate and microvertebrate localities increasing eastward (for a discussion of Albertan centrosaurine bone-

Late Cretaceous North American paleogeography. (A) Late Maastrichtian (approximately 66 Ma); (B) late Campanian (approximately 75 Ma); (C) early Campanian (approximately 85 Ma). Key ceratopsid-bearing formations (abbreviated) are noted for each interval as follows. A: Aguja; C: Cerro del Pueblo; D: Dinosaur Park; F: Kirtland and Fruitland; H: Hell Creek; I: Wapiti; J: Judith River; K: Kaiparowits; L: Oldman; M: Moreno Hill; N: North Horn; O: Ojo Alamo; P: Prince Creek; R: Horseshoe Canyon; S: Scollard; T: Two Medicine; W: Wahweap; V: Denver. Paleogeographic reconstructions after Blakey 2006. FIGURE 27.6.

Unraveling a Radiation 415

beds see Eberth et al. this volume). In a recent summary of the

can be tested through additional sampling of lower facies

paleoenvironmental associations of ceratopsids, Eberth (2007:

within the Two Medicine Formation, as well as through fur-

29) stated that most ceratopsid taxa ‘‘exhibit strong associa-

ther refinement of paleoenvironmental changes through this

tions with poorly drained alluvial to coastal plain facies . . . , but

and other formations.

a few seemed to have preferred better drained alluvial plain

Whereas dinosaur faunas of the Campanian WIB are limited

settings.’’ However, the coastal hypothesis is not without de-

to wet coastal plain and cooler alluvial plain paleoenviron-

tractors; in a relatively coarse analysis, Butler et al. (2007) made

ments, a third paleoenvironment occurs in the Maastrichtian.

use of a large database of dinosaur paleoenvironmental asso-

Following the onset of Laramide orogeny in the early Maas-

ciations to posit that ceratopsids preferred inland terrestrial

trichtian, upland, semi-arid intermontane basins appear to-

habitats.

gether with their own unique herbivorous dinosaur fauna

Our survey of the relevant literature, encompassing all cera-

that includes the titanosaur sauropod Alamosaurus and the

topsids described to date, found strong support for the coastal

chasmosaurine ceratopsid Torosaurus (Lehman 1997, 2001;

hypothesis. Just as Brinkman (1990) and Brinkman et al.

Sampson and Loewen 2005). Evidence for these intermontane

(1998) reported increasing numbers of southern Alberta cera-

basins is largely restricted to the southern region of the WIB,

topsid fossils in sediments approaching the Bearpaw trans-

from Utah to Texas. The occurrence of Torosaurus in upland,

gressive facies, so too it appears that ceratopsid taxa elsewhere

intermontane basins distal to the seaway margin may suggest

tend to occur predominantly in association with coastal set-

that this taxon was an ecological outlier, specializing in a pa-

tings (Fig. 27.5). More specifically, the bulk of ceratopsid taxa

leoenvironment that represents a major departure from the

occur in close stratigraphic proximity to marine facies asso-

coastal bias shown by the majority of ceratopsid taxa. Further

ciated with transgression-regression cycles (e.g., Agujaceratops

resolution of the paleoenvironmental associations of cera-

mariscalensis [Sankey et al. 2007; Sankey this volume]; Cerro

topsid species (and perhaps lineages) will be an important task

del Pueblo new taxon [Loewen et al. this volume]; Eotriceratops

for future workers attempting to reconstruct the horned dino-

xerinsularis [Wu et al. 2007]).

saur radiation.

A particularly significant geologic example is the Two Medicine Formation of northern Montana, one of the few Campanian-aged WIB formations described as preserving

Biogeography

more arid ‘‘upland’’ facies (Rogers 1998; Horner 1997). Ac-

Mesozoic biogeography has generally been limited to biotic

cordingly, Ryan et al. (2007) postulated that Styracosaurus al-

comparisons between and among continental landmasses.

bertensis, known from the Dinosaur Park Formation, was

However, the highly fossiliferous and well-sampled Upper Cre-

adapted to more coastal, mesic settings, whereas ‘‘S.’’ ovatus,

taceous formations of the Western Interior of North America

from the Two Medicine Formation, was specialized for more

offer a unique opportunity to examine finer-scale, subconti-

inland, xeric conditions. Most dinosaur remains have been

nental biogeographic patterns. Remarkably, the total com-

recovered from the upper portion of the formation (Rogers

bined area of these dinosaur-rich habitats during the Campa-

1998), from about 76–74 Ma. Yet ceratopsids are known only

nian encompassed approximately 4,000,000 km2, equivalent

from the uppermost facies between 75–74 Ma (Sampson 1995;

to about 16% of the present day area of North America (Fig.

McDonald and Horner this volume). Below this level, in the

27.6). Although Asia and western North America experienced

76–75 Ma interval, dinosaur remains are relatively abundant,

periodic land connections via a northern corridor (Russell

including the famous Egg Mountain fauna with such taxa as

1995), the lack of species common to both areas suggests that

Maiasaura, Orodromeus, and Troodon (Horner 1997). We postu-

this link acted largely as a sweepstakes filter, allowing limited

late that the lack of ceratopsid fossils from the fossiliferous

faunal exchange and effectively isolating the western part of

76–75 Ma interval represents a true absence, reflective of the

North America (Farlow et al. 1995; Godefroit et al. 2001). This

fact that ceratopsids entered Two Medicine ecosystems only

relative isolation on a peninsular landmass with an abundant

after the Bearpaw transgression sufficiently transformed local

fossil record makes the Late Cretaceous WIB an ideal focus for

habitats from inland to coastal. Returning to the Styracosaurus

studies of dinosaur ecology, evolution, and biogeography. In-

example, ‘‘S.’’ ovatus is known from somewhat younger sedi-

deed, with regard to dinosaurs, this time-space ‘‘slice’’ is likely

ments than S. albertensis (Fig. 27.3); therefore, since this was a

the best documented example for a major Mesozoic landmass.

period of rapid transgression, it is possible that both taxa were

It has been postulated that Late Cretaceous terrestrial floras

coastal or near-coastal specialists. Possible evidence against

and faunas inhabiting the WIB were separated into distinct

this hypothesis comes from the work of Rogers (1998), who

northern and southern biomes (Russell 1967; Lehman 1997,

argued on the basis of sedimentological data that the upper-

2001). This putative provincialism has been challenged, how-

most facies of the Two Medicine formation were deposited

ever, based on the argument that the key geologic formations

under relatively semi-arid conditions. The coastal hypothesis

were not deposited contemporaneously, but rather represent

416 sampson & loewen

temporally discrete intervals (Sullivan and Lucas 2003). The

availability and/or nesting season. The east-west migration

new radiometric dates noted in the stratigraphic discussion

hypothesis deserves further attention, and should be testable

above provide support for the hypothesis of latitudinally ar-

as the fossil record of the group becomes better resolved.

rayed faunas, demonstrating that the fossiliferous portions

Based on available evidence, it is unclear whether late Maas-

of several key late Campanian units (Dinosaur Park, Two Med-

trichtian ceratopsid taxa exhibited the apparently high levels

icine, and Judith River formations in the north; the Kaiparo-

of endemism that appear to characterize their late Campanian

wits and Fruitland formations in the south) overlap in time.

counterparts (Lehman 1987). Specifically, Torosaurus latus

Moreover, there is a growing body of evidence supporting the

has been documented in multiple formations in both the

provincialism hypothesis, augmented in particular by the re-

north (Hell Creek Formation of Montana, North Dakota,

covery of a new macrovertebrate fauna from the late Campa-

South Dakota; Frenchman Formation of Saskatchewan) and

nian of Utah (Sampson et al. 2004). Most remarkable of all is

south (North Horn Formation of Utah; McCrae Formation of

the fact that no species of Late Cretaceous dinosaur has been

New Mexico; Javelina Formation of Texas) of the WIB (Dod-

conclusively documented in both the north and south of the

son et al. 2004). Triceratops horridus also has a broad (though

WIB. Indeed, with key exceptions (e.g., Triceratops, Torosaurus;

not as extensive) distribution, occurring as far north as Alberta

see below), dinosaur species are typically restricted to single

and Saskatchewan, and as far south as Colorado and Wyo-

geologic formations.

ming (Dodson et al. 2004). Yet the species-level taxonomy of

This striking degree of endemism is inconsistent with pre-

both of these genera has been questioned. Whereas all north-

vious speculations that at least some dinosaurian megaherbi-

ern Torosaurus specimens have been placed into T. latus, the

vores—in particular ceratopsids and hadrosaurs—engaged in

more southern specimens have been regarded by some au-

annual long distances akin to those of modern day caribou and

thors as pertaining to a distinct species, T. utahensis (Lehman

wildebeest (Currie 1989). Rather, despite the possession of

1996). There has also been much discussion about species di-

large-to-giant body sizes, current evidence indicates that cera-

versity in Triceratops (Ostrom and Wellnhoffer 1986; Forster

topsids and hadrosaurs possessed relatively diminutive species

1996; Lehman 1990, 1998), with the bulk of specimens gener-

ranges grouped into at least two latitudinally arrayed regions

ally placed within T. horridus. However, there remains un-

(Lehman 1997, 2001; Sampson et al. 2004)—a northern region

certainty on this matter; preliminary studies by Happ and

that included Montana and Alberta, and a southern region

Morrow (1996) and Farke (1997) indicate the presence of geo-

that included Utah, New Mexico, Colorado, and Texas. More-

graphically separated species, and a study currently under way

over, the occurrence of disjunct geographic distributions

by Andrew Farke (pers. com. 2007) suggests that northern and

within well constrained, overlapping time intervals provides

southern specimens may be distinguishable morphometri-

strong support for the faunal provincialism hypothesis of Leh-

cally. If further analyses support the recognition of latitudi-

man (1987, 1997, 2001). The best documented evidence for

nally arrayed species within these genera, it would suggest

north-south regional faunas occurs in the 2-million-year late

that the dispersal barrier present for much of the Campanian

Campanian interval from 76 to 74 Ma. The provincialism hy-

may have been emplaced once again in the Maastrichtian.

pothesis can and should be tested further through a com-

Conversely, if it turns out that late Maastrichtian ceratopsids

bination of sources, including discoveries of new taxa and

had much more extensive species ranges, it might indicate

taxonomic reassessments (i.e., presence/absence of species),

that the external barriers separating northern and southern

phylogenetic analyses (i.e., presence/absence of geographi-

faunas during the late Campanian were removed, perhaps in

cally isolated subclades indicative of isolated endemic centers),

association with changing climates and retreat of the KWIS.

increasing stratigraphic resolution (i.e., presence/absence of

Alternatively, as predominantly coastal animals, ceratopsids

temporal overlap of taxa), and examination of unexplored

like Triceratops may simply have tracked their preferred hab-

formations.

itat southward as the seaway retreated during the Maastrich-

Brinkman et al. (1998) proposed an intriguing hypothesis regarding ceratopsid migration. They note that the abundant

tian. Either way, additional taxonomic work is needed to sort out the diversity of Maastrichtian ceratopsids.

monodominant ceratopsid bonebeds in the Dinosaur Park

The recent discoveries described above have resulted in a

Formation tend to occur in the middle of the unit, within facies

number of additional insights pertaining specifically to cera-

deposited relatively inland. This finding contrasts with the

topsid biogeography. For example, recognition of centro-

above-described pattern in which ceratopsid remains other-

saurine material from the Parras Basin of Mexico (Lund et al.

wise tend to be more frequent higher in section, within facies

2007; Loewen et al. this volume), together with reports of

deposited more proximal to the KWIS. The authors account for

Pachyrhinosaurus from Alaska (e.g., Fiorillo and Gangloff

this anomalous pattern by suggesting that ceratopsids may

2003), confirm that, contrary to several earlier statements,

have gathered together seasonally in large numbers to under-

this clade of short-frilled ceratopsids spanned the western

take east-west migrations, perhaps in association with food

North America landmass from Mexico to the Arctic Circle, a

Unraveling a Radiation 417

much larger range that has been documented for Chasmo-

amounts of high quality fodder than are generally available

saurinae. The discovery of new centrosaurine taxa from the

today; or (3) both (Farlow et al. 1995; Lehman 2007).

Campanian of Utah hints at an unrealized diversity of this short-frilled clade in the southern portion of the WIB. Moreover, Zuniceratops from the Turonian Moreno Hill Formation

Discussion

of New Mexico, and regarded as the neoceratopsian sister

BRIDGES AND GAPS

taxon to Ceratopsidae, provides strong evidence that elongate supraorbital horncores were present in the common ancestor

The ceratopsid fossil record is now relatively dense, with

of Ceratopsidae (Wolfe and Kirkland 1998; Wolfe 2000; Wolfe

about 32 species known from a single restricted landmass and

et al. 2007). The occurrence of similarly elongate horncores

a relatively brief interval (currently approximately14 million

in new centrosaurines from the north (Albertaceratops; Ryan

years). However, this record is far from uniform; rather it has

2007) and the south (Wahweap new taxon, Kirkland and De-

a number of strengths and weaknesses, here referred to as

Blieux this volume; Kaiparowits new taxon A) demonstrates

bridges and gaps, respectively. The new discoveries noted

that this feature persisted among some basal, early Campa-

above add important data points, expanding upon previously

nian members of the group, and that these long-horned forms

known bridges and erecting new bridges. For example, with

spanned a relatively broad latitudinal range. Among more de-

regard to the former, the new pachyrhinosaur-like centro-

rived centrosaurines, pachyrhinosaur-like forms with nasal

saurines emerging from the northern region of the WIB (Mon-

and supraorbital bosses are currently restricted to the north-

tana, Alberta, Alaska) greatly expand the existing data base,

ern portion of the WIB (Montana, Alberta, and Alaska).

transforming an animal previously regarded as a bizarre, odd-

The pairing of giant body sizes with such diminutive species

ball taxon (Pachyrhinosaurus canadensis) into a member of a

ranges is surprising. Studies of extant terrestrial vertebrates

highly successful, late persisting clade of centrosaurines. In

demonstrate that maximal body masses increase predictably

the realm of new bridges, Albertaceratops nesmoi (Ryan 2007)

with increasing land area, perhaps because larger masses re-

forces redefinition of our notion of Centrosaurinae to include

quire more expansive home ranges in order to ensure suffi-

elongate supraorbital horns, helping us to visualize basal cen-

cient food resources (Marquet and Taper 1998; Burness et al.

trosaurines and the likely transformation sequence that re-

2001; Kelt and Van Vuren 2001). In turn, for any given popula-

sulted in more derived members of the clade. The new cen-

tion size, larger home ranges ultimately entail lower popula-

trosaurine from the Wahweap Formation of Utah (Wahweap

tion densities. This evidence implies that maximal body sizes

new taxon; Kirkland and DeBlieux this volume), currently the

are constrained largely by the distribution and density of

oldest and most basal ceratopsid known, promises to expand

home ranges within a given area. Other factors being equal,

this notion further, perhaps even elucidating the transition

home ranges and species ranges are predicted to be greater

from basal neoceratopsians to ceratopsids.

among larger-bodied vertebrates, with the trend being more

Temporally, the greatest area of strength—or, to extend the

extreme among carnivores than herbivores, and more so

metaphor, the widest bridge with the most substantial sup-

among endotherms than ectotherms (Peters 1983; Farlow

ports—is the 2-million-year interval during the late Campa-

1993; Farlow et al. 1995; Van Valkenburgh and Janis 1993;

nian from approximately 76 to 74 Ma (Figs. 27.3, 27.4). Of the

Marquet and Taper 1998; Burness et al. 2001). Maximum body

32 currently recognized taxa, approximately one-half are

size in large-bodied carnivores thus reflects an evolution-

known from this period. If we expand this interval to encom-

ary balance between maintaining population densities low

pass the late Campanian more generally (approximately 76–

enough to avoid over-exploitation of prey species, yet large

70.5 Ma), the number of taxa increases to greater than 20

enough to reduce the probability of extinction (Farlow 1993;

(with one or two poorly dated taxa as question marks). In

Marquet and Taper 1998; Burness et al. 2001). Therefore, as-

other words, on the order of two-thirds of ceratopsid diversity

suming that the observation of latitudinal regional faunas on

comes from approximately one-third of their known strati-

late Campanian Laramidia is substantiated by further discov-

graphic duration. Notably, the late Campanian is by far the

eries, this finding raises problematic issues about ceratopsids

best sampled in terms of geography as well, with species span-

and the paleobiology of dinosaurs generally. A full discussion

ning Laramidia from Alaska to Mexico. Only for that same 76–

of this subject is beyond the scope of the present contribution.

74 Ma span can we currently document the presence of at

Nevertheless, assuming that future discoveries support the

least two distinct (northern and southern) lineages within

overall pattern of endemism, relatively few explanations can

both chasmosaurines and centrosaurines. Thus, by far our

account for the occurrence of roughly elephant-sized herbi-

best picture of the evolution of ceratopsid dinosaurs comes

vores occupying relatively small species ranges; alternatives

from the late Campanian (Figs. 27.3–27.6), and it is here that

include (1) metabolic rates substantially lower than those

attention should be focused when addressing questions relat-

of living endothermic mammals; (2) access to much greater

ing to the tempo and mode of evolution.

418 sampson & loewen

Geographically, the best known area is in the northern re-

many nagging questions remain. For example, was the diver-

gion of the WIB, in particular southern Alberta and north-

sity of ceratopsids decimated at the end of the Campanian,

ern Montana (Fig. 27.5). The great bulk of taxa from both

leaving only a handful of (mostly chasmosaurine) descen-

the Campanian and Maastrichtian are known only from this

dants? Or did the relatively rapid turnover of species that ap-

region. Indeed, almost half of the total known diversity of

pears to have characterized the late Campanian continue into

ceratopsids (15 taxa; Figs. 27.5, 27.6) is known from southern

early Maastrichtian times, culminating in the highly success-

Alberta, and most of these taxa are unknown elsewhere! Strati-

ful Triceratops?

graphic distributions are reasonably well established for the

Closely tied to above temporal gap is a geographic one. In

great majority of these species (Figs. 27.3, 27.4). Importantly,

contrast to the abundance of ceratopsid species recovered

the fossiliferous formations in the north also preserve a variety

from the northern WIB, and particularly the region surround-

of paleoenvironments—from nearshore coastal plain to more

ing the Alberta-Montana border, relatively little is known

distal alluvial plain settings (see above). Therefore, notwith-

of more southern forms. Certainly, recent finds from more

standing the many recent discoveries of ceratopsid taxa in the

southerly locales like Utah, New Mexico, and Coahuila (Mex-

southern portion of the WIB, the north still offers by far the

ico) hint at a previously unrealized array of southern cera-

highest resolution picture of ceratopsid evolution.

topsids that may ultimately rival horned dinosaur diversity in

The strengths of the ceratopsid fossil record are balanced by

the north. Yet much work remains to be done, and we are left

several key weaknesses. In particular, the known diversity of

with many unanswered questions. Did the southern ceratop-

the group is concentrated in the late Campanian of the north-

sids undergo the same degree of species turnover as their

ern WIB, resulting in several temporal gaps with little to no

counterparts in the north? Did the regional faunas that appar-

taxonomic representation. First, and perhaps most substan-

ently characterized late Campanian Laramidia persist into the

tial, is the paucity of pre–late Campanian ceratopsids (Figs.

Maastrichtian, or did environments become more homoge-

27.3–27.5). Prior to 76 Ma, our knowledge of ceratopsids is

nous, resulting in a decrease in standing diversity? In other

limited to Centrosaurus brinkmani (approximately 77 Ma; Ryan

words, did ceratopsids participate in latitudinally arrayed,

and Russell 2005), Albertaceratops nesmoi (approximately 78

temporally persistent centers of endemism? If so, how long

Ma; Ryan 2007), and a newly described Judith River Forma-

were these centers isolated, and what was the barrier to dis-

tion taxon (Ryan et al. this volume) in the north, and Wah-

persal? Answers to questions like these will be an important

weap new taxon in the south (approximately 79 Ma; Kirkland

component in understanding the ceratopsid radiation.

and DeBlieux this volume). Of these, only Wahweap new taxon likely has significant potential to illuminate the early history of ceratopsids. Prior to 79 Ma, we must go back about

SPECIES DIVERSITY AND TURNOVER

10 million years to find the sister taxon to Ceratopsidae, Zuni-

It has long been argued that the late Campanian ( Judithian)

ceratops christopheri, from the Turonian (approximately 89–90

was the acme of dinosaur evolution, at least in North America,

Ma) of New Mexico (Wolfe and Kirkland 1998). What hap-

characterized by a proliferation of contemporaneous species,

pened during the early evolution of ceratopsids from basal

both within and between nearby geologic formations (e.g.,

neoceratopsian ancestors? When did the first ceratopsids ap-

Horner 1984; Lehman 1997, 2001). For example, Lehman

pear? Was there an explosive early radiation of the clade prior

(2001) postulated the presence of a coastal Corthyosaurus-

to the Centrosaurinae-Chasmosaurinae split, or was the spe-

Centrosaurus association in southern Alberta (Dinosaur Park

cies diversity effectively subdivided into these two clades? An-

Formation) and a co-occurring, inland Maiasaura-Einiosaurus

swers to questions such as these are currently lacking, and will

fauna in northern Montana (Two Medicine Formation). Yet at

likely be found in rocks dating to 90–80 Ma.

least for ceratopsids, the revised stratigraphic record described

The other major stratigraphic gap is the Maastrichtian, the

above (Figs. 27.3, 27.4) does not support this conclusion.

final 5 million years of the Cretaceous, between about 70.5–

Within Centrosaurinae, Einiosaurus procurvicornis and Centro-

65.5 Ma. This was a period of dramatic environmental change,

saurus apertus do not overlap in time; rather they appear to be

including an overall cooling trend together with regression of

temporally separated by approximately 1 million years. Simi-

the KWIS, the latter reestablishing a subaerial connection be-

larly, there is no evidence of co-occurring late Campanian

tween Laramidia and Appalachia (Erickson 1978; Lillegraven

chasmosaurine taxa from approximately the same latitude. In

and Ostresh 1990). There are a few ceratopsid data points for

short, current evidence is consistent with a single species each of

this time interval, including the relatively well known latest

centrosaurine and chasmosaurine in the northern and southern re-

Maastrichtian Triceratops horridus and Torosaurus latus, and a

gions of the late Campanian WIB at any one time. Possible excep-

pair of latest Campanian to early Maastrichtian forms known

tions pertain to Diceratops hatcheri and a small bodied chas-

from single specimens—Arrhinoceratops brachyops and Eotri-

mosaurine living alongside Triceratops horridus in the late

ceratops xerinsularis (Wu et al. 2007; see above). However,

Maastrichtian ‘‘Hell Creek’’ ecosystem of the north (Dodson et

Unraveling a Radiation 419

al. 2004; Ott and Larson this volume), and Kaiparowits new

for the large-bodied adults; thus the addition of elaborate skull

species B and C in the late Campanian Kaiparowits ecosystem

structures like horns and frills would not have negatively af-

in the south. Of these exceptions, D. hatcheri is based on a

fected fitness except insofar as these structures added exces-

single specimen and its validity as a distinct genus and species

sive mass to the skull. Second, given that adductor muscula-

has been questioned. The stratigraphic occurrences of the new

ture likely did not extend onto the parietosquamosal frill

Kaiparowits ceratopsids documented in Figs. 27.3–27.5 are

(Dodson et al. 2004), there is little reason to suspect that horn

tentative and currently under study. This pattern of strati-

and frill variants negatively impacted upon skull function

graphic and geographic distribution may be based on an in-

(e.g., feeding, jaw mechanics). Finally, as noted above, evi-

complete fossil record. That is, if we had more specimens from

dence from a variety of vertebrate and invertebrate clades in-

overlapping time periods in different formations, the conclu-

dicates that species diversity can be driven in large part by

sions might be different, perhaps supporting the notion of

sexual selection acting on signaling characters subject to mate

‘‘inland’’ and ‘‘coastal’’ faunas with distinct ceratopsid species.

competition (e.g., Albertson et al. 1999; Panhuis et al. 2001;

Nevertheless, the current record is striking, with ceratopsids

Streelman and Danley 2003; and references therein).

largely confined to more coastal settings, and little to no evi-

Considered in unison, these conclusions are consistent with

dence of multiple, coexisting centrosaurines or chasmosau-

several speculations about the ceratopsid radiation. We sug-

rines in the north or south of the WIB.

gest that chasmosaurines and centrosaurines established dis-

The finding of relatively low species diversity at any one

tinct and stable ecological niches early in their evolution, and

time, combined with the high overall species diversity of cera-

further that these niches persisted relatively unchanged over

topsids (minimally 25 taxa during the final 8 million years of

millions of years. This postulate is inferred from the low de-

the Campanian) suggests that the clade experienced relatively

gree of evolutionary change in fitness-related characters.

rapid turnover, with average species durations considerably

However, against this backdrop of apparent ecological sta-

less than 1 million years (Figs. 27.3, 27.4; see Stratigraphy

bility, the clade underwent a dramatic radiation driven by en-

above). As noted above, although current sample sizes pro-

vironmental fluctuations on the one hand (causing the forma-

hibit rigorous quantitative assessments of species durations,

tion of population isolates) and by the evolution of mating

the occurrence of multiple stratigraphic sequences of non-

signals on the other (largely driven by forces unrelated to fit-

overlapping taxa within both Chasmosaurinae and Centro-

ness, such as sexual selection). Following the formation of

saurinae is notable, and increases confidence that the ob-

isolated populations, horn and frill morphologies are pre-

served temporal spans reflect that actual pattern of turnover

dicted to have diverged as the genetic basis for female prefer-

(Ryan et al. 2007). In short, current evidence from the late

ences evolved independently. Once a sufficient degree of dif-

Campanian suggests that Ceratopsidae was characterized by

ferentiation had accumulated, interbreeding would have been

relatively rapid species turnover and low diversity during any

limited or prohibited even if environmental changes subse-

given interval. In stark contrast to previous estimates, it is

quently reunited the two populations. Due to a lack of dis-

even conceivable that only one lineage each of centrosaurines

parity in ecological traits, the newly formed daughter species

and chasmosaurines existed for much of the latest Cretaceous.

generally would have been unable to coexist with the parent

The paucity of evolution in fitness-related traits (i.e., those

species. Thus it is likely that one variant would quickly come

associated with survival rather than reproduction) within cer-

to dominate, typically resulting in the extinction of the other.

atopsids may help explain the relatively low diversity of cera-

A corollary of this hypothesis is that the apparently short

topsids at any given time. That is, without such evolution,

species durations exhibited by late Campanian ceratopsids

there would have been few opportunities to partition the hab-

(and likely hadrosaurs as well) may have been due to a

itat such that sister taxa could coexist. Some of the features

concatenation of several factors: (1) cyclical environmental

distinguishing chasmosaurines from centrosaurines (e.g., rel-

change—in particular, transgression-regression cycles of the

ative depth of the preorbital facial skeleton, orientation of the

KWIS—subdividing large populations into smaller subpopula-

tritural surface on the predentary) may have evolved via char-

tions for extended intervals (Horner et al. 1992); (2) high lev-

acter displacement early in the evolution of the clade as a

els of mate competition within gregarious, mixed sex groups,

means of reducing trophic overlap (Henderson 2007). Con-

with the potential to cause rapid divergence in mating signals;

versely, the high degree of variation in signaling traits, pri-

(3) the presence of elaborate mating signals not closely tied to

marily horns and frills, both within and between species is

fitness-based traits; and (4) a lack of evolutionary change in

consistent with reduced genetic constraints on these struc-

fitness-based characteristics, thereby prohibiting the estab-

tures. The latter conclusion seems warranted on at least two

lishment of independent niches for closely related species. In

grounds. First, if most ceratopsid species were gregarious and

other words, a high rate of appearance of new species coupled

lived in open environments (Sampson 2001), crypsis would

with a lack of ecological space to maintain them may have

have been an unlikely strategy for avoiding predation, at least

translated into a pattern of relatively rapid replacement. This

420 sampson & loewen

hypothesis may account for the substantially longer durations

lection models. Viewed from the perspective of this multi-

of species within other, coeval dinosaur clades (e.g., the an-

stage hypothesis, it is intriguing to consider the possibility

kylosaurid Euoplocephalus tutus within the Dinosaur Park For-

that psittacosaurs and basal neoceratopsians represent di-

mation; Ryan and Evans 2005), which may have lacked one or

vergence along the axes of habitat and trophic morphology

more of these factors (i.e., high levels of mate competition).

(fitness-based traits), whereas ceratopsids comprise the third,

This admittedly highly speculative scenario, which shares

communication-based stage, in which sexual selection super-

several elements with that put forth independently by Mc-

sedes natural selection as the primary driver of divergence.

Donald and Horner (this volume), predicts a pattern of rapid

Streelman and Danley (2003: 126) suggest that ‘‘a paucity of

replacement in which two taxa from a single lineage overlap

genetic variation for means of signaling that are independent

for only a brief interval in ecological time, with little chance

of fitness traits might be the most important constraint limit-

that these taxonomic co-occurrences would be recorded in the

ing the diversification of vertebrate groups.’’ As discussed

geologic record. Stratigraphically, this pattern of rapid turn-

above, ceratopsids exhibited dramatic variation in signaling

over within single lineages would appear very much like ana-

structures (primarily horns and frills) that appear to have

genesis, with one form transforming into another without

had few fitness-related constraints. We tentatively propose,

lineage branching (Horner et al. 1992). Yet the scenario out-

then, that the ceratopsid radiation embodies the third stage

lined above is cladogenetic, with lineage branching immedi-

of Streelman and Danley (2003), in which diversification oc-

ately followed by an extinction event (Sampson 1997). In con-

curred principally along the communication axis.

trast to a strictly anagenetic hypothesis, the latter predicts

Finally, there remains a major outstanding question as to

no directional evolution of signaling structures within spe-

the pattern of species turnover of ceratopsids and other clades

cies. In other words, species are predicted to exhibit relative

of Late Cretaceous dinosaurs in the WIB. Specifically, did spe-

stasis (although potentially with significant variation about

cies turnover tend to occur sporadically and randomly or

a mean) between origin and extinction. If future studies with

in coordinated cross-taxic pulses? In other words, did the

increased sample sizes reveal step-wise directional evolu-

Campanian vertebrate faunas within the WIB undergo ran-

tion in horns and frills within ceratopsid species (perhaps

dom evolutionary change, with turnover in one lineage unre-

through morphometric analysis), this finding would con-

lated to others? Or was turnover concentrated in cross-taxic

stitute evidence against the above hypothesis. Currently,

pulses, as has been argued for Miocene-Recent mammals

however, the observed patterns are fully consistent with the

(Vrba 1987) and various clades of marine invertebrates (Brett

low diversity, rapid replacement model of ceratopsid turnover

and Baird 1995; see also discussions in Eldredge 1999 and

put forth here.

Gould 2002)? Of course, documentation of faunal turnover

Lastly, it is important to keep in mind that ceratopsids are

requires a relatively dense fossil record placed into context

members of a much larger, more widely distributed radiation

with a resolved stratigraphic framework. As noted above, the

of ceratopsian dinosaurs. In contrast to ceratopsids, psitta-

best examples from Late Cretaceous terrestrial formations

cosaurs and basal neoceratopsians exhibit broad taxonomic

within the WIB have thus far been documented in the north-

variation in fitness-related features, from locomotory style (bi-

ern region, with the Two Medicine Formation (Montana) and

peds versus quadrupeds) to jaw function (e.g., broad, planar

Dinosaur Park Formation (Alberta) being the most prominent

cutting surfaces on tooth crowns versus complex crowns that

examples (Horner et al. 1992; Currie and Russell 2005; Ryan

combine shearing and crushing; You and Dodson 2004).

and Evans 2005). Ryan and Evans (2005) described three ‘‘fau-

Whereas at least a portion of this variation is related to broad

nal zones’’ for the late Campanian Dinosaur Park Formation,

scale diversity within non-ceratopsid members of Ceratopsia,

based largely on distinct suites of hadrosaurids and ceratopsid

the distribution of phylogenetic characters across the skeleton

dinosaurs. As noted above, one important consequence of

in more basal (non-ceratopsid) clades suggests evolution of

stratigraphic zonation of taxa is a considerable reduction in

fitness-based traits within (as well as between) these clades.

perceived dinosaur diversity during any one interval within

Streelman and Danley (2003) argue on the basis of phylogene-

the late Campanian.

tic and population genetic evidence that many vertebrate ra-

Although still at the provisional stage, current evidence

diations exhibit evidence of passing through multiple stages.

is suggestive of cross-taxic pulses of faunal turnover, perhaps

Specifically, they posit that groups tend to diverge in a three-

causally linked to shifts in environment (Horner et al. 1992;

stage sequence reflecting a trio of axes—habitat, trophic mor-

Currie and Russell 2005). Horner et al. (1992) implicated

phology, and communication—with evolution often follow-

transgression of the KWIS as an external forcing factor that

ing this order. Whereas divergence along the first two axes,

precipitated rapid, simultaneous turnover across multiple

habitat and trophic morphology, is thought to proceed ac-

clades, an intriguing hypothesis that is consistent with avail-

cording to ecological speciation models, divergence along the

able evidence. Achelousaurus horneri, apparently a member

third axis, communication, is postulated to follow sexual se-

of the Two Medicine lineage (Sampson 1995; McDonald and

Unraveling a Radiation 421

Horner this volume), may occur in uppermost sediments (ap-

evolution of ceratopsids—only that it likely had minimal ef-

proximately 74.3 Ma) of both the Dinosaur Park Formation

fect on the niches occupied by members of the two subclades.

and as well as the Two Medicine Formation, just prior to the

Clade diversification, then, appears to have been driven

Bearpaw transgression maximum. Achelousaurus is the ear-

largely by modifications to signaling structures. Lacking sig-

liest known pachyrhinosaur-like form with nasal and supra-

nificant modifications to fitness-based traits during at least the

cranial bosses. However, all subsequent centrosaurines known

late Campanian and early Maastrichtian, ceratopsid diversity

to date from the northern region (i.e., between 74 and 70 Ma)

at any one moment in time may generally have consisted of a

are also pachyrhinosaur-like forms closely allied to Achelou-

single representative each of Chasmosaurinae and Centro-

saurus. One interpretation of this pattern is that the Bear-

saurinae in the northern and southern regions of the WIB.

paw transgression created a habitat bottleneck (sensu Horner

The apparent low diversity of coexisting ceratopsid taxa

et al. 1992) that resulted in extinction of the Centrosaurus-

within single ecosystems (or even regions) may have impor-

Styracosaurus lineage and persistence of the Achelousaurus-

tant implications for our understanding of the diversity of

Pachyrhinosaurus lineage.

Late Cretaceous dinosaurs generally. It has often been argued

Current evidence, particularly from the Dinosaur Park For-

that North American dinosaur diversity peaked during the

mation in Alberta, provides tantalizing clues suggestive of a

late Campanian and then underwent a steady decrease, cul-

turnover pulse pattern. If such speculation is borne out by

minating in relatively low species diversity during the termi-

further data collection from this and other formations, the

nal Cretaceous. The putative late Campanian peak has been

next important question will be whether or not these pulses

founded primarily on the anomalously high diversity of dino-

were coincident across formations. In particular, how did

saurs preserved in the Dinosaur Park Formation. However,

the tempo and mode of faunal turnover differ between the

whereas it is true that some dinosaur lineages present in the

southern and northern regions of the WIB? If transgressive-

latest Campanian are absent in the latest Maastrichtian (e.g.,

regressive cycles were critical factors driving faunal turnover,

lambeosaurine hadrosaurids, centrosaurine ceratopsids, alber-

as argued by Horner et al. (1992) and tentatively supported

tosaurine tyrannosaurids), recent assessments indicate that

here, it is unlikely that these turnover events occurred pene-

the Hell Creek dinosaur assemblage may have been signifi-

contemporaneously in the north and south, since current evi-

cantly more diverse than previously thought (Russell and

dence indicates that Late Cretaceous transgressive-regressive

Manabe 2002). More importantly for this discussion, recent

cycles were not in phase throughout the entire WIB (Lille-

evidence from the Dinosaur Park Formation indicates a suc-

graven and Ostresh 1990). Nevertheless, it may still be pos-

cession of dinosaur faunas, with standing diversity being

sible to link turnover pulses to major transgressive episodes of

much lower previously thought (Currie and Russell 2005;

the KWIS, particularly if these pulses do coincide with seaway

Ryan and Evans 2005). Thus, rather than being representative

migrations that were temporally staggered in the northern

of a single hyper-diverse ecosystem, the unusually high num-

and southern regions of the basin. Certainly answering ques-

bers of dinosaur species documented for the Dinosaur Park

tions such as these will be an important first step in unraveling

Formation are more likely reflective of relatively high rates of

the tempo and mode of evolutionary change in Late Creta-

faunal turnover, brief species durations, and thus higher num-

ceous terrestrial ecosystems within the WIB.

bers of species within a given temporal interval. Nevertheless, even if faunal turnover is accounted for and

Conclusion

intracladal diversity appropriately reduced, dinosaur diversity at any one time within late Campanian ecosystems was still

The recent discoveries discussed above greatly elucidate the

remarkably high (Lehman 1997). Considering only megaher-

evolution of ceratopsid dinosaurs. The two earliest known cer-

bivores, standing diversity typically included a single rep-

atopsids, Albertaceratops nesmoi (Ryan 2007) and Wahweap

resentative each of centrosaurine and chasmosaurine cera-

new taxon (Kirkland and DeBlieux this volume), provide di-

topsids, lambeosaurine and hadrosaurine hadrosaurids, and

rect evidence that the ceratopsid baüplan—including hyper-

nodosaurid and ankylosaurid ankylosaurs (as well as typically

trophied narial region, dental batteries, and an enlarged, orna-

smaller-bodied pachycephalosaurs and hypsilophodonts). If

mented parietosquamosal frill—was firmly in place by 77–79

we postulate that the northern and southern regions of the

Ma. With the exception of increases in body size, the subse-

WIB possessed approximately equal diversity levels during

quent 14 million years and greater than two dozen speciation

any particular interval—a contention supported by the re-

events (likely many more) do not appear to have resulted in

cently discovered Kaiparowits Formation fauna from south-

accrual of any substantial fitness-based evolutionary changes

ern Utah (Sampson et al. 2004, in press)—our rough estimate

in either chasmosaurines or centrosaurines. By this we do not

results in a minimum of 12 distinct species of rhinoceros-to-

mean to argue that natural selection had no impact on the

elephant-sized herbivores occupying a land area a fraction the

422 sampson & loewen

size of present day North America (Lehman 2001). If this conclusion even approximates actual diversity levels in the late

References Cited

Campanian WIB, it suggests that Late Cretaceous ecosystems

Adams, D. A. 1998. Structure and function of the ceratopsian forelimb. Ph.D. diss., University of California, Berkeley. Albertson, R. C., J. A. Markert, P. D. Danley, and T. D. Kocher. 1999. Phylogeny of a rapidly evolving clade: The cichlid fishes of Lake Malawi, East Africa. Proceedings of the National Academy of Sciences 96: 5107–5110. Atchley, S. C., L. C. Nordt, and S. I. Dworkin. 2004. Eustatic control on alluvial sequence stratigraphy: A possible example from the Cretaceous-Tertiary transition of the Tornillo Basin, Big Bend National Park, West Texas, U.S.A. Journal of Sedimentary Research 74: 391–404. Blakey, R. C. 2006. Paleogeography and Geologic Evolution of North America. Found on line at http:// www2.nau.edu/rcb7/nam.html. Brett, C. E., and G. C. Baird. 1995. Coordinated stasis and evolutionary ecology of Silurian to Middle Devonian faunas in the Appalachian Basin. In D. H. Erwin and R. L. Anstey, eds., New Approaches to Speciation in the Fossil Record, pp. 285–315. New York: Columbia University Press. Brinkman, D. B. 1990. Paleoecology of the Judith River Formation (Campanian) of Dinosaur Provincial Park, Alberta, Canada: Evidence from vertebrate microfossil localities. Palaeogeography, Palaeoclimatology, Palaeoecology 78: 37–54. Brinkman, D. B., M. J. Ryan, and D. A. Eberth. 1998. The paleogeographic and stratigraphic distribution of ceratopsids (Ornithischia) in the Upper Judith River Group of Western Canada. Palaios 13: 160–169. Burness, G. P., J. Diamond, and T. Flannery. 2001. Dinosaurs, dragons, and dwarves: The evolution of maximal body size. Proceedings of the National Academy of Sciences 98: 14518– 14523. Butler, R., P. Barrett, P. Kenrick, and M. Penn. 2007. Paleoenvironmental controls on the distribution of Cretaceous herbivorous dinosaurs. Journal of Vertebrate Paleontology 27(3, Suppl.): 54A. Chinnery, B. 2004. Morphometric analysis of evolutionary trends in the ceratopsian postcranial skeleton. Journal of Vertebrate Paleontology 24: 591–609. Colbert, E. H., and J. D. Bump. 1947. A skull of Torosaurus from South Dakota and a revision of the genus. Proceedings of the National Academy of Sciences 99: 93–106. Currie, P. J. 1989. Long-distance dinosaurs. Natural History 6: 60– 65. Currie, P. J., and P. Dodson. 1984. Mass death of a herd of ceratopsian dinosaurs. In W. E. Reif and F. Westphal, eds., Third Symposium of Mesozoic Terrestrial Ecosystems, pp. 52–60. Tubingen: Attempto Verlag. Currie, P. J., W. Langston, Jr., and D. H. Tanke. 2008. A new species of Pachyrhinosaurus (Dinosauria, Ceratopsidae) from the Upper Cretaceous of Alberta, Canada. In P. J. Currie, W. Langston, Jr., and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. 1–108. Ottawa: NRC Research Press. Currie, P. J., and D. A. Russell. 2005. The geographic and strati-

differed dramatically from those of the present day, including either greater food supplies or substantially lower metabolic rates among megaherbivores, or (perhaps most likely) both (Farlow et al. 1995; Lehman 1997, 2001). Despite the abundant new evidence now available for ceratopsids, another inescapable conclusion arising from this discussion is that our understanding of the ceratopsid radiation remains nascent. This statement can be made in spite of the fact that ceratopsids represent arguably the best documented clade of dinosaurs. Most importantly, given the sheer number of newly identified taxa, as well as temporal gaps in the fossil record, it appears that we have only begun to sample the diversity of horned dinosaurs in North America. This conclusion is underlined by the many new finds of ceratopsids in the southern portion of the WIB, by the complete lack of temporal overlap between southern and northern genera and species, and by the geographic concentrations of currently known species. It is likely, then, that the diversity of ceratopsids will increase greatly once less explored geographic regions and temporal intervals are subject to greater sampling. Major outstanding questions pertain to (1) the origin and early diversification of ceratopsids from basal neoceratopsian ancestors; (2) the diversity of ceratopsids during the Maastrichtian; (3) the number and geographic extent of regional faunas in the WIB during the Late Cretaceous; and (4) the pattern of species turnover (e.g., random versus cross-taxic pulses). Although some of these questions (1 and 2) can only be answered with the discovery of additional specimens, others (3 and 4) can be addressed at least in part by comprehensive studies that increase taxonomic, phylogenetic, stratigraphic, and paleoenvironmental resolution. Without doubt, thanks to an abundant fossil record, ceratopsid dinosaurs will be one of the key clades in efforts to understand the evolution of terrestrial faunas in the Late Cretaceous WIB, and indeed the paleobiology of dinosaurs generally. Acknowledgments

For discussions about ceratopsids generally, we sincerely thank Andy Farke, Cathy Forster, and Mike Getty. For providing thoughtful reviews of this manuscript, we thank Andy Farke, Eric Roberts, and Michael Ryan. For organizing a highly productive meeting, our thanks go to the co-conveners of the Royal Tyrrell Museum Ceratopsian Symposium—Don Brinkman, Michael Ryan, Brenda Chinnery-Allgeier, Dave Eberth, and Philip Currie. Finally, for their efforts in producing an important contribution to the dinosaur literature, we thank the editors of this volume—Michael Ryan, Brenda ChinneryAllgeier, and Dave Eberth.

Unraveling a Radiation 423

graphic distribution of articulated and associated dinosaur remains. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 537–569. Bloomington: Indiana University Press. Dodson, P., and P. J. Currie. 1990. Neoceratopsia. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 593–618. Berkeley: University of California Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A. 2007. Ceratopsians: A review of paleoenvironments and taphonomy. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 28–32. Drumheller: Royal Tyrrell Museum of Palaeontology. Eberth, D. A., and D. B. Brinkman. 1997. Paleoecology of an estuarine paleochannel complex in the Dinosaur Park Formation ( Judith River Group, Upper Cretaceous) of southern Alberta, Canada. Palaios 12: 43–58. Eberth, D. A., D. B. Brinkman, and V. Barkas. 2010. A centrosaurine mega-bonebed from the Upper Cretaceous of southern Alberta: Implications for behavior and death events. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 495–508. Bloomington: Indiana University Press. Eberth, D. A., and A. P. Hamblin. 1993. Tectonic, stratigraphic, and sedimentologic significance of a regional discontinuity in the upper Judith River Formation (Belly River Wedge) of southern Alberta, Saskatchewan, and northern Montana. Canadian Journal of Earth Sciences 30: 174–200. Eldredge, N. 1999. The Pattern of Evolution. New York: W. H. Freeman. Erickson, J. M. 1978. Bivalve mollusk range extensions in the Fox Hills Formation (Maestrichtian) of North and South Dakota and their implications for the Late Cretaceous geologic history of the Williston Basin. North Dakota Academy of Science Annual Proceedings 32: 79–89. Fanti, F. and P. J. Currie. 2007. A new Pachyrhinosaurus bonebed from the late Cretaceous Wapiti Formation. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 39–43. Drumheller: Royal Tyrrell Museum of Palaeontology. Farke, A. A. 1997. The distribution and taxonomy of Triceratops. In D. L. Wolberg, E. Stump, and G. Rosenberg, eds., Dinofest International: Proceedings of a Symposium Held at Arizona State University, pp. 47–49. Philadelphia: Academy of Natural Sciences. Farlow, J. O. 1993. On the rareness of big, fierce animals: Speculations about the body sizes, population densities, and geographic ranges of predatory mammals and large carnivorous dinosaurs. American Journal of Science 293–A: 167–199. Farlow, J. O., P. Dodson, and A. Chinsamy. 1995. Dinosaur biology. Annual Review of Ecology and Systematics 26: 445–471. Fassett, J. E., and M. B. Steiner. 1997. Precise age of C33n–C32r magnetic polarity reversal, San Juan Basin, New Mexico and

424 sampson & loewen

Colorado. In O. Anderson, B. Kues, and S. G. Lucas, eds., Mesozoic Geology and Paleontology of the Four Corners Area, pp. 29– 247. New Mexico Geological Society Guidebook 48. Fiorillo, A. R. and R. A. Gangloff. 2003. Preliminary notes on the taphonomic and paleoecologic setting of a Pachyrhinosaurus bonebed in northern Alaska. Journal of Vertebrate Paleontology 23(3, Suppl.): 50A. Forster, C. A. 1990. The cranial morphology of Triceratops, and a preliminary phylogeny of the Ceratopsia. Ph.D. diss., University of Pennsylvania, Philadelphia. ———. 1996. New information on the skull of Triceratops. Journal of Vertebrate Paleontology 16: 246–258. Forster, C. A., P. C. Sereno, T. W. Evans, and T. Rowe. 1993. A complete skull of Chasmosaurus mariscalensis (Dinosauria: Ceratopsidae) from the Aguja Formation (late Campanian) of west Texas. Journal of Vertebrate Paleontology 13: 161–170. Gill, J. R., and W. A. Cobban. 1973. Stratigraphic and geologic history of the Montana Group and equivalent rocks, Montana, Wyoming, and North and South Dakota. U.S. Geological Survey Professional Paper 776: 1–37. Godefroit, P. S., S. Zan, and L. Jin. 2001. The Maastrichtian (Late Cretaceous) lambeosaurine dinosaur Charonosaurus jiayinensis from north-eastern China. Bulletin de l’Institute Royale de Science Naturelle de Belgique 71: 119–157. Godfrey, S. J., and R. Holmes. 1995. Cranial morphology and systematics of Chasmosaurus (Dinosauria: Ceratopsidae) from the Upper Cretaceous of western Canada. Journal of Vertebrate Paleontology 15: 726–742. Goodwin, M. B., and A. L. Deino. 1989. The first radiometric ages from the Judith River Formation (Upper Cretaceous), Hill County, Montana. Canadian Journal of Earth Sciences 26: 1384– 1391. Gould, S. J. 2002. The Structure of Evolutionary Theory. Cambridge: Belknap Press. Happ, J. W., and C. M. Morrow. 1996. Separation of Triceratops (Dinosauria: Ceratopsidae) into two allopatric species by cranial morphology. Journal of Vertebrate Paleontology 16(3, Suppl.): 40A. Henderson, D. M. 2007. Skull shapes as indicators of niche partitioning by sympatric chasmosaurine and centrosaurine dinosaurs. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 77–78. Drumheller: Royal Tyrrell Museum of Palaeontology. Holmes, R. B., C. A. Forster, M. J. Ryan, and K. M. Shepherd. 2001. A new species of Chasmosaurus (Dinosauria: Ceratopsia) from the Dinosaur Park Formation of southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Horner, J. R. 1984. Three ecologically distinct vertebrate faunal communities from the Late Cretaceous Two Medicine Formation of Montana, with discussion of evolutionary pressures induced by interior seaway fluctuations. Montana Geological Society Field Conference Guidebook, pp. 299–303. ———. 1997. Dinosaur Lives. New York: Harcourt Brace & Co. Horner, J. R., D. J. Varricchio, and M. B. Goodwin. 1992. Marine transgressions and the evolution of Cretaceous dinosaurs. Nature 358: 59–61.

Jinnah, J. A., A. D. Deino, T. A. Gates, and E. M. Roberts. 2007. The first 40Ar/ 39Ar age date from the Wahweap Formation (Late Cretaceous of Utah): Implications for fauna correlations. Journal of Vertebrate Paleontology 27(3, Suppl.): 96A. Kelt, D. A., and D. H. Van Vuren. 2001. The ecology and macroecology of mammalian home range area. American Naturalist 157: 637–645. Kirkland, J. I., and D. D. DeBlieux. 2006. A new genus of ornate long-horned centrosaurine ceratopsian from the Middle Campanian Wahweap Formation, Grand Staircase-Escalante National Monument, southern Utah. Journal of Vertebrate Paleontology 26(3, Suppl.): 85A. ———. 2010. New basal centrosaurine ceratopsian skulls from the Wahweap Formation (Middle Campanian), Grand Staircase– Escalante National Monument, southern Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 117–140. Bloomington: Indiana University Press. Lehman, T. M. 1987. Late Maastrichtian paleoenvironments and dinosaur biogeography in the western interior of North America. Palaeogeography, Palaeoclimatology, Palaeoecology 60: 189– 217. ———. 1990. The ceratopsian subfamily Chasmosaurinae: Sexual dimorphism and systematics. In P. J. Currie and K. Carpenter, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 211– 229. New York: Cambridge University Press. ———. 1996. A horned dinosaur from the El Picacho Formation of west Texas, and review of ceratopsian dinosaurs from the American Southwest. Journal of Paleontology 17: 494–508. ———. 1997. Late Campanian dinosaur biogeography in the western interior of North America. In D. L. Wolberg, E. Stump, and G. D. Rosenburg, eds., Dinofest International: Proceedings of a Symposium Held at Arizona State University, pp. 223–240. Philadelphia: Academy of Natural Sciences. ———. 1998. A gigantic skull and skeleton of the horned dinosaur Pentaceratops sternbergi from New Mexico. Journal of Paleontology 72: 894–906. ———. 2001. Late Cretaceous dinosaur provinciality. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 310–328. Bloomington: Indiana University Press. ———. 2007. Growth and population age structure in the horned dinosaur Chasmosaurus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 259–318. Bloomington. Lilligraven, J. A., and L. M. Ostresh. 1990. Late Cretaceous (earliest Campanian/Maastrichtian) evolution of western shorelines of the North American Western Interior Seaway in relation to known mammalian faunas. In T. M. Bown and K. D. Rose, eds., Dawn of the Age of Mammals in the Northern Part of the Rocky Mountain Interior, North America, pp. 1–30. Geological Society of America Special Paper 243. Loewen, M. A., S. D. Sampson, E. K. Lund, A. A. Farke, M. C. Aguillón-Martínez, C. A. de Leon, R. A. Rodríguez-de la Rosa, M. A. Getty, and D. A. Eberth. 2010. Horned dinosaurs (Ornithischia: Ceratopsidae) from the Upper Cretaceous (Cam-

panian) Cerro del Pueblo Formation, Coahuila, Mexico. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 99–116. Bloomington: Indiana University Press. Lucas, S. G., R. M. Sullivan, and A. P. Hunt. 2006. Re-evaluation of Pentaceratops and Chasmosaurus (Ornithischia: Ceratopsidae) in the Upper Cretaceous of the Western Interior. In S. G. Lucas and R. M. Sullivan, eds., Late Cretaceous Vertebrates from the Western Interior, pp. 367–370. New Mexico Museum of Natural History and Science Bulletin 35. Lund, E. K., M. A. Loewen, S. D. Sampson, M. A. Getty, A. Aguillon Martinez, R. A. Rodriguez de la Rosa, and D. A. Eberth. 2007. Ceratopsian remains from the Late Cretaceous Cerro del Pueblo Formation, Coahuila, Mexico. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 108–113. Drumheller: Royal Tyrrell Museum of Palaeontology. Marquet, P. A., and M. L. Taper. 1998. On size and area: Patterns of mammalian body size extremes across landmasses. Evolutionary Ecology 12: 127–139. Marsh, O. C. 1891. Notice of new vertebrate fossils. American Journal of Science 42: 265–269. McDonald, A. T., and J. R. Horner. 2010. New material of ‘‘Styracosaurus’’ ovatus from the Two Medicine Formation of Montana. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 156–168. Bloomington: Indiana University Press. Nessov, L. A., L. F. Kaznyshkina, and G. O. Cherepanov. 1989. Mesozoic ceratopsian dinosaurs and crocodilians of central Asia. In T. N. Bogdanova and L. L. Khozatsky, eds., Theoretical and Applied Aspects of Modern Paleontology, pp. 144–154. Moscow. [In Russian.] Ostrom, J. H. and P. Wellnhoffer. 1986. The Munich specimen of Triceratops, with a revision of the genus. Zitteliana 14: 111–158. Ott, C. J., and P. L. Larson. 2010. A new, small ceratopsian dinosaur from the latest Cretaceous Hell Creek Formation, northwest South Dakota, United States: A preliminary description. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 203–218. Bloomington: Indiana University Press. Panhuis, T. M., R. Butlin, M. Zuk, and T. Tregenza. 2001. Sexual selection and speciation. Trends in Ecology and Evolution 16: 364–371. Penkalski, P., and P. Dodson. 1999. The morphology and systematics of Avaceratops, a primitive horned dinosaur from the Judith River Formation (late Campanian) of North America, with the description of a second skull. Journal of Vertebrate Paleontology 19: 692–711. Peters, R. H. 1983. The Ecological Implications of Body Size. New York: Cambridge Press. Roberts, E. M., A. L. Deino, and M. A. Chan. 2005. 40Ar/ 39Ar age of the Kaiparowits Formation, southern Utah and correlation of contemporaneous Campanian strata and vertebrate faunas

Unraveling a Radiation 425

along the margin of the Western Interior Basin. Cretaceous Research 26: 307–318. Rogers, R. R. 1998. Sequence analysis of the Upper Cretaceous Two Medicine and Judith River formations, Montana: Nonmarine response to Claggett and Bearpaw marine cycles. Journal of Sedimentary Research 68: 615–631. Rogers, R. R., C. C. Swisher III, and J. R. Horner. 1993. 40Ar/ 39Ar age and correlation of the nonmarine Two Medicine Formation (Upper Cretaceous), northwestern Montana. Canadian Journal of Earth Sciences 30: 1066–1075. Russell, D. A. 1967. A census of dinosaur specimens collected in western Canada. National Museum of Canada Natural History Papers 36: 1–13. ———. 1995. China and the lost worlds of the dinosaur era. Historical Biology 10: 3. Russell, D. A., and M. Manabe. 2002. Synopsis of the Hell Creek (uppermost Cretaceous) dinosaur assemblage. In J. H. Hartman, K. R. Johnson, and D. J. Nichols, eds., The Hell Creek Formation and the Cretaceous-Tertiary Boundary in the Northern Great Plains: An Integrated Continental Record of the End of the Cretaceous, pp. 169–176. Geological Society of America Special Paper 361. Ryan, M. J. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., D. B. Brinkman, D. A. Eberth, P. J. Currie, and D. H. Tanke. 2006. A new Pachyrhinosaurus-like ceratopsian from the upper Dinosaur Park Formation (Late Campanian) of southern Alberta. Journal of Vertebrate Paleontology 26(3, Suppl.): 117A. Ryan, M. J., and D. C. Evans. 2005. Ornithischian dinosaurs. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 312–348. Bloomington: Indiana University Press. Ryan, M. J., R. Holmes, and A. P. Russell. 2007. A revision of the late Campanian centrosaurine ceratopsid genus Styracosaurus from the Western Interior of North America. Journal of Vertebrate Paleontology 27: 944–962. Ryan, M. J., and A. P. Russell. 2005. A new centrosaurine ceratopsid from the Oldman Formation of Alberta and its implications for centrosaurine taxonomy and systematics. Canadian Journal of Earth Sciences 42: 1369–1387. Ryan, M. J., A. P. Russell, and S. Hartman. 2010. A new chasmosaurine ceratopsid from the Judith River Formation, Montana. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 181–188. Bloomington: Indiana University Press. Sampson, S. D. 1995. Two horned dinosaurs from the Upper Cretaceous Two Medicine Formation of Montana; with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15: 743–760. ———. 1997. Bizarre structures and dinosaur evolution. In D. L. Wolberg, E. Stump, and G. D. Rosenburg, eds., Dinofest International: Proceedings of a Symposium Held at Arizona State University, pp. 39–45. Philadelphia: Academy of Natural Sciences. ———. 2001. Speculations on the socioecology of ceratopsid dino-

426 sampson & loewen

saurs (Ornithischia: Neoceratopsia). In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 263–276. Bloomington: Indiana University Press. Sampson, S. D., T.A. Gates, E. M. Roberts, M. A. Getty, L. E. Zanno, M. A. Loewen, J. A. Smith, E. K. Lund, J. Sertich, and A. L. Titus. In press. Grand Staircase–Escalante National Monument: A new and critical window into the world of dinosaurs. Learning from the Land, Vol. 2. Bureau of Land Management. Sampson, S. D., and M. A. Loewen. 2005. Tyrannosaurus rex from the Upper Cretaceous (Maastrichtian) North Horn Formation of Utah: Biogeographic and paleoecologic implications. Journal of Vertebrate Paleontology 25: 469–472. ———. 2007. New information on the diversity, stratigraphic distribution, biogeography, and evolution of ceratopsid dinosaurs. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 125–133. Drumheller: Royal Tyrrell Museum of Palaeontology. Sampson, S. D., M. A. Loewen, E. M. Roberts, J. A. Smith, L. E. Zanno, and T. A. Gates. 2004. Provincialism in Late Cretaceous terrestrial faunas: New evidence from the Campanian Kaiparowits Formation of Utah. Journal of Vertebrate Paleontology 24(3, Suppl.): 108A. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Sankey, J. T. 2010. Faunal composition and significance of highdiversity, mixed bonebeds containing Agujaceratops mariscalensis and other dinosaurs, Aguja Formation (Upper Cretaceous), Big Bend, Texas. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 520–537. Bloomington: Indiana University Press. Sankey, J. T., S. Atchley, L. Nordt, S. Dworkin, and S. Driese. 2007. Vertebrates and paleoclimate from a Chasmosaurus mariscalensis bonebed, Late Cretaceous (late Campanian), Big Bend National Park, Texas. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 134–139. Drumheller: Royal Tyrrell Museum of Palaeontology. Scotese, C. R. 2001. Atlas of Earth History. Vol. I: Paleogeography. Arlington, Texas: PALEOMAP Project. Smith, J. A., S. D. Sampson, E. M. Roberts, M. A. Getty, and M. A. Loewen. 2004. A new chasmosaurine ceratopsian from the Upper Cretaceous Kaiparowits Formation, Grand StaircaseEscalante National Monument, Utah. Journal of Vertebrate Paleontology 24(3, Suppl.): 114A. Streelman, J. T., and P. D. Danley. 2003. The stages of vertebrate evolutionary radiation. Trends in Ecology and Evolution 18(3): 126–131. Sullivan, R. M., and S. G. Lucas. 2003. The Kirtlandian, a new land-vertebrate ‘‘age’’ for the Late Cretaceous of Western North America. In S. G. Lucas, S. C. Semken, W. R. Berglof, and D. S. Ulmer-Scholle, eds., Geology of the Zuni Plateau, New Mexico Geological Society Guidebook 54, pp. 369–377. Socorro: New Mexico Geological Society, Socorro.

———. 2007. A new chasmosaurine (Ceratopsidae, Dinosauria) from the Upper Cretaceous Ojo Alamo Formation (Naashoibito Member), San Juan Basin, New Mexico. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 141. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2010. A new chasmosaurine (Ceratopsidae, Dinosauria) from the Upper Cretaceous Ojo Alamo Formation (Naashoibito Member), San Juan Basin, New Mexico. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 169–180. Bloomington: Indiana University Press. Van Valkenburgh, B., and C. M. Janis. 1993. Historical diversity patterns in North American large herbivores and carnivores. In R. E. Ricklefs and D. Schluter, eds., Species Diversity in Ecological Communities, pp. 330–340. Chicago: University of Chicago Press. Vrba, E. S. 1987. Ecology in relation to speciation rates: Some case histories of Miocene-Recent mammal clades. Evolutionary Ecology 1: 283–300. Wolfe, D. G. 2000. New information on the skull of Zuniceratops christopheri, a neoceratopsian dinosaur from the Cretaceous

Moreno Hill Formation, New Mexico. In S. G. Lucas and A. B. Heckert, eds., Dinosaurs of New Mexico, pp. 93–94. New Mexico Museum of Natural History and Science Bulletin 17. Wolfe, D. G., and J. I. Kirkland. 1998. Zuniceratops christopheri n. gen. & n. sp., a ceratopsian dinosaur from the Moreno Hill Formation (Cretaceous, Turonian) of west-central New Mexico. Lower and Middle Cretaceous Terrestrial Ecosystems. New Mexico Museum of Natural History and Science Bulletin 24: 307–317. Wolfe, D. G., J. I. Kirkland, D. Smith, K. Poole, B. J. ChinneryAllgeier, and A. McDonald. 2007. Zuniceratops christopheri: An update on the North American Ceratopsid sister taxon, Zuni Basin, west-central New Mexico. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 159–167. Drumheller: Royal Tyrrell Museum of Palaeontology. Wu, X., D. B. Brinkman, D. A. Eberth, and D. R. Braman. 2007. A new ceratopsid dinosaur (Ornithischia) from the uppermost Horseshoe Canyon Formation (Upper Maastrichtian), Alberta, Canada. Canadian Journal of Earth Sciences 44: 1243–1265. You, H.-L., and P. Dodson. 2004. Basal Ceratopsia. In D. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 478–493. Berkeley: University of California Press.

Unraveling a Radiation 427

28 A Review of Ceratopsian Paleoenvironmental Associations and Taphonomy D AV I D A . E B E R T H

currently, much of the stratigraphic, paleoenviron-

Ceratopsian taphonomic data are patchy and, histori-

mental, and taphonomic data associated with most of

cally, have been drawn largely from monotaxic to mono-

the 70+ species of ceratopsians lacks detail. Accordingly,

dominant bonebeds, especially in the case of

there is a clear need for detailed paleoenvironmental and

centrosaurines. Although high-quality and exquisite

taphonomic work on ceratopsians, especially in the case

three-dimensional preservation of basal ceratopsians and

of relatively new Asian discoveries. Such information is

basal neoceratopsians from China and Mongolia have

key in evaluating paleobiological and paleoecological as-

been a catalyst for the taphonomic study of Asian speci-

pects and trends within the group. Here, a preliminary

mens, much more work remains to be done in both Asia

assessment is provided of the paleoenvironmental asso-

and North America. Further taphonomic work will be

ciations of ceratopsians through their approximately 95

critical in helping to resolve questions that are now

million year history (earliest Late Jurassic–end of Creta-

emerging about the putative life habits of psittacosaurs

ceous) using stratigraphic assignments, inferred paleo-

(partially or fully aquatic versus fully terrestrial), Pro-

climatic and depositional settings, and unusual

toceratops (nocturnal versus diurnal; burrower?), and

geological features (e.g., pyroclastics). A taphonomic re-

neoceratopsians (degree of gregariousness among cen-

view of some of the taxa is also provided.

trosaurines versus chasmosaurines).

Paleoenvironmental associations suggest that, as a group, ceratopsians retained a long-term preferential association with wetland paleoenvironments (lacustrine,

Introduction

alluvial and coastal plain). By the Late Cretaceous, how-

As recognized here (including data from this volume), there

ever, numerous basal neoceratopsians (Asia and North

are approximately 70 species of ceratopsian dinosaur (Ap-

America) were also exploiting marginal to fully eolian

pendix 28.1). The group existed for roughly 95 million years,

settings that were semi-arid to arid, or seasonally wet-

extending from the earliest Late Jurassic (Yinlong downsi,

dry. Neoceratopsians in Canada and Alaska flourished

160 Ma, northwest China) to the end of the Cretaceous

during the Campanian, expanding their diversity in the

(Triceratops horridus, 65.5 Ma [Ogg et al. 2004], western North

extensive warm-temperate to subtropical wetlands that

America). With the exception of Serendipaceratops—a poorly

dominated the coastal lowlands along the western shore

known possible basal neoceratopsian from Australia—mem-

of the Western Interior Seaway.

bership is strictly Laurasian, with forms known from Canada,

428

the United States, Mexico, Russia, Mongolia, China, Thailand,

of Psittacosaurus, and also regards a number of Psittacosaurus

and Uzbekistan.

species as synonymous. Thus, following Sereno (this volume),

Given the large diversity and abundance of ceratopsians,

only 13 basal ceratopsians species are considered valid. The

their broad paleogeographic and stratigraphic distributions,

age range for this group is early Late Jurassic (Oxfordian)

and the increasing availability of stratigraphic, sedimento-

through Early Cretaceous (Albian), a span of roughly 60 mil-

logic and taphonomic data via translations and English-

lion years (Gradstein et al. 2004).

language publications that relate to them, a review of cera-

In contrast, basal neoceratopsians are highly diverse, con-

topsian paleoenvironmental and taphonomic patterns/asso-

sisting of 21 genera, and 24 species (Archaeoceratops, Lepto-

ciations was deemed an appropriate topic for this volume. The

ceratops and Protoceratops each consist of two species). The

objectives of this manuscript are twofold: (1) assess paleoen-

group extends at least from mid-Early Cretaceous (Aptian-

vironmental associations and taphonomic patterns of cera-

Albian) to the end of the Cretaceous—a span of 55 million

topsians through time and across space; and (2) evaluate the

years—with most members ranging in age from mid-Early Cre-

potential significance of these patterns in the context of long-

taceous to Campanian (Appendix 28.1). Although most basal

standing and new ideas about ceratopsians.

neoceratopsians are Asian, Zuniceratops, Cerasinops, Montano-

The data examined here were drawn from the literature, per-

ceratops, Prenoceratops, and Leptoceratops are North Ameri-

sonal experience, and many of the manuscripts submitted to

can (Turonian-Maastrichtian), and Serendipaceratops is a very

this volume. Data were organized by assigning each genus/

poorly known form from the Early Cretaceous of Australia. All

species to one of four high-order taxonomic groups (Appendix

are small to medium in size (ⱕ4 m long adults).

28.1): basal ceratopsians (paraphyletic), basal neoceratopsians

Advanced neoceratopsians, or ceratopsids, are represented

(paraphyletic), and two groups of advanced neoceratopsians

by 26 genera and 31 species that can be assigned to either the

(ceratopsids): centrosaurines and chasmosaurines (monophy-

Centrosaurinae (11 gen., 14 spp.) or the Chasmosaurinae (15

letic). Data include stratigraphic placement, age, type of oc-

gen., 17 spp.). Most genera are monospecific, with the notable

currence (single versus multiple individuals), inferred pa-

exceptions of Chasmosaurus, Pachyrhinosaurus, and Centro-

leoclimate and depositional setting, and association with

saurus, which consist of 3, 3, and 2 species, respectively. Cera-

prominent climate indicators such as coal, caliche, and re-

topsids extend in time from the Santonian to the end of the

dbeds or association with unusual geological features (e.g.,

Cretaceous, a span of 20 million years. All are North Ameri-

pyroclastics).

can, and most, with the exception of Avaceratops, are large (?4

Because detailed stratigraphic, paleoenvironmental and

m long adults).

taphonomic data are lacking for most specimens, the results

In review, these taxonomic, stratigraphic, and paleogeo-

and patterns identified here are limited and hypothetical,

graphic data confirm (1) the Asian origins and long-term

but can be tested as more data are collected. Accordingly, I

dominance of ceratopsians through the early Cretaceous (e.g.,

strongly recommend that the interested reader review the pa-

Chinnery and Horner 2007), and (2) the middle Late Creta-

pers in this volume written by Chinnery-Allgeier and Kirk-

ceous explosion in ceratopsian diversity and increased body

land; Eberth et al.; Fiorillo et al.; Ford and Martin; Getty et al.;

sizes during and after entry into North America (Fig. 28.1).

Goodwin and Horner; Hunt and Farke; Kirkland and Bader; Longrich; Sampson and Loewen; and Sankey. Many of these authors present new data relating to these questions, as well

Paleoenvironmental Associations

as slightly modified to completely new perspectives on cera-

BASAL CERATOPSIANS

topsian paleoenvironments, and preservational mechanisms and taphonomy. Hopefully, their data and ideas, in combina-

Basal ceratopsians of China and Mongolia are largely asso-

tion with the patterns and data-gaps identified here, will stim-

ciated with paludal (wetland) settings near shallow to deep

ulate a reconsideration of some ceratopsian life habits, eco-

lakes, basin margins, and extensional volcanic terrains (Brink-

logical associations, adaptability, and preservational processes

man et al. 2001; Qi et al. 2007; Ford and Martin this volume;

through time.

Sereno this volume; You et al. this volume). A few associations with alluvial and eolian settings are also documented ( Jer-

Who’s Who, Where, and When?

zykiewicz and Russell 1991; Russell and Zhao 1996). At most localities, there is evidence for seasonal rainfall and/or semi-

Basal ceratopsians are relatively small forms ([2 m long) and

aridity (e.g., Brinkman et al. 2001), suggesting the possibility

include 5 genera (Yinlong, Chaoyangsaurus, Xuanhuaceratops,

of climatically induced, seasonal death events (Qi et al. 2007).

‘‘Hongshanosaurus,’’ and Psittacosaurus) and 16 species, all

Throughout northeastern China, basal ceratopsians (espe-

from Asia, and mostly from China (Appendix 28.1). Sereno

cially psittacosaurs) are associated with a variety of pyro-

(this volume) regards ‘‘Hongshanosaurus’’ as a junior synonym

clastics preserved in lacustrine and lacustrine-margin settings

A Review of Ceratopsian Paleoenvironmental Associations and Taphonomy 429

FIGURE 28.1.

Stratigraphy, paleogeography, and diversity of ceratopsians compiled from numerous sources, including this volume (see Appendix 28.1).

( Jiang and Sha 2006; Qi et al. 2007). In these deposits, excep-

in and around aquatic settings. These data support the notion

tionally high-quality preservation suggests rapid death and

that, given their consistently small size, basal ceratopsians

burial (obrution; Qi et al. 2007) and, possibly, additional con-

may have been behaviorally or ecologically linked to a life

ditions that inhibited bacterial decomposition and infaunal

within or near aquatic environments, and were unable to ex-

scavenging. Similarly, Yinlong, from the Junggar Basin (Xu et

ploit a more full range of terrestrial settings.

al. 2006), is represented by multiple occurrences of articulated skeletons that appear to have been rapidly entombed by mudflows (Eberth unpublished data).

BASAL NEOCERATOPSIANS

Although the breadth of paleoenvironmental associations

As a group, basal neoceratopsians exhibit the longest tempo-

and the wide paleogeographic distribution seen within psit-

ral range among the four groups of ceratopsians discussed

tacosaurs across Asia suggest a low level of paleoenvironmen-

here (Fig. 28.1), as well as the broadest variety of paleoen-

tal specificity, Ford and Martin (this volume) and others (e.g.,

vironmental associations and paleogeographic distributions

Suslov 1983; Averianov et al. 2006 [references from Ford and

(Appendix 28.1). Early Cretaceous forms from Asia include

Martin]) have suggested that the frequent association of psit-

Liaoceratops, which occurs with the basal ceratopsians, Psit-

tacosaurs with lacustrine and wetland paleoenvironments

tacosaurus lujiatunensis and ‘‘Hongshanosaurus houi,’’ from Lia-

may indicate that some of these dinosaurs were adapted to life

oning’s pyroclastic/lacustrine facies. Archaeoceratops and Au-

430 eberth

roraceratops are associated with seasonally arid-to-subtropical,

tween the groups was not so extreme as to be mutually exclu-

alluvial settings in Mongolia and western China (You et al.

sive in broad geographic and paleoenvironmental terms. How-

this volume). Two poorly known forms, Asiaceratops and Kul-

ever, in terrestrial beds that are Maastrichtian age (70.6–65.5

ceratops, are known from warm temperate, paralic settings

Ma [Ogg et al. 2004]), only chasmosaurs are present, indicating

of Uzbekistan (Nessov et al. 1989), whereas Serendipacera-

an extinction of centrosaurines at about the Campanian-

tops is known from Australia (paleoenvironmental data not

Maastrichtian boundary. At present there is no clearly under-

available).

stood reason as to what may have driven this extinction, al-

A variety of basal neoceratopsians are known from the Late

though it has been noted that chasmosaurs appear to have

Cretaceous (Santonian-early Maastrichtian) of Asia, including

been less gregarious than centrosaurs (e.g., Hunt and Farke this

Uzbekistan (Turanoceratops), China (Magnirostris; Protocera-

volume), which may have increased their survivability in some

tops), and Mongolia (Bagaceratops, Bainoceratops, Gracilicera-

way (M. J. Ryan pers. com. 2008). During the Campanian (both

tops, Lamaceratops, Platyceratops, Protoceratops, Udanoceratops,

groups) and the early Maastrichtian (chasmosaurines only)

and Yamaceratops). With the exceptions of Graciliceratops and

these groups exhibit strong associations with poorly drained

Yamaceratops, all are associated with semi-arid settings that

alluvial to coastal plain environments (Achelousaurus, Avacera-

are dominated by a variety of eolian, interdune alluvial-to-

tops, Centrosaurus, Einiosaurus, Pachyrhinosaurus, Styracosaurus,

paludal, and distal alluvial fan paleoenvironments (Gradzin-

Anchiceratops, Arrhinoceratops, Chasmosaurus, Pentaceratops,

ski and Jerzykiewicz 1974; Jerzykiewicz and Russell 1991;

Torosaurus, Eotriceratops). Throughout the Campanian a few

Dashzeveg et al. 2005). Graciliceratops appears to have been

centrosaurines (Albertaceratops) and chasmosaurines (Dicera-

associated with seasonally wet/dry alluvial wetlands (Sereno

tops) seemed to have preferred better drained alluvial plain

2000), whereas Yamaceratops is associated with a variety of

settings, and Late Maastrichtian forms like Triceratops appear

alluvial channel and interchannel sediments deposited in

to have continued expressing that ‘‘drier’’ paleoenvironmental

a seasonally semi-arid to subtropical setting (Eberth et al.

preference until the end of the Cretaceous.

2009). Basal neoceratopsians also include North American forms,

Paleoenvironmental Summary

which are an important indicator of an Asia-to-North America ceratopsian migration event(s) during the Late Creta-

The preceding review of paleoenvironmental associations

ceous (Wolfe and Kirkland 1998; Chinnery and Horner 2007;

suggests that, as a group, ceratopsians retained a long-term

Chinnery-Allgeier and Kirkland this volume). The oldest of

preferential association with wetland paleoenvironments, re-

these, Zuniceratops, a Turonian form from New Mexico, is as-

gardless of whether they were lacustrine, alluvial or coastal

sociated with alluvial-to-upper coastal plain, coal bearing sed-

plain settings. By the Late Cretaceous, however, numerous

iments that were deposited in a subtropical, seasonally wet

basal neoceratopsians were exploiting marginal to fully eolian

climate.

settings that were semi-arid to arid, as well as seasonally wet-

Montanoceratops, Prenoceratops, Cerasinops, and Leptoceratops are known from Montana and Alberta, and are mostly associ-

dry and well-drained alluvial settings across Asia and North America.

ated with seasonally dry paleoenvironmental settings charac-

Most of the neoceratopsians in Canada and Alaska appear

terized by calcareous paleosols and redbeds that developed in

to have flourished and expanded their diversity in the exten-

a warm temperate climate (Nadon 1994; Eberth and O’Con-

sive warm-temperate to subtropical wetlands that dominated

nell 1995; Chinnery and Weishampel 1998; Chinnery 2004;

the coastal lowlands along the western shore of the Western

Chinnery and Horner 2007). A few poorly preserved basal

Interior Seaway during the Late Cretaceous (especially the

neoceratopsians are also associated with non-calcareous,

Campanian), where seasonality was likely of the wet-and-

poorly drained, and coaly alluvial-to-coastal plain sediments

wetter variety.

in the same region (Ryan and Currie 1998; Ryan and Evans 2005).

Taphonomic Patterns Taphonomic studies of ceratopsians are typically focused

NEOCERATOPSIANS

on bonebed occurrences, and thus, our understanding of

Neoceratopsians (chasmosaurine and centrosaurine ceratop-

taphonomic patterns among ceratopsians is actually quite

sids) are a diverse group of strictly North American forms with

patchy, not lending itself to broad taxonomic categorization,

a Campanian-Maastrichtian temporal range (Fig. 28.1; Appen-

as in the case of paleoenvironmental associations. Bonebed

dix 28.1). Where Campanian age North American dinosaur

occurrences of Asian and North American ceratopsians attract

material is abundant and well exposed, the two subfamilies co-

considerable taphonomic interest (e.g., Currie and Dodson

occur, indicating that any ecological or niche partitioning be-

1984; Rogers 1990; Ryan et al. 2001; Eberth and Getty 2005;

A Review of Ceratopsian Paleoenvironmental Associations and Taphonomy 431

Mathews et al. 2007; Qi et al. 2007; Eberth et al. this volume;

numerous, fully articulated skeletons suggest (1) the absence

Getty et al. this volume; Hunt and Farke this volume; Sankey

of a lengthy history of ‘‘bloat-and-float’’ in Yixian lake waters

this volume). It is not surprising then that there is a significant

with attendant body part loss, and (2) high rates of sedi-

bias in the overall ceratopsian taphonomic data-sets that em-

mentation and/or conditions that suppressed decay once car-

phasizes ‘‘dramatic’’ bonebed associations—especially mono-

casses had settled on the lake bottom (e.g., cool temperatures,

taxic to monodominant varieties (Fiorillo and Eberth 2004;

anoxia).

Eberth et al. 2007). Whereas there are some excellent taphonomic studies that focus on monodominant, neoceratopsian bonebeds from the

BASAL NEOCERATOPSIANS

Western Interior of the United States and Canada (e.g., Rogers

The articulated Protoceratops specimen from Tugrikin-Shireh,

1990; Ryan et al. 2001; Eberth and Getty 2005) these often

Mongolia (Djadokhta Formation), that is preserved in direct

overlook the taphonomy of numerous isolated ceratopsid

contact with a Velociraptor (the ‘‘fighting’’ dinosaurs on dis-

specimens that occur in the same formations. Conversely, al-

play at the Mongolian Academy of Sciences in Ulaanbaatar;

though there are numerous anecdotal observations and pub-

Barsbold 1974) is arguably the most frequently discussed cera-

lished photographs that influence scientific opinions about

topsian taphonomic association and, probably among the

the taphonomy of individual ceratopsians from China and

most famous dinosaur associations known (Barsbold 1974;

Mongolia (e.g., Jerzykiewicz et al. 1993), detailed paleoen-

Kielan-Jaworowska 1975; Jerzykiewicz et al. 1993; Osmólska

vironmental and taphonomic studies of these specimens, or

1993; Unwin et al. 1995; Fastovsky et al. 1997; Carpenter

others from the same regions are, in fact, quite rare.

2001; Fiorillo and Eberth 2004). Although some authors regard the association as representative of an in-life interaction

BASAL CERATOPSIANS

between two species of dinosaur that in some manner resulted in their mutual demise (Barsbold 1974; Kielan-Jaworowska

Recent examination of an assemblage of Psittacosaurus speci-

1975), others reject the notion that both animals died to-

mens from the base of the Yixian Formation (Lujiatun beds,

gether (e.g., Osmólska 1993), preferring to interpret the Velo-

Liaoning, China) suggests rapid burial in a volcanic mudflow

ciraptor as trapped and buried during scavenging of a Proto-

(lahar), close to the site of volcanic eruption (Qi et al. 2007).

ceratops carcass.

However, even in this exceptionally preserved assemblage,

Fastovsky et al. (1997) documented the entombment and

there is uncertainty as to whether or not the lahar was gener-

skeletal preservation of numerous other Protoceratops speci-

ated by a volcanic eruption or secondary ‘‘wet’’’ reworking of

mens at Tugrikin-Shireh, showing that many animals died

volcanic debris on the flanks of the volcano. Higher in section,

on the lee side of prograding dunes. Although some degree

many tens to hundreds of articulated psittacosaurs have been

of postmortem exposure and drying was proposed, rapid

collected from tuffaceous lacustrine shales of the Yixian For-

burial by eolian storm events was considered a likely cause of

mation. Although there is yet no clear accounting for the

burial (cf. Eberth 1993). An alternative hypothesis proposed

death and burial histories of these ceratopsians and other

by Loope et al. (1998) suggested that many vertebrate fos-

large terrestrial vertebrates, Guo et al. (2003) have suggested

sils known from the Djadokhta Formation at Ukhaa Tolgod

that episodic lethal emissions of volcanogenic chemicals re-

(Mongolia), including Protoceratops, were buried and possibly

sulted in mass kills of the local terrestrial vertebrate assem-

killed by sand slides that became more frequent during mesic

blage. In this volume, Ford and Martin propose that these and

(wet) climatic phases across Central Asia. The range of preser-

other psittacosaurs are perhaps better interpreted as semi-

vation and facies associations in the Djadokhta Formation in

aquatic dinosaurs that inhabited lakes and ponds in the area.

Mongolia and China (Eberth 1993; Dashzeveg et al. 2005) sug-

In support of their hypothesis, they point to the overwhelm-

gest that both eolian sand storm and mass sediment flow

ing association of many of psittacosaur specimens with lac-

events were probably at work preserving ceratopsian and

ustrine shales, as well as a number of other skeletal features,

other vertebrate fossils in these beds. Even so, Longrich (this

the presence of gastroliths, and aspects of psittacosaur func-

volume) proposes a new way of looking at these Protoceratops

tional morphology. Whatever role life habits may have played

occurrences, suggesting that individuals may have been noc-

in psittacosaur remains being entombed in lacustrine shales,

turnal, residing in a burrow during the day, especially at

or whatever the causes/processes of death, burial and preser-

times when diurnal temperatures were very high. In Long-

vation that resulted in exquisite preservation, future tapho-

rich’s hypothesis, occupying a burrow on a regular basis

nomic interpretations must account for superb preservation

may have predisposed individuals of Protoceratops to be-

of articulated skeletons and the rarity of isolated elements or

coming entombed during times of high sediment supply. This

body parts. For now, it can be argued that the presence of

hypothesis provides an elegant solution as to why so many

432 eberth

Protoceratops are well preserved in these Djadokhta Forma-

a small sample size for chasmosaurines, centrosaurine and

tion eolian deposits and why they exhibit uniquely three-

chasmosaurine bonebeds may actually prove to be quantifia-

dimensional ‘‘standing’’ skeletal postures.

bly different in terms of the size of the biocoenoses and

Kirkland also studied the taphonomic history of a skeleton

thanatocoenoses that they represent, with chasmosaurine

of Protoceratops at Tugrikin-Shireh ( J. Kirkland pers. com.

bonebeds being generally smaller than those of centrosau-

2008), and his observations, as well as other data, are pre-

rines, and preferentially associated with channel deposits.

sented in this volume as evidence for carrion-eating beetle

Their cautionary approach challenges the degree to which we

activity in association with these and other vertebrate car-

can be certain about gregarious behaviors in any of these taxa.

casses preserved in the Djadokhta paleoenvironments (Kirk-

Their ideas are further supported in part by data presented by

land and Bader this volume).

Lehman (2007), Eberth et al. (this volume), Fiorillo et al. (this volume), Sankey (this volume), and, possibly, Getty et al. (this volume). The issues they raise will be important for neocera-

NEOCERATOPSIANS

topsian workers to consider in the future, especially in light of

Sternberg (1970) provided some observations about dinosaur

new data that are available about ceratopsid skull shape and

taphonomy from the Upper Cretaceous of southern Alberta.

inferred niche partitioning (Henderson this volume).

He noted that isolated skulls of ceratopsid dinosaurs are com-

Considering the lengthy stratigraphic range of monotaxic

mon in the sediments at Dinosaur Provincial Park. More com-

and monodominant ceratopsian bonebeds (Barremian to

mon, in fact, than their articulated postcranial remains. He

Maastrichtian), gregarious behavior is likely to have been

interpreted this pattern as reflecting preferential scavenging

plesiomorphic for ceratopsians. If so, the general absence of

(by theropods) of ceratopsian postcranial remains, and the

bonebeds for some well represented ceratopsian taxa (e.g., Tri-

overall robust nature of sutured ceratopsid cranial remains

ceratops; but see Mathews et al. 2007 for an important excep-

that would have resisted decomposition and reworking. He

tion to the ‘‘rule’’) may reflect a significant divergence from

also interpreted the presence of ceratopsian bonebeds at the

this pattern, or alternatively, may reflect a significant change

Park as a function of accumulation of ceratopsian carcasses in

in the depositional and preservational characteristics of the

swamps, and subsequent trampling of the remains by a variety

host formation (cf. Goodwin and Horner this volume).

of large dinosaurs. Dodson (1971, 1983) confirmed that cera-

Data from integrative geologic/paleontologic studies can be

topsian skulls are more common than are their postcranial

of help in assessing the significance of differences in the com-

remains, that articulated skeletons from juvenile and subadult

position of syntaxonomic ceratopsian bonebeds within and

ceratopsians are rare at Dinosaur Provincial Park, but that cer-

between formations, and this approach shows promise in

atopsian remains are associated with both paleochannel and

helping to provide more subtle paleobiological insights re-

overbank deposits. Eberth and Currie (2005) suggested that

lated to breeding and migration habits. For example, although

many of the dinosaurs preserved at Dinosaur Provincial Park

centrosaurines are known throughout the Dinosaur Park For-

(including the ceratopsians) succumbed to large scale coastal

mation in southern Alberta, bonebeds containing their re-

plain flooding, but that only those deposited in river channels

mains are restricted to paleogeographic locations more than

were buried quickly enough to be preserved with their articu-

200 km up-dip from paleoshoreline. This peculiar and non-

lated skeletons intact.

random paleogeographic distribution of centrosaurine bone-

The taphonomy of monodominant ceratopsian bonebeds

beds has been interpreted as reflecting that this group may

has been the focus of numerous studies during the past

have had a preference for isolation or small family groupings

25 years (Lehman 1982; Currie and Dodson 1984; Visser 1986;

when they were close to the shoreline (during nesting?) versus

Rogers 1990; Ryan 1992; Ryan 2003; Ryan et al. 2001; Fiorillo

a preference for large aggregations (after nesting?) during or

2004; Eberth and Getty 2005; Eberth et al. this volume;

after migration away from shoreline (Brinkman et al. 1998;

Fiorillo et al. this volume; Getty et al. this volume). The pic-

Eberth et al. this volume). As more data are collected, similar

ture that has emerged is remarkably consistent from study to

kinds of patterns may ultimately be revealed and tested for

study. In virtually every case, ceratopsian bonebed assem-

other ceratopsid taxa that are known from numerous localities

blages are interpreted as the remains of a group of gregarious

and stratigraphic positions (e.g., Chasmosaurus, Triceratops).

ceratopsians that were overcome, en masse, by extreme paleoenvironmental conditions (e.g., drought, flood). In most cases, it is proposed that postmortem exposure, trampling,

Taphonomic Summary

scavenging, and hydraulic reworking all contributed to the

Taphonomic data from ceratopsians and their localities are

development of complex taphonomic signatures. However,

quite patchy and, historically, have focused on monotaxic to

Hunt and Farke (this volume) raise the possibility that despite

monodominant bonebeds, especially in the case of centro-

A Review of Ceratopsian Paleoenvironmental Associations and Taphonomy 433

saurines. High-quality and exquisite three-dimensional preservation of basal ceratopsians and basal neoceratopsians from

References Cited

China and Mongolia have been a catalyst for new taphonomic

Alifanov, V. R. 2003. Two new dinosaurs of the infraorder Neoceratopsia (Ornithischia) from the Upper Cretaceous of the Nemegt depression, Mongolian People’s Republic. Paleontological Journal 37(5): 524–534. Averianov, A., A. V. Voronkevich, S. V. Leshchinskiy, and A. V. Fayngertz. 2006. A ceratopsian dinosaur Psittacosaurus sibiricus from the Early Cretaceous of West Siberia, Russia and its phylogenetic relationships. Journal of Systematic Palaeontology 4: 359–395. Barsbold, R. 1974. Duelling dinosaurs. Priroda 2: 81–83. [In Russian.] Bohlin, B. 1953. Fossil Reptiles from Mongolia and Kansu. Reports from the Scientific expedition to the North-Western Provinces of China under Leadership of Dr. Sven Hedin. SinoSwedish Expedition Publication 37: 9–113. Brinkman, D. B., D. A. Eberth, M. J. Ryan, and P. J. Chen. 2001. The occurrence of Psittacosaurus xinjiangensis Sereno and Chow, 1988, in the Urho area, Junggar Basin, Xinjiang, People’s Republic of China. Canadian Journal of Earth Sciences 38: 1781–1786. Brinkman, D. B., M. J. Ryan, and D. A. Eberth. 1998. The paleogeographic and stratigraphic distribution of ceratopsids (Ornithischia) in the Upper Judith River Group of Western Canada. Palaios 13: 160–169. Brown, B. B. 1914. Leptoceratops, a new genus of Ceratopsia from the Edmonton Cretaceous of Alberta. Bulletin of the American Museum of Natural History 33: 567–580. Brown, B. B., and E. M. Schlaikjer. 1942. The skeleton of Leptoceratops with the description of a new species. American Museum Novitates 1169: 1–15. Buffetaut, E., and V. Suteethorn. 1992. A new species of the ornithischian dinosaur Psittacosaurus from the Early Cretaceous of Thailand. Palaeontology 35: 810–822. Carpenter, K. 2001. Evidence of predatory behavior by carnivorous dinosaurs. Gaia 15: 135–144. Chinnery, B. J. 2004. Description of Prenoceratops pieganensis gen et sp. nov. (Dinosauria: Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Chinnery, B. J., and J. R. Horner. 2007. A new neoceratopsian dinosaur linking North American and Asian taxa. Journal of Vertebrate Paleontology 27: 625–641. Chinnery, B. J., and D. B. Weishampel. 1998. Montanoceratops cerorhynchus (Dinosauria: Ceratopsia) and relationships among basal neoceratopsians. Journal of Vertebrate Paleontology 18: 569–585. Chinnery-Allgeier, B. J., and J. I. Kirkland. 2010. An update on the biogeography of ceratopsian dinosaurs. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 387–404. Bloomington: Indiana University Press. Currie, P. J., and P. Dodson. 1984. Mass death of a herd of ceratopsian dinosaurs. In W. E. Reif and F. Westphal, eds., Third

studies focusing on a handful of these specimens, but much more work remains to be done on the taphonomy and preservational histories of ceratopsians from those regions. Further taphonomic work in Asia and North America will be critical in helping to resolve questions that are now emerging about the life habits of psittacosaurs (partially or fully aquatic versus fully terrestrial), Protoceratops (nocturnally versus diurnally active; burrower?), neoceratopsians (degree of gregariousness; paleogeography), as well as other associated vertebrate taxa.

Conclusions Paleoenvironmental associations suggest that, as a group, ceratopsians retained a long-term preferential association with wetland paleoenvironments (lacustrine, alluvial and coastal plain). By the Late Cretaceous numerous basal neoceratopsians (Asia and North America) were exploiting marginal to fully eolian settings that were semi-arid to arid or seasonally wet-dry. Neoceratopsians in Canada and Alaska flourished during the Campanian, expanding their diversity in the extensive warm-temperate to subtropical wetlands that dominated the coastal lowlands along the western shore of the Western Interior Seaway at that time. The patchy taphonomic data sets relating to ceratopsians have been drawn largely from monotaxic to monodominant bonebeds, especially those comprising centrosaurines. Highquality, three-dimensionally preserved specimens of basal ceratopsians and basal neoceratopsians from China and Mongolia have been a catalyst for a few taphonomic studies in those regions. Further taphonomic work on ceratopsians will be critical in helping to resolve questions that are now emerging about ceratopsian paleobiology. Acknowledgments

I thank my co-editors, Michael Ryan and Brenda ChinneryAllgeier, for their help and forbearance. Don Henderson and Michael Ryan graciously reviewed this manuscript on very short notice. I thank Layne Syvertsen for initiating the review of ceratopsian-bearing formations at my request. Lastly, I thank all of the contributors to this volume, whose papers added significantly to this manuscript. In particular, the contributions of Nick Longrich, Tracy Ford and Larry Martin, ReBecca Hunt and Andy Farke, and Jim Kirkland and Ken Bader were stimulating and helpful. Any errors or misrepresentations are mine alone.

434 eberth

Symposium of Mesozoic Terrestrial Ecosystems, pp. 52–60. Tubigen: Attempto Verlag. Currie, P. J., W. Langston, Jr., and D. H. Tanke. 2008. A new species of Pachyrhinosaurus (Dinosauria, Ceratopsidae) from the Upper Cretaceous of Alberta, Canada. In P. J. Currie, W. Langston Jr., and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. 1–108. Ottawa: National Research Council Press. Dashzeveg, D., L. Dingus, D. B. Loope, C. C. Swisher III, T. Dulam, and M. R. Sweeney. 2005. New stratigraphic subdivision, depositional environment, age estimate for the Upper Cretaceous Djadokhta Formation, southern Ulan Nur Basin, Mongolia. American Museum Novitates 3498: 1–31. Dodson, P. 1971. Sedimentology and taphonomy of the Oldman Formation (Campanian), Dinosaur Provincial Park, Alberta (Canada). Palaeogeography, Palaeoclimatology, Palaeoecology 10: 21–74. ———. 1983. A faunal review of the Judith River (Oldman) Formation, Dinosaur Provincial Park, Alberta. Mosasaur 1: 89–118. ———. 1986. Avaceratops lammersi: A new ceratopsid from the Judith River Formation of Montana. Proceedings of the Academy of Natural Sciences of Philadelphia 138(2): 305–317. Dong, Z.-M., and Y. Azuma. 1997. On a primitive neoceratopsian from the Early Cretaceous of China. In D. Zhiming, ed., SinoJapanese Silk Road Dinosaur Expedition, pp. 68–89. Beijing: China Ocean Press. Eberth, D. A. 1993. Depositional environments and facies transitions of dinosaur-bearing Upper Cretaceous redbeds at Bayan Mandahu (Inner Mongolia, People’s Republic of China). Canadian Journal of Earth Sciences 30: 2196–2213. Eberth, D. A., D. B. Brinkman, and V. Barkas. 2010. A centrosaurine mega-bonebed from the Upper Cretaceous of southern Alberta: Implications for behavior and death events. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 495–508. Bloomington: Indiana University Press. Eberth, D. A., and P. J. Currie. 2005. Vertebrate taphonomy and taphonomic modes. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 453–477. Bloomington: Indiana University Press. Eberth, D. A., and M. A. Getty. 2005. Ceratopsian bonebeds at Dinosaur Provincial Park: Stratigraphy, geology, taphonomy, origins, and significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 501–536. Bloomington: Indiana University Press. Eberth, D. A., Y. Kobayashi, Y.-N. Lee, O. Mateus, F. Therrien, D. K. Zelenitsky, and M. A. Norell. 2009. Assignment of Yamaceratops dorngobiensis and associated redbeds at Shine Us Khudag (Eastern Gobi, Dorngobi Province, Mongolia) to the redescribed Javkhlant Formation (Upper Cretaceous). Journal of Vertebrate Paleontology 29:295–302. Eberth, D. A., and S. C. O’Connell. 1995. Notes on changing paleoenvironments across the Cretaceous-Tertiary boundary (Scollard Formation) in the Red Deer River valley of southern Alberta. Bulletin of Canadian Petroleum Geology 43: 44–53.

Eberth, D. A., M. Shannon, and B. G. Noland. 2007. A bonebeds database: Classification, biases, and patterns of occurrence. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 103–219. Chicago: University of Chicago Press. Fastovsky, D. E., D. Badamgarav, H. Ishimoto, M. Watabe, and D. B. Weishampel. 1997. The paleoenvironments of TugrikinShireh (Gobi Desert, Mongolia) and aspects of the taphonomy and paleoecology of Protoceratops (Dinosauria: Ornithischia). Palaios 12: 59–70. Fiorillo, A. R. 2004. The dinosaurs of Arctic Alaska. Scientific American 291: 84–92. Fiorillo, A. R., and D. A. Eberth. 2004. Dinosaur taphonomy. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 607–613. Berkeley: University of California Press. Fiorillo, A. R., P. J. McCarthy, P. P. Flaig, E. Brandlen, D. W. Norton, P. Zippi, L. Jacobs, and R. A. Gangloff. 2010. Paleontology and paleoenvironmental interpretation of the Kikak-Tegoseak Quarry (Prince Creek Formation: Late Cretaceous), northern Alaska: A multi-disciplinary study of a high-latitude ceratopsian dinosaur bonebed. In M. J. Ryan, B. J. ChinneryAllgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 456–477. Bloomington: Indiana University Press. Ford, T. L., and L. M. Martin. 2010. A semi-aquatic life habit for Psittacosaurus. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 328–339. Bloomington: Indiana University Press. Getty, M. A., M. A. Loewen, E. Roberts, A. L. Titus, and S. D. Sampson. 2010. Taphonomy of horned dinosaurs (Ornithischia: Ceratopsidae) from the Late Campanian Kaiparowits Formation, Grand Staircase–Escalante National Monument, Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 478–494. Bloomington: Indiana University Press. Gilmore, C. W. 1930. On dinosaurian reptiles from the Two Medicine Formation of Montana. Proceedings of the U.S. National Museum 77: 1–39. Goodwin, M. B., and J. R. Horner. 2010. Historical collecting bias and the fossil record of Triceratops in Montana. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 551–563. Bloomington: Indiana University Press. Gradstein, F., J. Ogg, and A. Smith, eds. 2004. A Geologic Time Scale 2004. Cambridge: Cambridge University Press. Gradzinski, R., and T. Jerzykiewicz. 1974. Dinosaur and mammal-bearing aeolian and associated deposits of the Upper Cretaceous in the Gobi Desert (Mongolia). Sedimentary Geology 12: 249–278. Granger, W., and W. K. Gregory. 1923. Protoceratops andrewsi, a pre-ceratopsian dinosaur from Mongolia. American Museum Novitates 72: 1–9.

A Review of Ceratopsian Paleoenvironmental Associations and Taphonomy 435

Guo Z. F, J. Q. Liu, and X. L. Wang. 2003. Effect of Mesozoic volcanic eruptions in the western Liaoning Province, China on paleoclimate and paleoenvironment. Science in China (Series D) 46: 1261–1272. Henderson, D. M. 2010. Skull shapes as indicators of niche partitioning by sympatric chasmosaurine and centrosaurine dinosaurs. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 293–307. Bloomington: Indiana University Press. Holmes, R., C. A. Forster, M. J. Ryan, and K. M. Shepherd. 2001. A new species of Chasmosaurus (Dinosauria: Ceratopsia) from the Dinosaur Park Formation of southern Alberta. Canadian Journal of Earth Sciences 38: 1423–1438. Hunt, R. K., and A. A. Farke. 2010. Behavioral interpretations from ceratopsid bonebeds. In M. J. Ryan, B. J. ChinneryAllgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 447–455. Bloomington: Indiana University Press. Jerzykiewicz, T., P. J. Currie, D. A. Eberth, P. A. Johnston, E. H. Koster, and J. Zheng. 1993. Djadokhta Formation correlative strata in Chinese Inner Mongolia: An overview of the stratigraphy, sedimentary geology and paleontology and comparisons with the type locality in the Pre-Altai Gobi. Canadian Journal of Earth Sciences 30: 2180–2195. Jerzykiewicz, T., and D. A. Russell. 1991. Late Mesozoic stratigraphy and vertebrates of the Gobi Basin. Cretaceous Research 12: 345–377. Jiang, B.-Y., and J.-G. Sha. 2006. Preliminary analysis of the depositional environments of the Lower Cretaceous Yixian Formation in the Sihetun area, western Liaoning, China. Cretaceous Research: 1–11. Kielan-Jaworowska, Z. 1975. Late Cretaceous mammals and dinosaurs from the Gobi Desert. American Scientist 63: 150– 159. Kirkland, J. I., and K. Bader. 2010. Insect trace fossils associated with Protoceratops carcasses in the Djadokhta Formation (Upper Cretaceous), Mongolia. In M. J. Ryan, B. J. ChinneryAllgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 509–519. Bloomington: Indiana University Press. Kirkland, J. I., and D. D. DeBlieux. 2010. New basal centrosaurine ceratopsian skulls from the Wahweap Formation (Middle Campanian), Grand Staircase–Escalante National Monument, southern Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 117–140. Bloomington: Indiana University Press. Kurzanov, S. M. 1990. A new Late Cretaceous Protoceratopsid genus from Mongolia. Palaeontological Journal 24: 85–91. ———. 1992. Gigantskiy protoseratopsid iz verkhnengo mela Mongolii. Paleontologischeskii Zhural 1992(3): 81–93. Kurzanov, S. M., and K. E. Mikhailov. 1988. The finding of dinosaur eggshells in the Lower Cretaceous of Mongolia. Fossil reptiles and birds of Mongolia. Joint Soviet-Mongolian Palaeontological Expedition Transaction 34: 72–76.

436 eberth

Lambe, L. M. 1902. New genera and species from the Belly River Series (Mid-Cretaceous). Contributions to Canadian Palaeontology 3: 22–81. ———. 1904. On the squamoso-parietal crest of horned dinosaur Centrosaurus apertus and Monoclonius canadensis from the Cretaceous of Alberta. Transactions of the Royal Society of Canada 4: 3–13. ———. 1913. A new genus and species of Ceratopsia from the Belly River Formation of Alberta. Ottawa Naturalist 27: 109–116. ———. 1914. On Gryposaurus notabilis, a new genus and species of trachodont dinosaur from the Belly River Formation of Alberta, with a description of the skull of Chasmosaurus belli. Ottawa Naturalist 27(11): 145–153. Lambert, O., P. Godefroit, H. Li, C.-Y. Shang, and Z.-M. Dong. 2001. A new species of Protoceratops (Dinosauria, Neoceratopsia) from the Late Cretaceous of Inner Mongolia (P. R. China). Bulletin de l’institut royal des Sciences Naturelles de Belgique, Sciences de la Terre 71(Suppl.): 5–28. Lehman, T. M. 1982. A ceratopsian bone bed from the Aguja Formation (Upper Cretaceous) Big Bend National Park, Texas. M.A. thesis, University of Texas, Austin. ———. 1989. Overview of Late Cretaceous sedimentation in TransPecos Texas. In A. B. Busbey III and T. M. Lehman, eds., Vertebrate Paleontology, Biostratigraphy and Depositional Environments, Latest Cretaceous and Tertiary, Big Bend Area, Texas. Field Trip Guidebook for the Society of Vertebrate Paleontology, pp. 23–46. ———. 2007. Growth and population age structure in the horned dinosaur Chasmosaurus. In K.Carpenter, ed., Horns and Beaks, Ceratopsian and Ornithopod Dinosaurs, pp. 259–317. Bloomington: Indiana University Press. Loewen, M. A., S. D. Sampson, E. K. Lund, A. A. Farke, M. C. Aguillón-Martínez, C. A. de Leon, R. A. Rodríquez-de la Rosa, M. A. Getty, and D. A. Eberth. 2010. Horned dinosaurs (Ornithischia: Ceratopsidae) from the Upper Cretaceous (Campanian) Cerro del Pueblo Formation, Coahuila, Mexico. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 99–116. Bloomington: Indiana University Press. Longrich, N. 2010. The function of large eyes in Protoceratops: A nocturnal ceratopsian? In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 308–327. Bloomington: Indiana University Press. Loope, D. B., L. Dingus, C. C. Swisher III, and C.-L. Minjin. 1998. Life and death in a Late Cretaceous dune field, Nemegt Basin, Mongolia. Geology 25: 27–30. Lucas, S. G., R. M. Sullivan, and A. P. Hunt. 2006. Re-evaluation of Pentaceratops and Chasmosaurus (Ornithischia: Ceratopsidae) in the Upper Cretaceous of the western interior. In S. G. Lucas and R. M. Sullivan, eds., Late Cretaceous Vertebrates from the Western Interior, pp. 367–370. New Mexico Museum of Natural History & Science Bulletin 35. Lull, R. S. 1905. Restoration of the horned dinosaur Diceratops. American Journal of Science, Series 3, 20: 420–422. Makovicky, P. J., and M. A. Norell. 2006. Yamaceratops dorngo-

biensis, a new primitive ceratopsian (Dinosauria: Ornithischia) from the Cretaceous of Mongolia. American Museum Novitates 3530: 1–42. Marsh, O. C. 1889. Notice of giant horned Dinosauria from the Cretaceous. American Journal of Science, 3rd Series, 38: 173–175. ———. 1891. Notice of new vertebrate fossils. American Journal of Science 42: 265–269. Maryanska, ´ T., and H. Osmólska. 1975. Protoceratopsidae (Dinosauria) of Asia. Palaeontologica Polonica 33: 135–181. Mathews, J. C., M. Henderson, and S. Williams. 2007. Taphonomy, sedimentology, and paleoenvironmental reconstruction of a unique Triceratops site in the Hell Creek, Southeastern Montana. Geological Society of America Abstracts with Programs 39: 9. Nadon, G. C. 1994. The genesis and recognition of anastomosed fluvial deposits; data from the St. Mary River Formation, southwestern Alberta, Canada. Journal of Sedimentary Research 64: 451–463. Nessov, L. A. 1995. Dinozavri severnoi Yevrasii: Novye dannye o sostave kompleksov, ekologii i paleobiogeografii [Dinosaurs of Northern Eurasia: New Data about Assemblages, Ecology and Paleobiogeography]. St. Petersburg: Scientific Research Institute of the Earth’s Crust. Nessov, L. A., L. Kaznyshkina, and G. Cherepanov. 1989. Mesozoic dinosaurians—ceratopsians and crocodiles of Central Asia. Trudy Sessii Vsesoyuznogo Paleontologicheskogo Obshchestva 33: 144–154. Ogg, J. G., F. P. Agterberg, and F. M. Gradstein. 2004. The Cretaceous Period. In F. Gradstein, J. Ogg, and A. Smith, eds., A Geologic Time Scale 2004, pp. 344–383. Cambridge: Cambridge University Press. Osborn, H. F. 1923. Two Lower Cretaceous dinosaurs of Mongolia. American Museum Novitiates 95: 1–10. Osmólska, H. 1993. Were the Mongolian ‘‘fighting dinosaurs’’ really fighting? Review de. Paléobiologie 7: 161–162. Ott, C. J., and P. L. Larson. 2010. A new, small ceratopsian dinosaur from the latest Cretaceous Hell Creek Formation, northwest South Dakota, United States: A preliminary description. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 203—218. Bloomington: Indiana University Press. Parks, W. A. 1925. Arrhinoceratops brachyops, a new genus and species of Ceratopsia from the Edmonton Formation of Alberta. University of Toronto Studies, Geological Series 19: 1–15. Penkalski, P. G., and P. Dodson. 1999. The morphology and systematics of Avaceratops, a primitive horned dinosaur from the Judith River Formation (Late Campanian) of Montana, with the description of a second skull. Journal of Vertebrate Paleontology 19(4): 692–711. Qi, Z., P. M. Barrett, and D. A. Eberth. 2007. Evidence for postnestling social behaviour in the primitive ceratopsian dinosaur. Palaeontology 50: 1023–1029. Rich, T. H., and P. Vickers-Rich. 2003. Protoceratopsian? ulnae from Australia. Records of the Queen Victoria Museum 113: 1–10. Rogers, R. R. 1990. Taphonomy of three dinosaur bone beds in

the Upper Cretaceous Two Medicine Formation of northwestern Montana: Evidence for drought related mortality. Palaios 5: 394–413. Russell, D. A., and X.-J. Zhao. 1996. New psittacosaur occurrences in Inner Mongolia. Canadian Journal of Earth Sciences 33: 637–648. Ryan, M. J. 1992. The taphonomy of a Centrosaurus (Reptilia: Ornithischia) bone bed (Campanian), Dinosaur Provincial Park, Alberta, Canada. M.Sc. thesis, University of Calgary, Calgary. ———. 2003. Taxonomy, systematics and evolution of centrosaurine ceratopsids of the Campanian Western Interior Basin of North America. Ph.D. diss., University of Calgary, Calgary. ———. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81(2): 376–396. Ryan, M. J., and P. J. Currie. 1998. First report of protoceratopsians (Neoceratopsia) from the Late Campanian Judith River Group, Alberta, Canada. Canadian Journal of Earth Sciences 35: 820–826. Ryan, M. J., D. A. Eberth, D. B. Brinkman, P. J. Currie, and D. H. Tanke. 2010. A new Pachyrhinosaurus-like ceratopsid from the upper Dinosaur Park Formation (Late Campanian) of southern Alberta, Canada. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 141–155. Bloomington: Indiana University Press. Ryan, M. J., and D. C. Evans. 2005. Review of the Ornithischia of Dinosaur Provincial Park. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 313–348. Bloomington: Indiana University Press. Ryan, M. J., and A. P. Russell. 2005. A new centrosaurine ceratopsid from the Oldman Formation of Alberta and its implications for centrosaurine taxonomy and systematics. Canadian Journal of Earth Sciences 42: 1369–1387. Ryan, M. J., A. P. Russell, D. A. Eberth, and P. J. Currie. 2001. The taphonomy of a Centrosaurus (Ornithischia: Certopsidae [sic]) bone bed from the Dinosaur Park Formation (Upper Campanian), Alberta, Canada, with comments on cranial ontogeny. Palaios 16: 482–506. Sampson, S. D. 1995. Two new horned dinosaurs from the Upper Cretaceous Two Medicine Formation of Montana; with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15(4): 743–360. Sankey, J. T. 2010. Faunal composition and significance of highdiversity, mixed bonebeds containing Agujaceratops mariscalensis and other dinosaurs, Aguja Formation (Upper Cretaceous), Big Bend, Texas. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 520–537. Bloomington: Indiana University Press. Sereno, P. C. 2000. The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia. In M. J. Benton, M. A. Shishkin, D. M. Unwin, and E. N. Kurochkin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. 480–516. New York: Cambridge University Press.

A Review of Ceratopsian Paleoenvironmental Associations and Taphonomy 437

———. 2010. Taxonomy, cranial morphology, and relationships of parrot-beaked dinosaurs (Ceratopsia: Psittacosaurus). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 21–58. Bloomington: Indiana University Press. Sereno, P. C., and S. Chao. 1988. Psittacosaurus xinjiangensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of northwestern China. Journal of Vertebrate Paleontology 8(4): 353–365. Sereno, P. C., S. Chao, Z. Cheng, and C. Rao. 1988. Psittacosaurus meileyingensis (Ornithischia: Ceratopsia), a new psittacosaur from the Lower Cretaceous of Northeastern China. Journal of Vertebrate Paleontology 8(4): 366–377. Sereno, P. C., X.-J. Zhao, L. Brown, and L. Tan. 2007. New psittacosaurid highlights skull enlargement in horned dinosaurs. Acta Palaeontologica Polonica 52: 275–284. Sternberg, C. M. 1940. Ceratopsidae from Alberta. Journal of Palaeontology 14(5): 468–480. ———. 1950. Pachyrhinosaurus canadensis, representing a new family of the Ceratopsia, from Southern Alberta. National Museum of Canada Bulletin 118: 109–114. ———. 1951. Complete skeleton of Leptoceratops gracilis Brown from the Upper Edmonton member on Red Deer River, Alberta. National Museum of Canada Bulletin 123: 225–255. ———. 1970. Comments on dinosaurian preservation in the Cretaceous of Alberta and Wyoming. National Museums of Canada, Publications in Paleontology 4: 9 p. Sullivan, R. M., and S. G. Lucas. 2010. A new chasomosaurine (Ceratopsidae, Dinosauria) from the Upper Cretaceous Ojo Alamo Formation (Naashoibito Member), San Juan Basin, New Mexico. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 169–180. Bloomington: Indiana University Press. Suslov, J. V. 1983. The locality of Psittacosaurus in Chamrin-us (East Gobi, MPR). Transactions of the Joint Soviet Mongolian Paleontological Expedition 24: 118–121. Tereshchenko, V. S., and V. R. Alifanov. 2003. Bainoceratops efremovi, a new protoceratopid dinosaur (Protoceratopidae, Neoceratopsia) from the Bain-Dzak Locality (South Mongolia). Paleontological Journal 37: 293–302. Unwin, D. M., A. Perle, and C. Trueman. 1995. Protoceratops and Velociraptor preserved in association: Evidence for predatory behavior in dromaeosaurid dinosaurs? Journal of Vertebrate Paleontology 15: 57. Visser, J. 1986. Sedimentology and taphonomy of a Styracosaurus bonebed in the Late Cretaceous Judith River Formation, Dinosaur Provincial Park, Alberta. M.Sc. thesis. University of Calgary, Calgary. Wolfe, D. G., and J. I. Kirkland. 1998. Zuniceratops christopheri n. gen. & n. sp., a ceratopsian dinosaur from the Moreno Hill Formation (Cretaceous, Turonian) of West-Central New Mexico. In S. G. Lucas, J. I. Kirkland, and J. W. Estep, eds., Lower and Middle Cretaceous Terrestrial Ecosystems, pp. 303–317. New Mexico Museum of Natural History and Science Bulletin 14.

438 eberth

Wolfe, D. G., J. I. Kirkland, R. Denton, and B. G. Anderson. 1997. A new terrestrial vertebrate record from the Moreno Hill Formation (Turonian, Cretaceous), west-central New Mexico. Journal of Vertebrate Paleontology 17(3, Suppl.): 85A–86A. Wu, X-C., D. B. Brinkman, D. A. Eberth, and D. R. Braman. 2007. A new ceratopsid dinosaur (Ornithischia) from the uppermost Horseshoe Canyon Formation (upper Maastrichtian), Alberta, Canada. Canadian Journal of Earth Sciences 44(9): 1243–1265. Xu, X. 1997. A new psittacosaur (Psittacosaurus mazongshanensis sp. nov.) from Mazongshan area, Gansu Province, China. In Z.-M. Dong, ed., Sino-Japanese Silk Road Dinosaur Expedition, pp. 48–67. Beijing: China Ocean Press. Xu, X., C. A. Forster, J. M. Clark, and J. Mo. 2006. A basal ceratopsian with transitional features from the Late Jurassic of northwestern China. Proceedings of the Royal Society B. 273: 2135–2140. Xu, X., P. J. Makovicky, X.-L. Wang, M. A. Norell, and H.-L. You. 2002. A ceratopsian dinosaur from China and the early evolution of ceratopsia. Nature 416: 314–317. You, H.-L., and D.-M. Dong. 2003. A new protoceratopsid (Dinosauria: Neoceratopsia) from the Late Cretaceous of Inner Mongolia, China. Acta Geologica Sinica 77: 299–303. You, H.-L., D. Li, Q. Ji, M. C. Lamanna, and P. Dodson. 2005. On a new genus of basal neoceratopsian dinosaur from the Early Cretaceous Gansu Province, China. Acta Geologica Sinica 79: 593–597. You, H., K. Tanoue, and P. Dodson. 2010. A new species of Archaeoceratops (Dinosauria: Neoceratopsia) from the Early Cretaceous of the Mazongshan area, northwestern China. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 59–67. Bloomington: Indiana University Press. You, H.-L., X. Xu, and X. Wang. 2003. A new genus of psittacosauridae (Dinosauria: Ornithopoda) and the origin and early evolution of marginocephalian dinosaurs. Acta Geologica Sinica 77: 15–20. Young, C.-C. 1958. The dinosaurian remains of Laiyang, Shantung. Palaeontologia Sinica 142C: 1–138. Zhao, Q. I., P. M. Barrett, and D. A. Eberth. 2007. Social behaviour and mass mortality in the basal Ceratopsian dinosaur Psittacosaurus (Early Cretaceous), People’s Republic of China. Palaeontology 50(5): 1023–1029. Zhao, X. 1985. The reptiles of Jurassic in China. In S.-E. Wang, ed., The Jurassic System of China, pp. 286–289. Beijing: Geological Publishing House. [In Chinese.] Zhao, X., Z. Cheng, and X. Xu. 1999. The earliest ceratopsian from the Tuchengzi Formation of Liaoning. Journal of Vertebrate Paleontology 19: 681–691. Zhao, X., Z. Cheng, X. Xu, and P. J. Makovicky. 2006. A new Ceratopsian from the Upper Jurassic Houcheng Formation of Hebei, China. Acta Geologica Sinica 80(4): 467–473. Zhou, C.-F., K.-Q. Gao, R. C. Fox, and S.-H. Chen. 2006. A new species of Psittacosaurus (Dinosauria; Ceratopsia) from the Early Cretaceous Yixian Formation, Liaoning, China. Palaeoworld 14: 100–114.

439

# of

(Inner Mongolia, China)

Bayan Gobi Formation

(Aptian)

single

wet/dry

seasonally

semi-arid to

subtropical,

margin, delta

fan-delta, lacustrine

Sereno et al. 2007

Psittacosaurus gobiensis

Mongolia, China)

unknown unit (Inner

(Liaoning, China);

(China); Jiufotang

Early Cretaceous

Shestakovskaya Svita

volume

Sereno this volume)

(Sereno et al. 2007)

indeterminate seasonally Aptian)

(Mongolia);

Sereno 1988; Sereno this

osborni; P. guyangensis; (Russia); Lisangou

airfall; terrestrial (Barremian-

Khulsyngoskaya Svita

wet/dry

lacustrine; pyroclastic subtropical, semi-arid to

Early Cretaceous

Khukhtekskaya Svita,

and Mikhailov 1988;

floodplain/paludal

alluvial-lacustrine;

delta, pyroclastic airfall

Osborn 1923; Kurzanov

semi-arid

subtropical,

wet/dry

lacustrine, lacustrine

pyroclastic flows&airfall wet/dry seasonally

mudflows&debris flows; seasonally

volcanic

pyroclastic flows&airfall

wet/dry subtropical,

mudflows&debris flows;

volcanic

pyroclastics

lacustrine-alluvial;

pyroclastics

lacustrine-alluvial;

alluvial; mudflows

basin margin; paludal-

Paleoenvironment

seasonally

subtropical,

wet/dry

seasonally

subtropical,

wet/dry

seasonally

subtropical,

wet/dry

seasonally

subtropical,

Climate

mongoliensis (includes P.

multiple

multiple

multiple

(bonebeds)

multiple

single

single

single

multiple

individuals

Psittacosaurus

(Barremian)

China)

Xinminpu (Gansu,

Early Cretaceous

Early Cretaceous

‘‘mazongshanensis’’

Xu 1997

Jiufotang (China)

(Hauterivian?)

Early Cretaceous

Psittacosaurus

Sereno et al. 1988

Psittacosaurus

China

Yixian (lujiatun beds)

(Hauterivian?)

(Aptian-Albian)

al. 2007

lujiatunensis

Early Cretaceous

(Liaoning, China)

late Late Jurassic

late Late Jurassic

Yixian (lujiatun beds)

China)

Tuchengzi (Liaoning,

early Late Jurassic

Age

meileyingensis

Zhou et al. 2006; Zhao et

Psittacosaurus

1985

You et al. 2003

China)

Zhao et al. 2006; Zhao

Xuanhuaceratops niei

‘‘Hongshanosaurus houi’’

Houcheng (Hebei,

Zhao et al. 1999

(Xinjiang, China)

Chaoyangsaurus youngi

Shishugou Formation

Stratigraphy

Xu et al. 2006

References

Yinlong downsi

Basal Ceratopsians

Taxa

pyroclastics

redbeds

pyroclastics

pyroclastics

pyroclastics

redbeds

coals/pyroclastics

Redbeds/caliches/

Appendix 28.1. Known Ceratopsian Genera and Species, Their Stratigraphy, Paleoenvironmental Associations, and Presence/Absence of Taphonomic Studies.

no

no

no

no

yes

no

no

no

no

Studies

Taphonomic

440 Early Cretaceous (?Aptian-Albian )

Gansu, China) Xinminpu Group (Gansu, China)

Dong and Azuma 1997

Nessov et al. 1989

volume)

Archaeoceratops oshimai

Asiaceratops sal-

wetlands semi-arid, seasonally wet Maastrichtian) Djadokhta redbeds

Kurzanov 1990

includes Breviceratops)

(Mongolia)

Ukhaa Tolgod

eolian; alluvial/paludal subtropical, (Campanian-

Khermeen Tsav;

Osmólska 1975;

multiple Late Cretaceous

‘‘Baruungoyot’’ red beds,

Maryanska ´ and

rozhdestvenskyi (genus

and dry

floodplain/paludal

alluvial-lacustrine;

Bagaceratops

semi-arid

subtropical, (?Aptian-Albian)

(Gansu, China)

single

Early Cretaceous

Xinminpu Group

wetlands?

alluvial/paralic/

You et al. 2005

temperate

warm

floodplain/paludal

alluvial-lacustrine;

floodplain/paludal

alluvial-lacustrine;

streams

alluvial, meandering

Auroraceratops rugosus

fragments

semi-arid

subtropical,

semi-arid

subtropical,

?

alluvial-lacustrine?

(Uzbekistan)

(Albian)

multiple

single

single

?

sopaludalis (n.d.)

Early Cretaceous

(?Aptian-Albian)

Khodzhakul Formation

Early Cretaceous

Archaeoceratops new sp. (Mazongshan area,

(Aptian-Albian)

Early Cretaceous

Xinminpu Group

Khok Kruat (Thailand)

(You & Dodson this

sians You et al. 2009

Suteethorn 1992

Basal Neoceratop-

Buffetaut and

sattayaraki’’

(Aptian-Albian)

Early Cretaceous

single

‘‘Psittacosaurus

Ilek (west Siberia)

paludal

Averianov et al. 2006

wet/dry

Psittacosaurus sibiricus

Early Cretaceous

alluvial lacustrine,

Tugulu (China)

Brinkman et al. 2001

margin/paludal, fluvial,

Lacustrine-

Sereno and Chao 1988;

seasonally

subtropical

eolian?

margin/paludal, fluvial,

lacustrine-

jiangensis

multiple

multiple

seasonally arid

Psittacosaurus xin-

(Aptian-Albian)

Early Cretaceous

single

eolian?

wet/dry

lacustrinemargin/paludal, fluvial,

semi-arid, seasonally-

pyroclastic

China)

volume

Early Cretaceous

multiple

Sereno this volume)

Qingshan (Shandong,

Young 1958; Sereno this

(includes P. youngi;

China)

Psittacosaurus sinensis

Russell and Zhao 1996

Ejinhoro (Ordos Basin,

‘‘ordosensis’’

Psittacosaurus

Early Cretaceous

pyroclastic flows&airfall

wet/dry

or early Albian)

ing, China) Ejinhoro (Ordos Basin,

mudflows&debris flows;

seasonally

(late Barremian

China)

Russell and Zhao 1996

Psittacosaurus

volcanic

subtropical,

single

Early Cretaceous

Paleoenvironment

(Lujiatun Beds; Liaon-

Climate

Yixian Formation

# of individuals

Age

Stratigraphy

neimongoliensis

Sereno et al. 2007

References

Psittacosaurus major

Taxa

Appendix 28.1. (Continued)

redbeds

redbeds

redbeds

redbeds

redbeds

redbeds

pyroclastics

redbeds

no

no

no

no

no

no

no

no

no

no

no

no

pyroclastics

redbeds

Studies

Taphonomic

coals/pyroclastics

Redbeds/caliches/

441

1953; Maryanska ´ and

Brown and Schlaikjer 1942; Sternberg 1951 Alifanov 2003

Montanoceratops

cerorhynchus

Platyceratops tatarinovi

wetlands sonally wet

(CampanianMaastrichtian)

meen Tsav (Mongolia)

and dry

eolian; alluvial/paludal

calcareous paleosols

alluvial, well-drained,

alluvial; alluvial fan

semi-arid, sea-

ate, semi-arid

warm temper-

arid

semi-arid-to-

eolian; interdune

pyroclastic flows&airfalls wet/dry? subtropical,

mudflows&debris flows;

volcanic/alluvial

phylic

coastal plain, hydro-

seasonally

subtropical,

wet

ate, seasonally

warm temper-

wet/dry

seasonally

temperate,

drained to hydrophylic

alluvial plain well-

subtropical,

single

two

warm temperate-

Late Cretaceous

(Maastrichtian)

single

multiple

single

multiple

Baruun Goyot? Kher-

tana)

Late Cretaceous

(Campanian)

St. Mary River Fm. (Mon-

Late Cretaceous

(Hauterivian?)

Early Cretaceous

dahu (China)

Yixian (Lujiatun beds;

(Campanian)

Late Cretaceous

Djadokhta, Bayan Man-

Xu et al. 2002

Liaoceratops

Dinosaur Park (Alberta)

(Maastrichtian)

Creek (Montana)

You and Dong 2003

Ryan and Currie 1998

Leptoceratops sp.

Late Cretaceous

Scollard; (Alberta); Hell

seasonally wet

Maastrichtian)

Magnirostris dodsoni

Brown 1914

Leptoceratops gracilis

wetlands

semi-arid, and dry

eolian; alluvial/paludal

subtropical,

single

(Campanian-

wetlands?

alluvial/paralic/

Late Cretaceous

temperate

warm

fossils

micro-

wetlands

wet/dry?

Mongolia

(Albian)

(Uzbekistan)

alluvial/paludal

seasonally

Baruun Goyot, Khulsan

Early Cretaceous

Khodzhakul Formation

Liaoning, China)

Alifanov 2003

Lamaceratops tereschenkoi

(Santonian-

Gashuun (Mongolia) Campanian)

Late Cretaceous

Bayn Shiren, Shireegiin

single

seasonally

Campanian) wet/dry

paleosols

paleosols

arid alluvial with calcareous

Late Cretaceous ate, semi-arid,

drainage, and calcareous

eolian with interdune

semi-arid-to-

subtropical,

(early-to-mid-

multiple

single

warm temper-

(Campanian)

Late Cretaceous

yanzigouensis

Nessov 1995

Kulceratops kulensis (n.d.)

Osmólska 1975

Sereno 2000; Bohlin

mation (Montana)

2007

mongoliensis

lower Two Medicine For-

Mongolia)

Alifanov 2003 Chinnery and Horner

Djadokhta (Bayn Dzak,

Tereschenko and

Graciliceratops

Cerasinops hodgskissi

Bainoceratops efremovi

redbeds

caliches

redbeds, caliches

pyroclastics

coals

coals

redbeds

caliches

redbeds, caliches

redbeds, caliches

no

no

yes

no

no

no

no

no

no

no

no

442 Kurzanov 1992

Makovicky and Norell 2006; Eberth et al. in

Udanoceratops tschizhovi

Yamaceratops dorngo-

biensis

Achelousaurus horneri

(Centrosaurinae)

Ceratopsids

Zuniceratops christopheri

Nessov et al. 1989

Turanoceratops tardabilis

Sampson 1995

Late Cretaceous (late Campanian)

tana)

(Turonian)

Late Cretaceous

Two Medicine (Mon-

Mexico)

(Santonian-

(eastern Gobi, Mongolia)

Moreno Hill Fm. (New

Late Cretaceous

Javkhlant Formation

and Kirkland 1998

(Campanian)

Campanian)

Late Cretaceous

Mongolia)

(Maastrichtian)

Late Cretaceous

Early Cretaceous

(Campanian)

Late Cretaceous

Djadokhta (Udan-Sayr,

(Uzbekistan)

Wolfe et al. 1997; Wolfe

press

2003

clarkei

(Australia)

Wanthaggi Formation

Rich and Vickers-Rich

Serendipaceratops arthurc-

Djadokhta (Zizhuqu, Inner Mongolia, China)

Lambert et al. 2001

hellenikorhinus

Protoceratops

Mongolia, China)

Mongoliazizhiqu, Inner

Tugreeken Shiren,

(Campanian)

(bonebed)

multiple

(bonebed)

multiple

multiple

single

single

single

single

multiple

Dazk, Alag Teeg,

1923

Late Cretaceous

Djadokhta; Minhe (Bayn

Granger and Gregory

bonebed

(Campanian)

multiple;

Late Cretaceous

Protoceratops andrewsi

# of individuals

tana)

Age

Two Medicine (Mon-

Stratigraphy

Chinnery 2004

References

Prenoceratops pieganensis

Taxa

Appendix 28.1. (Continued)

arid

ate, seasonally

warm temper-

wet/dry

seasonally-

tropical,

temperate-sub

warm

semi-arid

subtropical,

arid

semi-arid-to-

subtropical,

?

alluvial-coastal plain

coal bearing

alluvial/coastal plain;

alluvial-distal fan

age, alluvial fan

eolian, interdune drain-

?

?

eolian; paludal

arid ?

alluvial; alluvial fan;

semi-arid-to-

eolian, interdune

eolian; paludal

arid

subtropical,

alluvial; alluvial fan;

eolian, interdune

coal

coals

redbeds

redbeds, caliches

redbeds, caliches

redbeds, caliches

caliches

alluvial with calcareous paleosols

coals/pyroclastics

Redbeds/caliches/ Paleoenvironment

semi-arid-to-

subtropical,

wet/dry

seasonally

ate, semi-arid,

warm temper-

Climate

no

yes

no

no

no

no

no

yes

no

Studies

Taphonomic

443

Dodson 1986; Penkalski

Avaceratops lammersi

densis

Pachyrhinosaurus cana-

Sternberg 1950

multiple

seasonally wet

ate; cool tem-

warm temperperate, wet-

(late Campanian)

(Alaska)

multiple (bonebed)

Late Cretaceous

dering to anastomosed and dry

shoreline

coastal plain, estuarine,

streams

accomodation, mean-

alluvial plain, high seasonally wet

coals, caliches

Warm-hot,

lower-middle

1 single

rivers; hydromorphic to and dry

Campanian)

Campanian

organics dering and anastomosed well drained soils

minor caliches; alluvial plain with mean-

caliches

seasonally wet

alluvial-coastal plain

coals

caliches

coals

Subtropical,

arid

ate, seasonally

warm temper-

estuarine, shoreline

alluvial-coastal plain,

alluvial-coastal plain

alluvial plain

estuarine, shoreline

alluvial-coastal plain,

coastal plain-shoreline

alluvial plain

(middle

single

(bonebed)

wet and dry

ate, seasonally

Warm temper-

arid

ate, seasonally

warm temper-

(Alberta); Prince Creek

Wapiti; St Mary River

Horseshoe Canyon;

member’’

the middle mudstone

(Utah) ‘‘middle of

this volume

this volume)

Wahweap Formation

Kirkland and Deblieux

ern Utah)

Late Cretaceous

(late Campanian)

(Montana)

Kaiparowits Fm (south-

Late Cretaceous

Two Medicine

land and Deblieux

Getty et al. this volume

Gilmore 1930

New genus & sp. (Kirk-

et al. this volume)

New genus & sp. (Getty

this volume)

MacDonald and Horner

Styracosaurus ovatus;

New genus & sp. (=

Zone (Dinosaur Provin-

et al. this volume) cial Park, Alberta)

(late Campanian

single

Late Cretaceous

Ryan et al. this volume

Lethbridge Coal Zone

(late Campanian)

tana)

multiple (bonebed)

Late Cretaceous

Two Medicine (Mon-

arid

ate, seasonally

Campanian)

warm temper-

multiple (bonebed)

(mid-late

wet

ate, seasonally

warm temper-

arid

ate, seasonally

warm temper-

Late Cretaceous

tion; Lethbridge Coal

Sampson 1995

Einiosaurus procurvicornis

Oldman (Alberta)

multiple (bonebed)

(late Campanian)

(bonebed)

multiple

Late Cretaceous

(late Campanian)

Late Cretaceous

arid

warm temper-

nian)

single ate, seasonally

Late Cretaceous (mid-late Campa-

Dinosaur Park Forma-

Ryan and Russell 2005

Centrosaurus brinkmani

Dinosaur Park (Alberta)

Judith River (Montana)

Oldman (Alberta)

‘‘pachyrhinosaur’’ (Ryan

Lambe 1904

Centrosaurus apertus

and Dodson 1999

Ryan 2007

Albertaceratops nesmoi

no

no

yes

no

no

yes

no

yes

no

no

444 # of

al. 2006

Brown 1914

Parks 1925

Lambe 1902; Lambe

scalensis

Anchiceratops ornatus

Arrhinoceratops brachyops

Chasmosaurus belli

Holmes et al. 2001

Sternberg 1940

Chasmosaurus irvinensis

Chasmosaurus russelli

1914

Lehman 1989; Lucas et

Agujaceratops mari-

(Chasmosaurinae)

Ceratopsids

Lambe 1913

(late Campanian-

temperate-sub

Dinosaur Park (Alberta)

Dinosaur Park (Alberta)

(late Campanian)

Late Cretaceous

(late Campanian)

Late Cretaceous

(late Campanian)

Late Cretaceous

multiple

multiple

multiple

wet

ate, seasonally

warm temper-

wet

ate, seasonally

warm temper-

wet

ate, seasonally

warm temper-

estuarine, shoreline

alluvial-coastal plain,

shoreline

coastal plain, estuarine,

estuarine, shoreline

alluvial-coastal plain,

wet

early Maastrich-

Dinosaur Park (Alberta)

shoreline ate, seasonally

(late Campaniantian)

coastal plain, estuarine, warm temper-

Late Cretaceous

shoreline

coastal plain, estuarine,

coastal plain

(Alberta)

single

ate, seasonally wet

warm temper-

multiple (bonebed)

sonally wet

tropical; sea-

warm

multiple

estuarine, shoreline

alluvial-coastal plain,

Horseshoe Canyon

tian)

early Maastrich-

Late Cretaceous

shoe Canyon (Alberta)

Campanian)

(early-late

wet

ate, seasonally

(bonebed)

(late Campanian)

Late Cretaceous

multiple (bonebed)

Late Cretaceous

Dinosaur Park, Horse-

Aguja (Texas)

Dinosaur Park (Alberta)

warm temper-

sp.? (Fiorillo et al. this

Styracosaurus albertensis

alluvial-coastal plain,

estuarine, shoreline

alluvial-coastal plain,

Paleoenvironment

morphic soils

Cool temper-

and dry

seasonally wet

temperate,

Warm/cool

Climate

wet and dry

multiple (bonebed)

(Maastrichtian)

(bonebed)

multiple

individuals

Late Cretaceous

Late Cretaceous

Age

volume)

Prince Creek formation

(Grande Prairie)

Wapiti Formation

Stratigraphy

low energy, hydro-

Fiorillo et al. this volume

Currie et al. 2008

References

ate, seasonally

Pachyrhinosaurus new

(Currie et al. in press)

Pachyrhinosaurus new sp.

Taxa

Appendix 28.1. (Continued)

coals

coals

coals

coals

coals

caliches

coals

organic rich

coals

coals/pyroclastics

Redbeds/caliches/

no

no

no

no

no

yes

yes

yes

yes

Studies

Taphonomic

445

Ryan et al. this volume

low sinuosity to braided wet and dry Campanian)

Pentaceratops sternbergii

Osborn 1923

tropical, sea-

Maastrichtian)

sonally wet

temperate-sub

warm(Campanian-

Mexico)

multiple Late Cretaceous

Fruitland, Kirtland (New

shoreline

estuarine, paludal,

delta, coastal plain,

canic ash input

floodplain, high vol-

streams, well drained wet and dry tian

San Juan Basin, New Mexico

and meandering dry, seasonally

volume)

alluvial plain, braided Warm-hot, Early Maastrich-

Naashoibito Member,

Late Cretaceous,

Ojo Alamo Formation,

volume

floodplain Sullivan and Lucas this

1 single

to low accomodation, ate, seasonally

streams, well drained

alluvial plain, moderate Warm temper-

1 single

(mid-late

Late Cretaceous

livan and Lucas this

(Montana)

Judith River Formation

New genus & sp. (Sul-

this volume)

dering streams, hydrowet and dry

tian)

morphic soils

accomodation, mean-

1 single

alluvial plain, high

lower coastal plain

ate, seasonally

seasonally arid

Warm temper-

New genus & sp. (Ryan

(South Dakota)

(bonebed)

tiple

and 1 mul-

(Late Maastrich-

Late Cretaceous

volume

and Larson this volume)

Hell Creek Formation

Ott and Larson this

New genus & sp. (Ott

(latest

Late Cretaceous

Warm-hot,

coal

silicic, barite

caliches

organic rich

carbonates

rivers; hydromorphic to

and dry

(bonebed)

Campanian)

2 single

organics dering and anastomosed

seasonally wet

multiple

well drained soils

minor caliches;

alluvial plain with mean-

Subtropical,

single and

(middle

Late Cretaceous

Campanian)

tion (Coahuilla, Mexico)

volume

rivers; hydromorphic to

and dry

Campanian)

volume)

Cerro del Pueblo Forma-

Loewen et al. this

(Loewen et al. this

ern Utah)

New genus & sp.

Getty et al. this volume

Kaiparowits Fm (south-

et al. this volume)

organics

dering and anastomosed well drained soils

minor caliches;

alluvial plain with mean-

seasonally wet

Late Cretaceous

Subtropical,

coal

(middle

single

tropical, sea-

tian)

Kaiparowits Fm (south-

swamps

temperate-sub sonally wet

alluvial floodplain; peat

warm-

alluvial floodplain

(mid Maastrich-

seasonally wet

subtropical;

temperate-

warm

Late Cretaceous

single

single

(Alberta)

New genus & sp. (Getty

Getty et al. this volume

New genus & sp. (Getty

Late Cretaceous (Maastrichtian)

Horseshoe Canyon

ern Utah)

Wu et al. 2007

Eotriceratops xerinsularis

Lance (Wyoming)

et al. this volume)

Lull 1905

Diceratops hatcheri

no

no

no

no

no

yes

yes

no

no

446 Marsh 1891

Marsh 1889

Triceratops horridus

References

Torosaurus latus

Taxa

(Maastrichtian)

and dry

(Saskatchewan)

Frenchman

sonally wet

tian)

rado); Hell Creek (MonScollard (Alberta);

tropical, sea-

(late Maastrich-

ming); Laramie (Colotana, South Dakota);

temperate-sub

Late Cretaceous

Lance, Evanston (Wyo-

Mexico); Javelina (Texas)

(Utah); Kirtland (New

rado); North Horn

tana); Laramie (Colo-

warm-

sonally wet

South Dakota, Mon-

tropical, sea-

Creek (North Dakota,

temperate-sub

warm-

Climate

(Saskatchewan); Hell

multiple

multiple

Late Cretaceous

Lance (Wyoming); Frenchman

# of individuals

Age

Stratigraphy

Appendix 28.1. (Continued)

alluvial and coastal plain

alluvial & coastal plain

Paleoenvironment

coals/pyroclastics

Redbeds/caliches/

yes

no

Studies

Taphonomic

29 Behavioral Interpretations from Ceratopsid Bonebeds REBECCA K. HUNT AND ANDREW A. FARKE

monodominant ceratopsid bonebeds represent an

Introduction

important source of information for understanding ceratopsid paleobiology, and monodominant centrosaurine

Monodominant ceratopsid bonebeds (sensu Eberth et al.

bonebeds are often cited as evidence for herding be-

2007), where individuals of a ceratopsid species greatly out-

haviors and complex social hierarchies for Ceratopsidae.

number individuals of other vertebrate species, are immensely

However, caution should be used when applying such in-

useful for clarifying a number of paleobiological research

terpretations across the entire clade. Ceratopsid

questions (Brinkman et al. 2007), including those related to

bonebeds most commonly contain the remains of cen-

systematics, growth rates, behavior, and individual variation

trosaurines, with more than 20 sites representing 8 out of

(e.g., Sampson et al. 1997; Lehman 2006). It is not an exag-

10 taxa. Chasmosaurine bonebeds, in contrast, are rarer

geration to claim that these bonebeds have been central to the

and are currently known from only 8 sites representing

modern reinterpretation of ceratopsid paleobiology. The ta-

only 6 out of 14 species. Furthermore, the minimum

phonomy and sedimentology of these bonebeds have been

number of individuals tends to be greater in cen-

investigated in great depth over the past 20 years and provide

trosaurine bonebeds than in chasmosaurine bonebeds.

important insight about the circumstances leading to such

These differences may be attributed to collecting or

unusual accumulations (e.g., Currie and Dodson 1984; Rogers

taphonomic biases, or behavioral differences between the

1990; Ryan et al. 2001; Eberth and Getty 2005; Eberth et al.

two subfamilies. For instance, factors such as limited wa-

this volume). An important conclusion from the study of cera-

ter resources during drought may result in a concentrated

topsian bonebeds is the possibility that these deposits con-

death assemblage for a taxon that does not normally ex-

stitute evidence of gregarious, herding behavior. There is little

hibit gregarious behavior. Thus, an understanding of the

doubt that the animals preserved in these bonebeds were to-

depositional environments and taphonomic histories for

gether at the time of their death, and that the accumulations

ceratopsid bonebeds can help clarify the origins of large

are not time-averaged accumulations. But what do these ag-

taphocoenoses. When such factors are assessed within

gregations really signify? Do they represent a herd with a com-

ceratopsidae, there is positive evidence for non-stress-

plex social and behavioral hierarchy governed by displays of

related, gregarious behavior for only a few genera of cen-

frills and horns (e.g., Norman and Wellnhofer 1988; Sampson

trosaurines. We conclude that social behavior probably

1995b)? Did all ceratopsids herd (Fastovsky and Weishampel

varied greatly across ceratopsids.

2005), or was there intraspecific variation in these behaviors?

447

The idea of ceratopsids as herding animals is ingrained in

2001; Fiorillo and Gangloff 2003), and Pachyrhinosaurus n. sp.

the popular literature (e.g., Currie 1981; Norman and Welln-

(Wapiti Formation, Alberta; Tanke 1988; Ryan et al. 2006; Cur-

hofer 1988; Svarney and Barnes-Svarney 1999; Haines 2000),

rie et al. 2007). We note that the Achelousaurus horneri type

and to a lesser extent, the scientific literature (Currie and Dod-

locality includes at least two individuals (recognized by two

son 1984; Dodson 1996; Sampson 1997; Sampson 2001; Dod-

right squamosals, one as part of the holotype skull and a sec-

son et al. 2004). For example, Fastovsky and Weishampel

ond adult squamosal under the same catalog number, MOR

(2005: 173–174) claim:

485), although this has not been reported previously in the

At the very least, it can be claimed with much justification that many if not all ceratopsians lived in large herds, at least during part if not all of the year. This justification comes from our ever-increasing catalog of ceratopsian bonebeds.

A similar generalization appears in Dodson et al. (2004):

literature. Thus, monodominant bonebed accumulations are known for eight out of ten recognized centrosaurine species at more than 20 sites. These sites, containing estimated minimum numbers of individuals between two and several hundred (Eberth 1996, 1998; Eberth and Getty 2005; Eberth et al. this volume), are currently known only from Alaska, Alberta, and Montana.

Such remarkable concentrations of bones belonging to a single species suggest that ceratopsids congregated in large, gregarious accumulations.

Among chasmosaurines, relatively fewer taxa occur in bonebeds than is the case for centrosaurines (Fig. 29.1; Table 29.1), and chasmosaurine bonebeds are infrequently de-

How universal was gregarious behavior in ceratopsids? Most

scribed. Among the 14 taxa of chasmosaurines, only 6 are

statements regarding these behaviors are based on bonebeds

known from bonebeds: Agujaceratops mariscalensis, Aguja For-

dominated by centrosaurines rather than chasmosaurines.

mation, Texas (Lehman 1982); Anchiceratops ornatus, Horse-

Can such results be generalized across the entire ceratopsid

shoe Canyon Formation, Alberta (Sternberg 1926); Torosaurus

clade? Is there any evidence in the fossil record for behavioral

utahensis, North Horn Formation, Utah, and Javelina Forma-

differences between taxa? What is the significance of the fact

tion, Texas (Gilmore 1946; Hunt 2005); Triceratops sp., Hell

that centrosaurine and chasmosaurine bonebeds actually dif-

Creek Formation, Montana (Mathews et al. 2007); Kaiparo-

fer quite sharply in their relative abundance in the rock record

wits new taxon B, Kaiparowits Formation, Utah (Getty et

and the size of their bonebed assemblages (Table 29.1)?

al. 2007, this volume); and a new taxon from the Cerro del

Here we review the geological and paleontological data drawn from studies of centrosaurine and chasmosaurine

Pueblo Formation, Coahuila, Mexico (Eberth et al. 2004; Lund et al. 2007; Loewen et al. this volume).

bonebeds, and explore the possible significance of the differ-

Furthermore, chasmosaurine bonebeds, in contrast to those

ences between them. Specifically, we examine whether cen-

comprising centrosaurines, often preserve a smaller number

trosaurines congregated more often and in larger groups than

of individuals per site. A minimum of 3 individuals are pre-

chasmosaurines, or whether they inhabited environments

served at the type locality in the North Horn Formation, Utah,

that favored preservation of their remains in large numbers.

for Torosaurus utahensis (Gilmore 1946), and a minimum

We also suggest some new directions for future ceratopsian

number of 3 individuals are also preserved at a quarry in the

bonebed research.

Javelina Formation, Texas (Hunt 2006). Both of the undescribed chasmosaurine taxa (Kaiparowits taxon B and the

Taxonomic and Geographic Distributions

Cerro del Pueblo chasmosaurine taxon) are represented by a minimum of 3 individuals each. The single outlier among

Among centrosaurines, bonebed accumulations (Fig. 29.1;

chasmosaurines in this regard is represented by a bonebed of

Table 29.1) are known for Centrosaurus apertus (Dinosaur Park

Agujaceratops mariscalensis that preserves at least 20 indi-

Formation, Alberta; Currie and Dodson 1984; Ryan et al.

viduals (Lehman 2006). No data are available yet on the An-

2001; Eberth and Getty 2005; Eberth et al. this volume), Cen-

chiceratops ornatus bonebed, as it is awaiting further study.

trosaurus brinkmani (Oldman Formation, Alberta; Ryan and Russell 2005), Albertaceratops nesmoi ( Judith River Formation, Montana; Ryan 2007), Styracosaurus albertensis (Dinosaur Park

Depositional Environments

Formation, Alberta; Visser 1986; Sampson et al. 1997; Ryan

Do any differences exist between the depositional settings

2003), Achelousaurus horneri and Einiosaurus procurvicornis

for centrosaurine and chasmosaurine bonebeds? If so, what

(Two Medicine Formation, Montana; Rogers 1990; Sampson

might they reveal about the paleoecology of the two groups?

1995a, b), Pachyrhinosaurus canadensis (Horseshoe Canyon

Here, we divide bonebed depositional environments into

Formation and Saint Mary River Formation, Alberta, and

paleochannel (here termed ‘‘higher energy’’) and overbank

Prince Creek Formation, Alaska; Langston, Jr. 1975; Ryan et al.

(here referred to as ‘‘lower energy’’) settings. Although this

448 hunt & farke

Table 29.1. Summary of Bonebed Occurrences for Ceratopsid Dinosaurs by Subfamily Bonebed Taxon

occurrences

Depositional environments and MNI

Achelousaurus horneri

1

Overbank (MNI = 2)

Albertaceratops nesmoi

1

Unknown (MNI = 6+)

Avaceratops lammersi

0

Centrosaurinae

‘‘Brachyceratops montanus’’

1

Unknown (MNI = 5)

Centrosaurus apertus

5+

Paleochannel sandstones, overbank mudstones (MNI = 3, 3+, 4, 4+, 57)

Centrosaurus brinkmani

2

Splay mudstone (MNI = 5, 13)

Einiosaurus procurvicornis

2

Shallow lake in abandoned stream channel; floodplain waterhole (MNI = 7, 8)

Pachyrhinosaurus canadensis

2

Overbank, Paleochannel sandstones (MNI = 2, 15+)

Pachyrhinosaurus n.sp.

1

Overbank (MNI = 27+)

Styracosaurus albertensis

2

Paleochannel sandstone, IHS (MNI = 7+)

Styracosaurus ovatus

0

Chasmosaurinae Agujaceratops mariscalensis

Deltaic interdistributary marshes (MNI = 20)

Anchiceratops ornatus

1

Arrhinoceratops brachyops

0

Unknown

Cerro del Pueblo n.sp.

1

Chasmosaurus belli

0

Chasmosaurus irvinensis

0

Chasmosaurus russelli

0

Diceratops hatcheri

0

Eotriceratops xerinsularis

0

Pentaceratops sternbergi

0

Torosaurus latus

0

Torosaurus utahensis

2

Overbank (MNI = 3, 3)

Triceratops sp.

1

Overbank (MNI = 2)

Unknown (MNI = 2)

Data from: Gilmore 1914, 1946; Langston 1975; Rogers 1990; Ryan et al. 2001; Ryan 2003; Eberth and Getty 2005; Hunt 2005; Lehman 2006; Currie et al. 2007; Fanti and Currie 2007; Fiorillo et al. 2007; Lund et al. 2007; Mathews et al. 2007.

dual classification is simplistic, it appears to be a useful means

redeposited as a paleochannel-hosted bonebed. Bonebeds

of assessing preferred paleoenvironmental associations for

might also have been destroyed during reworking (Eberth and

these two groups of ceratopsids.

Getty 2005).

Nineteen centrosaurine ceratopsian bonebeds are known

The apparent non-random stratigraphic distribution of the

from the Dinosaur Park Formation within Dinosaur Provin-

Centrosaurus apertus bonebeds at Dinosaur Provincial Park and

cial Park (Alberta, Canada; Centrosaurus apertus and Styraco-

in the surrounding region (Eberth et al. this volume) suggests

saurus albertensis). The bonebeds were deposited across an im-

that many of these animals may have perished in at least 2

mense alluvial lowland, which was semi-constantly reworked

(and as many as 17) coastal-plain flooding events, and were, in

by paleochannel flow (Eberth 2005). Flows varied from high-

some cases, part of life assemblages that numbered in the high

energy and erosive to low-energy with standing water. Many

hundreds and, possibly, low thousands. Drowned individuals

paleochannels in the Dinosaur Park Formation overlie and

would have settled out on the floodplain or would have been

incise overbank-hosted bonebeds. This may indicate that

moved into the channels draining the lowlands (Eberth and

some material is sourced from overbank-hosted bonebeds and

Getty 2005). Other centrosaurine bonebeds are also known

Behavioral Interpretations from Ceratopsid Bonebeds 449

FIGURE 29.1.

Geographic distribution of chasmosaurine and centrosaurine ceratopsid bonebeds. (A) Pachyrhinosaurus canadensis, Prince Creek Formation, Alaska; (B) Pachyrhinosaurus n. sp., Wapiti Formation, Alberta; (C) Anchiceratops ornatus and Pachyrhinosaurus canadensis, Horseshoe Canyon Formation, Alberta; (D) Centrosaurus apertus and Styracosaurus albertensis, Dinosaur Park Formation; Centrosaurus brinkmani, Oldman Formation, Alberta; (E) Centrosaurus brinkmani, Oldman Formation, Alberta; (F) Einiosaurus procurvicornis and Achelousaurus horneri, Two Medicine Formation, Montana; (G) Albertaceratops nesmoi, Judith River Formation, Montana; (H) Triceratops sp., Hell Creek Formation, Montana; (I) Torosaurus utahensis, North Horn Formation, Utah; ( J) Kaiparowits New Taxon B, Kaiparowits Formation, Utah; (K) Agujaceratops mariscalensis, Aguja Formation; Torosaurus cf. utahensis, Javelina Formation, Texas; (L) Cerro del Pueblo new chasmosaurine, Cerro del Pueblo Formation, Coahuila, Mexico.

from low-energy settings (e.g., overbank/floodplain). These

are interpreted as having been deposited in shallow lakes and

include

waterholes on a floodplain after the animals succumbed to

Centrosaurus

brinkmani,

Albertaceratops

nesmoi,

Achelousaurus horneri, and Einiosaurus procurvicornis (Rogers

drought (Rogers 1990).

1990; Sampson 1995a, b; Sampson et al. 1997; Eberth and

Nearly all known chasmosaurine bonebeds (with the excep-

Getty 2005; Ryan 2007). In all of these overbank bonebeds,

tion of Kaiparowits new taxon B) are preserved in overbank/

evidence for hydraulic sorting is poor, specimens are disarticu-

floodplain settings. The Kaiparowits new taxon B bonebed

lated and fragmented, and the fossil assemblages are domi-

is known from the base of a sandy channel, and it is com-

nated by small broken specimens (Eberth and Getty 2005).

pletely disarticulated with characteristics commonly associ-

The presence of thoroughly disarticulated remains that are

ated with predepositional transportation (Getty et al. 2007,

poorly sorted and that exhibit green breaks, removal of tra-

this volume).

becular bone, and an absence of weathering indicate the Dino-

The first described Triceratops bonebed from the Hell Creek

saur Provincial Park centrosaurine carcasses were deposited in

Formation of Montana was discovered in 2006. This bonebed

standing water and were trampled and scavenged in saturated

is especially unusual considering the comparative abun-

to damp conditions until final burial (Eberth and Getty 2005).

dance of isolated individuals for which Triceratops remains

Centrosaurine bonebeds within the Two Medicine Formation

are known within the Hell Creek and Lance formations. The

450 hunt & farke

Triceratops bonebed is located in an overbank-hosted mud-

during initial floodplain draining (Eberth and Getty 2005).

stone and contains the remains of at least three individuals

Ultimately, the relative scarcity of data on chasmosaurine

(Mathews et al. 2007). The remains are disarticulated, but

bonebeds hinders a more complete interpretation of paleo-

show no orientation, size or shape bias due to hydraulic in-

environmental preferences at this time.

fluences. The only taphonomic modifications at this site are green fractures and very mild weathering (Mathews et al.

Social Behaviors

2007; Mathews pers. com. 2007). In Big Bend National Park, Texas, the chasmosaurine Agu-

The popular image of great herds of ceratopsids, reminiscent

jaceratops mariscalensis is preserved within the low-energy

of the herds of bison that used to cover portions of the Ameri-

coastal marsh and swamp facies that would have been depos-

can West, has been spurred by bonebed discoveries. As de-

ited in the deltaic interdistributary marshes adjacent to the

scribed above, this idea has permeated both the popular and

Late Cretaceous shoreline (Lehman 1982, 1989). A single

scientific literature. With this image of ‘‘herding’’ ceratop-

bonebed, containing at least 20 individuals, along with the

sians, it is an easy leap to envision a complex social hierarchy,

solitary remains of other Agujaceratops specimens repeatedly

built around visual displays from the elaborate horns and frills

found associated with swamp and marsh facies, suggest that

(e.g., Sampson 1995b). Yet, as is often cautioned (see especially

this was either their preferred habitat or where they com-

Lehman 2006), social behavior is extremely difficult to infer

monly perished (Lehman 2006). The Agujaceratops bonebed

from bonebeds alone. The presence of a bonebed does not

contains thoroughly disarticulated and well sorted remains,

automatically indicate that the preserved individuals lived to-

and the only taphonomic modifications present at the site are

gether in social aggregations, or even died together (Voorhies

green fractures (Lehman 1982) suggesting trampling and/or

1985).

scavenging.

A number of factors—such as a concentration of water re-

A markedly different bonebed occurs in the Javelina Forma-

sources during a drought, fire, breeding seasons or an abun-

tion of Big Bend National Park, containing the remains of at

dance of a specific food resource—may result in high popu-

least three Torosaurus cf. T. utahensis. The high organic matter

lation densities of otherwise nonherding animals. This fact

at this bonebed, ferroan calcite nodules, and coatings that

has been recognized by other workers (e.g., Rogers 1990; Leh-

occur on the bones in the mudstone suggest water-saturated

man 2006; Rogers and Kidwell 2007), but apparently has not

reducing conditions existed, rather than the typical, well-

been appreciated fully for ceratopsids. To what extent can the

drained and oxidized paleoenvironments of Javelina flood-

known ceratopsid bonebeds listed above be attributed to these

plains (Lehman 1990; Hunt 2005). The thoroughly disarticu-

phenomena?

lated and hydraulically well-sorted remains at this bonebed

Monodominant Einiosaurus assemblages in the overbank

contain green breaks, and exhibit mild weathering and a

facies of the Two Medicine Formation of Montana have been

loss of trabecular bone on poorly ossified articular surfaces on

attributed to attritional drought accumulations (Rogers 1990;

several of the limb bones (Hunt 2005; Hunt and Lehman in

Sampson 1995a, b). In contrast, paleoenvironmental data for

press). The type specimen of Torosaurus utahensis is known

the many centrosaurine bonebeds known from the Dinosaur

from the North Horn Formation of Utah, along with the re-

Park Formation of Alberta, Canada, exclude drought as a

mains of two other individuals. This bonebed appears to be

mechanism for bonebed accumulations (Currie and Dodson

hosted by overbank-hosted mudstones; green fractures are the

1984; Eberth and Getty 2005; Eberth et al. this volume). The

only known taphonomic modifications present (Gilmore

Centrosaurus apertus bonebeds at Dinosaur Provincial Park and

1946). Few geological data are available for this site, however,

the surrounding region are attributed, instead, to coastal-

and restudy is necessary to develop a more complete picture of

plain-scale flood-induced drowning events (Eberth and Getty

the taphonomy and paleoenvironment.

2005; Eberth et al. this volume). Similarly, the monodominant

In all, there appears to be little evidence for differences in

Agujaceratops assemblage of the Aguja Formation of Texas does

gross depositional environment between centrosaurine and

not appear to fit a drought-induced bonebed formation sce-

chasmosaurine bonebeds. It is possible that the associated

nario (Lehman 2006). Thus, some sort of ‘‘grouping’’ behavior

lowland paleoenvironments simply reflect the preferred habi-

(whether as a herd, breeding assemblage, or simply an ‘‘in-

tat for both taxa, particularly centrosaurines, at least for a por-

festation’’) in the absence of drought is likely for these taxa

tion of the year (Brinkman et al. 1998; Eberth and Getty 2005;

(Ryan et al. 2001).

Lehman 2006). Although some centrosaurine bonebeds are

The relative absence of bonebeds for Chasmosaurus, which

associated with relatively high-energy channel accumula-

co-existed with Centrosaurus (Ryan and Evans 2005), may pro-

tions, these probably represent bones reworked from low-

vide a clue to the habits of the former animal. The relative

energy deposits, or remains that were hydraulically moved

absence of bonebeds containing Chasmosaurus may indicate a

Behavioral Interpretations from Ceratopsid Bonebeds 451

lack of group-behavior in Chasmosaurus, or at least a lack of

ences do exist between some centrosaurine sites; Eberth and

such behavior when it inhabited the regions also occupied by

Getty 2005). Juveniles, subadults, and adults are known from

Centrosaurus. As noted by Eberth and Getty (2005), the low-

most of the bonebeds consisting of either clade (Lehman

gradient coastal and alluvial lowlands of Dinosaur Provin-

1982, 2006; Currie and Dodson 1984; Sampson 1995a, b;

cial Park during the Late Cretaceous would have left no high

Ryan and Russell 2005; Hunt 2006; Mathews et al. 2007; Ryan

ground protection for terrestrial vertebrates in the case of a

2007; Hunt and Lehman in press). Bonebeds preserving only

flooding event. The absence of Chasmosaurus bonebeds may

subadult material are also known from the ‘‘Brachyceratops’’

indicate a lack or smaller capacity of herding or breeding as-

and Triceratops bonebeds (Gilmore, 1917; Mathews et al.

semblages present in their social structure. A similar pattern

2007). The presence of juvenile and subadult with adult re-

may also explain the rarity of hadrosaur bonebeds in Dinosaur

mains in bonebed accumulations demonstrate that some pos-

Provincial Park (Ryan et al. 2001). Conversely, the difference

sible family ‘‘structure’’ may have existed for some duration

in preservational modes between centrosaurines and Chasmo-

of time.

saurus could be argued as additional evidence in favor of the reality of herding behavior in Centrosaurus.

Discussion

The differences in number of individuals present in the Agujaceratops versus Centrosaurus-Styracosaurus bonebeds may

Based on the preceding data and inferences, we propose that

represent interspecific differences in group-living behavior,

evidence for non-stress-related, group-living behavior is pres-

or a difference in the factors that acted to concentrate the

ent for only a few taxa, and cannot be extended with confi-

remains of these animals in a death assemblage. Equally plau-

dence to other taxa. As recognized by other authors (Currie

sibly, differences in local paleoenvironmental and paleoeco-

and Dodson 1984; Visser 1986; Ryan et al. 2001; Eberth and

logical factors may have resulted in these numerical differ-

Getty 2005; Fiorillo et al. this volume), bonebed assemblages

ences. For example, population densities for ceratopsids as a

of Centrosaurus, Styracosaurus, and Pachyrhinosaurus probably

whole may have varied between these regions because of vari-

represent catastrophic death events affecting a large aggregate

ations in food and other resource requirements. In this con-

of individuals. We posit (in agreement with these authors)

text the major climatic and other paleoenvironmental dif-

that the bonebeds for these taxa may indeed capture a ‘‘herd’’

ferences that existed between northern environments (e.g.,

of animals. Although these groupings may have been per-

Alaska to Montana) versus more southern environments (e.g.,

manent or temporary, it seems unlikely that they formed

Utah to Mexico) likely impacted the carrying capacity of

purely in response to environmental stresses. For bonebeds

these regions, and thus influenced group size and access (year-

apparently associated with drought accumulations, such

round, seasonal, absent). Such variable paleoenvironmental

as Einiosaurus, it is quite easy to posit that these group-

conditions may ultimately explain why most southern chas-

associations were due to congregation around a water re-

mosaurine bonebeds include considerably fewer individuals

source, rather than any long-term social behavior. Thus, gre-

than the more northern centrosaurine bonebeds.

garious herds for these taxa are not well-supported based on

A further consideration relating to paleoenvironmental differences is that differential access to any region by different

the fossil evidence (although it does not conclusively rule out aggregation behavior).

ceratopsid groups may have resulted in different degrees of

The evidence for herding/social aggregation in Agujacera-

mortality and preservation of the taxa that visited those areas.

tops is also equivocal. There is no evidence in these deposits for

In this context, bonebeds may simply reflect seasonally vari-

specific environmental stresses that may have acted to force

able mortality and preservation potential due to floods or

the assembly of this group. Alternatively, it is also conceivable

storm activity, etc. Along these lines, absolute differences in

that the bonebed is an attritional assemblage reflecting a lo-

depositional environments inhabited regularly by centro-

calized preference for habitat or food (Lehman 2006) rather

saurines versus chasmosaurines (e.g., the Hell Creek Forma-

than the relatively rapid demise of a ‘‘herd’’ due to flood or

tion versus the Dinosaur Park Formation) could also explain

drought. Without a more thorough geological knowledge of

the disparities in these bonebeds.

the few chasmosaurine bonebeds, it will remain difficult to

It is also important to recognize that not all taxa may have

determine if the small number of chasmosaurine bonebeds

exhibited the same behaviors. While one group may have ex-

relative to centrosaurine bonebeds represents behavioral dif-

hibited at least temporary herding behavior, as is possible for

ferences between the taxa or an artifact of preservation.

Centrosaurus apertus, it might have been different for other

Group accumulations have also been noted for Psittacosau-

taxa (e.g., Pentaceratops; Brinkman et al. 1998; Eberth and

rus, a basal ceratopsian from Asia (Meng et al. 2004; Qi et

Getty 2005). However, with this said, striking differences be-

al. 2007), Prenoceratops, a basal neoceratopsian from Mon-

tween the ontogenetic compositions of centrosaurine and

tana (Chinnery 2004), and Leptoceratops, a basal neocera-

chasmosaurine bonebeds appear to be absent (although differ-

topsian from Alberta (Sternberg 1951). Four immature in-

452 hunt & farke

dividuals are preserved within this Two Medicine Formation

son, D. Eberth, J. Foster, M. Getty, W. Hammer, J. Kirkland,

Prenoceratops bonebed (Chinnery 2004). Two individuals of

W. Langston, Jr., M. Loewen, R. Rogers, M. Ryan, S. Samp-

Leptoceratops are preserved in the Edmonton Formation bone-

son, and D. Tanke were helpful in formulating the ideas

bed described by Sternberg (1951). Most recently, Qi et al.

and arguments presented here. The reviews of D. Eberth and

(2007) proposed that an accumulation of juvenile psittaco-

M. Getty greatly improved this manuscript. We especially

saurs represents a social aggregation of individuals from suc-

thank J. Mathews and S. Williams for sharing preliminary data

cessive annual clutches. This would contrast with the rather

on the Triceratops bonebed. This work was supported in part

frequent occurrence of aggregations between juveniles of the

by an NSF Graduate Research Fellowship to AAF.

same clutch among modern animals. We do not dispute the fact that the psittacosaurs probably were forming an aggregation. We do, however, question the claim that the juveniles were from disparate clutches. Qi et al. (2007) based their claim of multiple age classes on the distribution of femoral lengths, and by inference body masses, in the psittacosaur specimens. We posit that the range of femoral lengths (the minimum and maximum values differ only by 15% from the mean length) is compatible with a single age class, and suggest that it may fall within the bounds seen for modern associations of juvenile animals. Histological information is necessary to verify any claims on the relative age of the specimens.

Conclusions Clearly, much additional work is necessary in order to more fully understand the significance of the differences and similarities between chasmosaurine and centrosaurine bonebeds. At present, chasmosaurine bonebeds are not nearly as well studied or characterized as centrosaurine bonebeds. In particular, more complete sedimentological, taphonomic, and paleoenvironmental interpretations are needed. Future discoveries will certainly help indicate if the known differences in centrosaurine and chasmosaurine bonebed frequencies are real or simply an artifact of sampling. Based on known bonebed occurrences, some ceratopsid species apparently gathered in groups at least some of the time, but there is no evidence that most or all species gathered in groups all of the time, as suggested by some authors (e.g., Fastovsky and Weishampel 2005). As noted by Lehman (2006) and Voorhies (1985), presence of a monodominant bonebed assemblage does not necessary indicate herding behavior. Aggregations can form for a variety of different reasons— resource availability, breeding, protection against predators, and so forth. It is a beautiful, but frustrating, realization that ceratopsid behavior was likely quite diverse, but many aspects of ceratopsid behavior, including some instances of aggregation behavior, are ultimately unknowable. Acknowledgments

We thank the conveners of the Ceratopsian Symposium for the opportunity to prepare this paper for inclusion in this volume. Discussions with B. Chinnery-Allgeier, D. Boyer, P. Dod-

References Cited Brinkman, D. B., D. A. Eberth, and P. J. Currie. 2007. From bonebeds to paleobiology: Applications of bonebed data. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 221–263. Chicago: University of Chicago Press. Brinkman, D. B., M. J. Ryan, and D. A. Eberth. 1998. The palaeogeographic and stratigraphic distribution of ceratopsids (Ornithischia) in the Upper Judith River Group of Western Canada. Palaios 13: 160–169. Chinnery, B. 2004. Description of Prenoceratops peiganenis gen.et sp. nov. (Dinosauria, Neoceratopsia) from the Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 24: 572–590. Currie, P. J. 1981. Hunting dinosaurs in Alberta’s great bonebed. Canadian Geographic 101: 34–39. Currie, P. J., and P. Dodson. 1984. Mass death of a herd of ceratopsian dinosaurs. In W. E. Reif and F. Westphal, eds., Third Symposium on Mesozoic Terrestrial Ecosystems, pp. 61–66. Tubingen: Attemto Verlag. Currie, P. J., W. Langston Jr., and D. H. Tanke. 2007. A new Pachyrhinosaurus from the Wapiti Formation of Grand Prairie, Alberta. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, p. 22. Drumheller: Royal Tyrrell Museum of Palaeontology. Dodson, P. 1996. The Horned Dinosaurs. Princeton: Princeton University Press. Dodson, P., C. A. Foster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Eberth, D. A. 1996. Ceratopsian bonebeds in the Dinosaur Park Formation (Campanian) of southern Alberta: Bigger than we thought? Journal of Vertebrate Paleontology 16(3, Suppl.): 32A. ———. 1998. Clustered ceratopsian bonebeds, southern Alberta, Canada: Primary evidence for the size of ceratopsian-herd death assemblages. In D. L. Wolberg, K. Gittis, S. Miller, L. Carey and A. Raynor, eds., The Dinofest Symposium, pp. 13. Philadelphia: Academy of Natural Sciences. ———. 2005. The geology. In P. J. Currie and E.B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 54–82. Bloomington: Indiana University Press. Eberth, D. A., D. B. Brinkman, and V. Barkas. 2010. A centrosaurine mega-bonebed from the Upper Cretaceous of southern Alberta: Implications for behavior and death events. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New

Behavioral Interpretations from Ceratopsid Bonebeds 453

Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 495–508. Bloomington: Indiana University Press. Eberth, D. A., C. R. Delgado-de Jesús, J. F. Lerbekmo, D. B. Brinkman, R. A. Rodríguez-de la Rosa, and S. D. Sampson. 2004. Cerro del Pueblo Formation (Difunta Group, Upper Cretaceous), Parras Basin, southern Coahuila, Mexico: Reference sections, age, and correlation. Revista Mexicana de Ciencias Geológicos 21: 335–352. Eberth, D. A., and M. A. Getty. 2005. Ceratopsian bonebeds: Occurrence, origins, and significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 501–536. Bloomington: Indiana University Press. Eberth, D. A., M. Shannon, and B. G. Noland. 2007. A bonebeds database: Classification, biases, and patterns of occurrence. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 103–219. Chicago: University of Chicago Press. Fanti, F., and P. J. Currie. 2007. A new Pachyrhinosaurus bonebed from the late Cretaceous Wapiti Formation. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 39–43. Drumheller: Royal Tyrrell Museum of Palaeontology. Fastovsky D. E., and D. B. Weishampel. 2005. The Evolution and Extinction of the Dinosaurs. Cambridge: Cambridge University Press. Fiorillo, A. R., and R. A. Gangloff. 2003. Preliminary notes on the taphonomic and paleoecologic setting of a Pachyrhinosaurus bonebed in northern Alaska. Journal of Vertebrate Paleontology 23(3, Suppl.): 50A. Fiorillo, A. R., P. J. McCarthy, E. Brandlen, P. P. Flaig, D. Norton, L. Jacobs, P. Zippi, and R. A. Gangloff. 2007. Paleontology, sedimentology, paleopedology, and palynology of the KikakTegoseak Quarry (Prince Creek Formation: Late Cretaceous), northern Alaska. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 48–49. Drumheller: Royal Tyrrell Museum of Palaeontology. Fiorillo, A. R., P. J. McCarthy, P. P Flaig, E. Brandlen, D. W. Norton, P. Zippi, L. Jacobs, and R. A. Gangloff. 2010. Paleontology and paleoenvironmental interpretation of the Kikak-Tegoseak Quarry (Prince Creek Formation: Late Cretaceous), northern Alaska: A multi-disciplinary study of a high-latitude ceratopsian dinosaur bonebed. In M. J. Ryan, B. J. ChinneryAllgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 456–477. Bloomington: Indiana University Press. Getty, M. A., M. A. Loewen, A. L. Titus, and S. D. Sampson. 2007. Ceratopsid taphonomy from the Kaiparowits Formation, Grand Staircase-Escalante National Monument, Utah. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 56–60. Drumheller: Royal Tyrrell Museum of Palaeontology. Getty, M. A., M. A. Loewen, E. Roberts, A. L. Titus, and S. D. Sampson. 2010. Taphonomy of horned dinosaurs (Ornithischia: Ceratopsidae) from the Late Campanian Kaiparowits

454 hunt & farke

Formation, Grand Staircase–Escalante National Monument, Utah. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 478–494. Bloomington: Indiana University Press. Gilmore, C. W. 1914. A new ceratopsian dinosaur from the Upper Cretaceous of Montana, with note on Hypacrosaurus. Smithsonian Miscellaneous Collections 63: 1–10. ———. 1917. Brachyceratops: A ceratopsian dinosaur from the Two Medicine Formation of Montana. U. S. Geological Society Professional Paper 103. ———. 1946. Reptilian fauna of the North Horn Formation of central Utah. U.S. Geological Society Professional Paper 210-C: 29–53. Haines, T. 2000. Walking with Dinosaurs. New York: Dorling Kindersley Publishing, Inc. Hunt, R. K. 2005. Ceratopsid dinosaurs from the Javelina Formation (Maastrichtian), Big Bend National Park, Texas. M.S. thesis. Texas Tech University, Lubbock. ———. 2006. The taphonomy of a chasmosaurine dinosaur bone bed from the Javelina Formation (Maastrichtian) of Big Bend National Park, Texas. Geological Society of America Abstracts with Programs 38: 62. Hunt, R. K., and T. M. Lehman. In press. Attributes of the ceratopsian dinosaur Torosaurus, and new material from the Javelina Formation (Maastrichtian) of Texas. Journal of Paleontology 82. Langston, W. L., Jr. 1975. The ceratopsian dinosaurs and associated lower vertebrates from the St. Mary River Formation (Maastrichtian) at Scabby Butte, southern Alberta. Canadian Journal of Earth Sciences 12: 1576–1608. Lehman, T. M. 1982. A ceratopsian bone bed from the Aguja Formation (Upper Cretaceous), Big Bend National Park, Texas. M.Sc. thesis, University of Texas, Austin. ———. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. ———. 1990. Upper Cretaceous (Maastrichtian) paleosols in TransPecos Texas. Geological Society of America Bulletin 101: 188–203. ———. 2006. Growth and population age structure in the horned dinosaur Chasmosaurus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 259–317. Bloomington: Indiana University Press. Loewen, M. A., S. D. Sampson, E. K. Lund, A. A. Farke, M. C. Aguillón-Martínez, C. A. de Leon, R. A. Rodríguez-de la Rosa, M. A. Getty, and D. A. Eberth. 2010. Horned dinosaurs (Ornithischia: Ceratopsidae) from the Upper Cretaceous (Campanian) Cerro del Pueblo Formation, Coahuila, Mexico. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 99–116. Bloomington: Indiana University Press. Lund, E. K., M. A. Loewen, S. D. Sampson, M. A. Getty, M. C. Aguillón Maertinez, R. A. Rodriguez de la Rosa, and D. A. Eberth. 2007. Ceratopsian remains from the Late Cretaceous Cerro del Pueblo Formation, Coahuila, Mexico. In D. R. Bra-

man, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 108–113. Drumheller: Royal Tyrrell Museum of Palaeontology. Mathews, J. C., M. Henderson, and S. Williams. 2007. Taphonomy, sedimentology, and paleoenvironmental reconstructions of a unique Triceratops site in the Hell Creek, southeastern Montana. Geological Society of America Abstracts with Programs 39: 9. Meng, Q., J. Liu, D. J. Varricchio, T. Huang, and C. Gao. 2004. Parental care in an ornithischian dinosaur. Nature 431: 145– 146. Norman, D., and P. Wellnhofer. 1988. Illustrated Encyclopedia of Dinosaurs. Baltimore: Salamander Books. Qi, Z., P. M. Barrett, and D. A. Eberth. 2007. Social behavior and mass mortality in the basal ceratopsian dinosaur Psittacosaurus (Early Cretaceous, People’s Republic of China). Palaeontology 50: 1023–1029. Rogers, R. R. 1990. Taphonomy of three dinosaur bone beds in the Upper Cretaceous Two Medicine Formation of northwestern Montana: Evidence for drought-related mortality. Palaios 5: 394–413. Rogers, R. R., and S. M. Kidwell. 2007. A conceptual framework for the genesis and analysis of vertebrate skeletal concentrations. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 1–63. Chicago: University of Chicago Press. Ryan, M. J. 2003. Taxonomy, systematics, and evolution of centrosaurine ceratopsids of the Campanian Western Interior Basin of North America. Ph.D. diss., University of Calgary, Alberta. ———. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, Southeastern Alberta. Journal of Paleontology 81: 376–396. Ryan, M. J., and D. C. Evans. 2005. Ornithischian dinosaurs. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 312–348. Bloomington: Indiana University Press. Ryan, M. J., and A. P. Russell. 2005. A new centrosaurine ceratopsid from the Oldman Formation of Alberta and its implications for centrosaurine taxonomy and systematics. Canadian Journal of Earth Sciences 42: 1369–1387. Ryan, M. J., and A. P. Russell, P. J. Currie, and D. A. Eberth. 2001. Taphonomy of a Centrosaurus (Dinosauria: Ornithischia) bone bed from the Dinosaur Park Formation (Campanian) of Al-

berta, Canada with comments on cranial ontogeny. Palaios 16: 482–506. Ryan, M. J., D. Tanke, D. Brinkman, D. A. Eberth, and P. J. Currie. 2006. A new Pachyrhinosaur-like ceratopsian from the Upper Dinosaur Park Formation (Late Campanian) of southern Alberta, Canada. Journal of Vertebrate Paleontology 26(3, Suppl.): 117A. Sampson, S. D. 1995a. Two new horned dinosaurs from the Upper Cretaceous Two Medicine Formation of Montana, with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15: 743–760. ———. 1995b. Horns, herds, and hierarchies. Natural History 104: 36–40. ———. 1997. Bizarre structures and dinosaur evolution. In D. L. Wolberg, E. Stump, and G. D. Rosenberg, eds., DinoFest International: Proceedings of a Symposium Held at Arizona State University, pp. 39–45. Philadelphia: Academy of Natural Sciences. ———. 2001. Speculations on the socioecology of ceratopsid dinosaurs (Ornithischia: Neoceratopsia). In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 263–276. Bloomington: Indiana University Press. Sampson, S. D., M. J. Ryan, and D. H. Tanke. 1997. Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 121: 293–337. Sternberg, C. M. 1926. Notes on the Edmonton Formation of Alberta. Canadian Field-Naturalist 40: 102–104. ———. 1951. Complete skeleton of Leptoceratops gracilis Brown from the Upper Edmonton Member on Red Deer River, Alberta. National Museum of Canada Bulletin 123: 225–255. Svarney, T. E., and P. Barnes-Svarney. 1999. The Handy Dinosaur Answer Book. Canton: Visible Ink Press. Tanke, D. H. 1988. Ontogeny and dimorphism in Pachyrhinosaurus (Reptilia: Ceratopsidae), Pipestone Creek, N.W. Alberta, Canada. Journal of Vertebrate Paleontology 8(3, Suppl.): 41A. Visser, J. 1986. Sedimentology and taphonomy of a Styracosaurus bonebed in the Late Cretaceous Judith River Formation, Dinosaur Provincial Park, Alberta. M.Sc. thesis, University of Calgary, Alberta. Voorhies, M. R. 1985. A Miocene rhinoceros herd buried in a volcanic ash. National Geographic Society Research Reports 19: 671–688.

Behavioral Interpretations from Ceratopsid Bonebeds 455

30 Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry (Prince Creek Formation: Late Cretaceous), Northern Alaska: A Multi-Disciplinary Study of a High-Latitude Ceratopsian Dinosaur Bonebed A N T H O N Y R . F I O R I L L O , PA U L J . M C C A R T H Y, P E T E R P. F L A I G , E R I K B R A N D L E N , D AV I D W. N O R T O N , P I E R R E Z I P P I , L O U I S J A C O B S , A N D R O L A N D A . G A N G L O F F

the northern part of Alaska contains a number of

the understory likely comprised ferns and smaller angio-

Cretaceous dinosaur localities that are situated along the

sperms. Overlapping age ranges for some of the paly-

Colville River and hosted by the Prince Creek Formation

nomorphs indicate an earliest Maastrichtian age for

(Campanian to Maastrichtian), a rock unit comprised

the site.

largely of alluvial/coastal plain sediments shed off the

In combination with sedimentological data, the fossil

rising Brooks Range to the south. Dinosaur sites include

assemblage suggests that the depositional environment

bonebeds like the well-known Liscomb Bonebed, and a

of the Kikak-Tegoseak Quarry was that of a low-energy al-

more recently discovered dinosaur bonebed, the Kikak-

luvial/coastal plain. In this area the floodplain was wet,

Tegoseak Quarry.

but water levels probably fluctuated from shallow stand-

Extensive excavation of the Kikak-Tegoseak Quarry has shown that it contains a preponderance of remains of the centrosaurine dinosaur Pachyrhinosaurus sp. Excava-

ing water to dry and subaerially exposed, possibly on a seasonal basis. It is inferred here that the Kikak-Tegoseak Quarry as-

tion of a 5 m by 4 m area within the bonebed has yielded

semblage succumbed to a flood event that captured the

the skeletal remains of at least nine individuals (deter-

remains of several taxa of dinosaurs. This event likely re-

mined from a count of occipital condyles). Based on the

sulted from strong seasonal runoff (e.g., snowmelt) from

nearly adult size of the condyles, but incomplete fusion

the rising Brooks Range presenting a substantial hazard

of other cranial elements, all these individuals appear to

to dinosaurs all along the ancestral North Slope.

have been subadults. Other taxa documented at the Kikak-Tegoseak Quarry include osteichthyan fishes and other dinosaurs including dromaeosaurs, tyrannosaurids, ornithomimids, and hadrosaurs.

Introduction Evolutionary and paleoecological inferences about dinosaurs

An abundant and diverse assemblage of pollen, spores,

have long tended to be derived from localities at middle and

algae, amber, and microscopic plant debris was recovered

lower latitudes, rather than from high latitudes. Given that

from the quarry site and indicates that the vegetation in

modern high-latitude ecosystems are valued for the insights

the area at or near the time of burial consisted of both co-

they provide into ecosystem structure and function in gen-

niferous forests (across the coastal plain) and broad-

eral, as well as global climate dynamics (Pielou 1994; Dowde-

leafed deciduous forests (in riparian areas). In all areas,

swell and Hambrey 2002), it makes sense that we may be able

456

to gain similar insights into ancient terrestrial ecosystems and

Subsequent work in 1998–2002 by joint Museum of Nature

their fossil vertebrates by examining ancient examples that

and Science/University of Alaska field parties, using tradi-

occurred in the high latitudes. In the Early Cretaceous of Aus-

tional excavation methods, confirmed the presence of a rich

tralia, for example, it has been suggested that large orbits in

monodominant bonebed dominated by ceratopsian remains

the hypsilophodontid Leaellynasaura amicagraphica were an

(Fiorillo and Gangloff 2003). With the support of the United

adaptation for life under the low-angle light conditions of the

States Army in 2002, excavation at the site was undertaken in

southern polar region (Rich and Rich 1989; Vickers-Rich et al.

earnest, yielding approximately two tons of jacketed mate-

1999). Similarly, as an explanation for the remarkably high

rial during that field season. As with most localities that are

abundance of isolated teeth of Troodon formosus, the large or-

broadly exposed to the effects of weathering, partially ex-

bits of the small theropod were suggested to be an adaptation

posed to shallowly buried specimens proved to be highly frag-

for the low-angle light conditions of the northern polar re-

mented, likely as the result of a combination of modern

gions (Fiorillo and Gangloff 2000).

weathering processes and the freeze-thaw activity that results

The Late Cretaceous (Campanian-Maastrichtian) Prince

from proximity to the active permafrost layer.

Creek Formation contains the densest concentrations of dino-

Accordingly, the excavation plan was revised and successful

saur bones of any high-latitude location in the northern or

excavations were begun in 2005, lasting through the 2007

southern hemispheres (Rich et al. 1997, 2002), and the spe-

field season. During this phase of operation, approximately

cifics of this rich record have come to light in recent years

4.5 tons of jacketed fossils and matrix were removed from the

(Davies 1987; Parrish et al. 1987; Fiorillo and Gangloff 2000,

Kikak-Tegoseak Quarry using a Bell 206 Jet Ranger equipped

2001; Gangloff et al. 2005; Fiorillo 2008, in press). However,

with a sling. These materials are curated at the Museum of

regardless of specimen richness, the taxonomic diversity of

Nature and Science where they are being prepared by Museum

fossil vertebrate faunas from the Prince Creek Formation is

staff and volunteers.

quite modest compared to faunas of similar age from lower latitudes (Fiorillo and Gangloff 2000; Fiorillo 2004). In addition to their rich specimen but low taxonomic-

Institutional Abbreviations. AK: University of Alaska Museum of the North, Fairbanks; DMNH: Museum of Nature and Science, Dallas.

diversity yields, these fossil localities occur not only at high latitudes today, but, given the tectonic history of North America, the Prince Creek Formation localities were deposited

Geologic Setting

at latitudes as high or somewhat higher than their current

The Kikak-Tegoseak Quarry is located near the confluence of

geographic positions (Witte et al. 1987). Thus, the Prince

Kikak Creek and the Colville River (Fig. 30.2). Precise locality

Creek Formation represents the gateway that connected the

data gathered by GPS are on file with the authors and the

faunas and floras of the latest Cretaceous of Asia to those in

Bureau of Land Management in Anchorage, Alaska.

the latest Cretaceous of North America (Fiorillo in press).

The site is stratigraphically situated within the Prince Creek

Despite their paleontological and paleogeographic signifi-

Formation. Mull and others (2003) have recently revised the

cance, there are few published paleoenvironmental studies

Cretaceous and Tertiary stratigraphic nomenclature in the

of these sites that combine paleontology, sedimentology, pa-

Colville Basin of the North Slope of Alaska, restricting the

leopedology, and palynology. Here we integrate these aspects

Prince Creek Formation to units overlying, and interfinger-

with respect to one quarry, the Kikak-Tegoseak Quarry (Fio-

ing with, the marine Schrader Bluff Formation (Fig. 30.3). Pre-

rillo and Gangloff 2003).

vious stratigraphic nomenclature (Chapman et al. 1964) de-

The Kikak-Tegoseak Quarry (Figs. 30.1, 30.2) was discovered

fined two tongues of the Prince Creek Fm., the Tuluvak

in the mid-1990s by a University of Alaska field crew engaged

Tongue, a lower, coarser-grained tongue generally below and

in paleontological surveys along the Colville River between

interfingering with the Schrader Bluff Formation, and the

Umiat and Nuiqsut. In 1994, investigators found pieces of fos-

Kogosukruk Tongue, a finer-grained tongue mostly above the

sil bone ‘‘float’’ along the banks of the river at the base of

Schrader Bluff Formation. In terms of this older nomencla-

a bluff with relief in excess of 100 m, but could only deter-

ture, the Kikak-Tegoseak Quarry is situated within the Kogosu-

mine that bones were eroding from sedimentary layers more

kruk Tongue. Unconformably overlying the Prince Creek For-

than 90 m above the base of the bluff. The bluffs along this

mation in the study area is the Pliocene to Holocene Gubik

stretch of river are largely unconsolidated and subject to par-

Formation. Although its current latitude is 70\ N, paleomag-

ticularly active hydrologic and thermal erosion, resulting in

netic studies of the Cretaceous rocks of this region suggest

a thick weathered zone along the face. In 1997, Ron Mancil,

that it was located 67\ to 85\ N during the Late Cretaceous

a University of Alaska student, and D. W. Norton located

(Witte et al. 1987; Besse and Courtillot 1991).

the in situ source of the fossils in the uppermost 3 m of the bluff.

The Prince Creek Formation crops out nearly continuously for approximately 72 km along the north-south course of the

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 457

FIGURE 30.1.

View looking north at the Kikak-Tegoseak Quarry. Quarry in the foreground.

Colville River. It contains more than 20 rhyolitic tuff beds

In the Ocean Point area pollen work suggests that the age of

that range in thickness from 0.25 to 6 m and are distributed

the vertebrate fossil deposits are close to the Campanian-

through 400 m of the unit (Brosge et al. 1966; Moore et al.

Maastrichtian boundary (Frederiksen 1990). Our pollen anal-

1994). Conrad and others (1990) have published potassium-

yses (see below) of Prince Creek Formation sediments approxi-

argon and argon-argon dates from tuffs at Ocean Point (Fig.

mately 50 km up river (south) of Ocean Point, in the area of

30.2) which range from approximately 68 to 71 Ma, with a

Kikak Creek, suggest an earliest Maastrichtian age for the mass

best age estimate of 69.1 +/– 0.3 Ma.

accumulation of horned dinosaur bones and fossil vertebrates

Previous biostratigraphic work roughly places the Maas-

recovered from the Kikak-Tegoseak Quarry.

trichtian/Paleocene boundary as occurring along the Colville

The Prince Creek Formation consists largely of alluvial/

River northeast of the Kikak-Tegoseak Quarry site in the

coastal plain sediments shed from the rising Brooks Range

vicinity of Ocean Point (Frederiksen et al. 1988; Fredericksen

during the Late Cretaceous. These sediments contributed to

1990). The Campanian/Maastrichtian boundary is less well

the filling of the Colville Basin, a dominantly east-west ori-

defined. Previous studies approximated the position to be

ented trough (Mull 1985; Mull et al. 2003). Paleogeographic

‘‘just slightly south’’ (Wiggins 1976) to ‘‘approximately 35

reconstructions of the Prince Creek Formation place the rising

miles south’’ (Smiley 1969; Fredericksen and McIntyre 2000)

Brooks Range up to hundreds of kilometers to the south of the

of Ocean Point. North of the confluence of the Chandler and

known dinosaur quarries along the present Colville River.

Colville Rivers, where the strata are relatively flat-lying with broad gentle folds. Thus, the Campanian/Maastrichtian tran-

Palynology

sition is likely to crop out over a long stretch of the lower Colville River between the Chandler River and the head of the

Medium to dark brown siltstones from within the Kikak-

modern Colville River delta, just downriver of Ocean Point.

Tegoseak Quarry were sampled by P.Z. for palynological analy-

458 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

FIGURE 30.2.

Location of the Kikak-Tegoseak Quarry near the confluence of Kikak Creek and the Colville River.

sis. Samples were lightly crushed, and treated with HCl to re-

the Palynodata 2000 database. Absolute and relative ages in

move carbonates, HF to remove silicates, and cold HNO3 to

Table 30.1 are after Gradstein and others (2004).

remove sulfides and dissaggregate massed organics. The remaining organic residue was sieved through a 10–mm mesh screen. The residue was strewn in glycerin jelly on a glass

Sedimentology

coverslip and mounted to a microscope slide with glycerin

Alluvial deposits of the Prince Creek Formation were exam-

jelly. The slides were examined with bright field and inter-

ined in the field, and all stratigraphic sections were logged

ference contrast illumination. The absolute abundance of pal-

noting both sedimentological and paleopedological features

ynomorphs in the encasing siltstone is very high.

according to standard procedures (Figs. 30.5, 30.6; Day 1983).

Seventy-one palynomorph taxa were recovered, identified

Paleosols were sampled by collecting undisturbed blocks at

and quantified. Twenty-five of these pollen and spore taxa

10 cm intervals. Thin sections were prepared from undis-

have useful biostratigraphic ranges that, in combination, de-

turbed blocks following air-drying and impregnation with

fine the age of the site as earliest Maastrichtian (Fig. 30.4; Table

resin (Murphy 1986). Samples were described using the termi-

30.1). Comprehensive research of previously published paly-

nology of Bullock et al. (1985) and Stoops (2003). Polished

nomorph age citations was facilitated, in part, by the use of

surfaces of impregnated samples were studied with a binocu-

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 459

FIGURE 30.3.

Stratigraphic relationships between the Prince Creek, Schrader Bluff, and Gubik formations near the KikakTegoseak Quarry site (modified from Mull et al. 2003).

lar microscope under low magnification (up to 10— mag-

plant fragments and impressions, mudstone rip-up clasts,

nification), and thin sections were subsequently examined

tuffaceous pebbles/nodules, siderite nodules/concretions,

under plane- and cross-polarized light using a Leica petro-

rare scour and fill structures, dewatering structures (e.g., con-

graphic microscope.

volute bedding), and fossil bone. The sandstone body often

The facies architecture of the Prince Creek Formation near

contains rippled intervals that become more common near

the Kikak-Tegoseak Quarry consists of small, shallow, ribbon-

the top. Carbonized organic material is common along both

like, sandy anastomosed channel fills with width/depth ratios

ripples and trough cross-beds. Paleoflow directions deter-

ⱕ30 and sharp margins, surrounded by abundant overbank

mined from an analysis of trough cross-beds coupled with

deposits (siltstone to sandstone ratio 70:30) including levees,

bounding surface orientations indicate that lateral accretion

crevasse splays and weakly developed paleosols. Larger, mean-

dominates with downstream accretion as a secondary compo-

dering trunk channel fills are located nearby, but they are not

nent. The medium-grained, multistory sandstone body fines

present in the immediate vicinity of the quarry. Alluvial sedi-

upward into a 2–6 m thick succession of heterolithic fill, com-

ments of the Prince Creek Formation at the Kikak-Tegoseak

posed of alternating, inclined beds of interbedded very-fine

Quarry can be assigned to one of six facies associations.

rippled sandstone and silt/mudstone. This inclined heterolithic stratification (IHS) grades upward into a 3–4 m suc-

FACIES ASSOCIATION 1: THICK, MULTISTORY SANDSTONE (MEANDERING TRUNK CHANNEL)

cession of facies that may include a combination of planar laminated, rippled, or massive olive-grey or brown siltstone with coalified plant fragments and carbonaceous roots, blocky

Facies Association (FA) 1 is characterized by a thick (13–20 m),

rooted mudstone with coalified plant fragments and roots,

fining upward succession (FUS) dominated by an erosionally

coaly shale, and coal. This facies association may extend as

based, multistory, medium-grained, trough-cross-bedded

much as hundreds of meters laterally, and is typically capped

sheet sandstone body with a basal lag. At the base of the FUS is

by a thin ([ 1 m thick) coaly shale, often containing car-

a 5–50 cm thick lag composed of a combination of pebble- to

bonized roots and plant fragments or, rarely, a thin ([ 1 m

boulder-sized (up to 30 cm long) clasts of chert, quartz, and

thick) coal. This facies association grades or pinches out later-

quartzite. Coalified or silicified logs, wood fragments, wood

ally into finer-grained floodplain facies.

impressions, coalified plant fragments, mud rip-up clasts, and

Interpretation. The FUS is interpreted as a succession of

bone fragments are common within the lag. The basal lag fines

point bar deposits in a large, meandering trunk channel

upward into a multistory, 7–11 m thick, massive-to-trough-

(sensu Allen 1964, 1965, 1970; Elliott 1976). A fining upward

cross-bedded, medium-grained sandstone body, which lacks

succession exhibiting evidence for lateral accretion is typical

roots but also may contain coalified or silicified logs, wood/

for point bar deposits of meandering streams (Nanson 1980;

460 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

Table 30.1. List of Palynomorphs with Age Ranges (Mega Annum) That Help Define the Age of the Kikak-Tegoseak Sample Taxon

FAD

LAD

Region

Aquilapollenites decorus

71.3

Maas.

65

Maas.

Canada

Aquilapollenites fusiformis

70.4

Early Maas.

65

Late Maas.

Canada AK

Aquilapollenites quadrilobus

85.8

Santonian

61

Danian

Worldwide

Aquilapollenites reticulatus

76.2

Late Camp.

61

Danian

Worldwide

Aquilapollenites scabridus

80.7

Middle Camp.

71.3

Late Camp.

Worldwide

Azonia cribrata

80.7

Middle Camp.

67.25

early L. Maas.

AK Canada

Azonia parva

83.5

Campanian

71.3

Late Camp.

AK Canada

Azonia pulchella

84.6

Late Sant.

67.25

early L. Maas.

AK Canada

Azonia strictiparva

71.3

Early Maas.

67.25

early L. Maas.

AK Canada

Callistopollenites radiostriatus

71.3

Maas.

65

Maas.

Canada E.Russia

Barremian

65

Maas.

w. N. America

61

Danian

Worldwide

Cedripites canadensis

127

Cranwellia rumseyensis

76.2

Late Camp.

Cranwellia striata

76.2

Late Camp.

61

Danian

Worldwide

Erdtmanipollis procumbentiformis

80.7

Middle Camp.

49

Ypresian

Worldwide

Faguspollenites granulatus

80.7

Middle Camp.

65

Maas.

Canada E.Russia

Integricorpus sp 1 of Samoilovich 1967

83.5

Camp.

61

Danian

E.Russia

Liliacidites variegatus Loranthacites pilatus Mancicorpus trapeziforme

102 71.3 76.2

Late Albian Early Maas.

5.3 65

Miocene

Worldwide

Maas.

Canada

Late Camp.

65

Maas.

Worldwide

Aptian

71.3

Camp.

w. N. America

Microreticulatisporites uniformis

121

Polycingulatisporites triangularis

220.7

Norian

65

Maas.

AK Canada

71.3

Maas.

61

Danian

Worldwide

Sigmopollis psilatus

71.3

Maas.

0

Recent

North America

Stereisporites regium

144.2

Berriasian

54.8

Paleocene

Worldwide

Tricolpopollenites parvulus

112.2

Early Albian

61

Danian

Worldwide

Santonian

70.8

early E. Maas.

Canada US

Quercoidites genustriatus

Trudopollis meekeri

85.8

Abbreviations: FAD: first appearance datum; LAD: last appearance datum.

Ethridge et al. 1981; Smith 1987). The trunk channel desig-

FACIES ASSOCIATION 2: THIN, LENTICULAR

nation for FA 1 is based on the incidence of both the largest

SANDSTONE (ANASTOMOSING SPLAY CHANNEL)

grain size and greatest sandstone body thickness in the region (cf. Ethridge et al. 1981). In this interpretation, the basal

FA 2 is characterized by thin (1–3 m), medium to very fine-

lag represents deposition in the thalweg of the channel (Allen

grained beds of sandstone with erosive, concave-up basal con-

1970). Medium-grained, trough cross-bedded sandstones are

tacts, overlain by a succession of thin sandstones with rare,

interpreted as three-dimensional dunes deposited along the

ripple cross-lamination or trough cross-bedding. Most sand-

bed of the channel during high flow (Plint 1983). Rare, con-

stones in FA 2 show slight upward-fining trends. These ribbon-

volute bedding suggests high sedimentation rates with shear-

like sandstone bodies pinch out laterally (maximum width of

ing of sediments by current drag while sediments were still

15 m). Small-scale (30 cm thick) inclined surfaces exist in sev-

liquefied (Banks 1973). Scour and fill structures are inter-

eral exposures. Paleocurrents from trough cross-beds and rip-

preted as chute channels on point bars (Plint 1983). Inclined

ple cross-laminations from sandstone bodies indicate flow to

heterolithic stratification (IHS) consisting of fine-grained

the north-northwest.

sandstone and silt/ mudstone couplets suggest a possible tidal

FA 2 typically incises or grades upward into thick inter-

influence (Smith 1987; Thomas et al. 1987). Organic-rich silt-

bedded siltstone and sandstone, massive siltstone or thin

stone, coaly shale, rooted mudstone, and coal are interpreted

coaly shale. Inclined heterolithic stratification is common

as vertical accretion deposits laid down at the tops of point

and rare ferruginous nodules and in situ root traces are present

bars during channel migration and/or fine-grained channel

throughout some of the sandstone beds.

fills that developed after abandonment (Allen 1970; Nanson 1980).

Interpretation. These sandstones are interpreted as anastomosing splay channels. Channelized flow incised into soft

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 461

Palynomorphs collected from the KikakTegoseak Quarry. 1: Baculatisporites comaumensis (trilete spore [t.s.], 37.4 mm); 2: Biretisporites sp. (t.s., 37.4 mm); 3: Cicatricosisporites sp. (t.s., 46 mm); 4: Leptolepidites sp. (t.s., 44 mm); 5: Osmundacidites wellmanii (t.s., 34 mm); 6: Polycingulatisporites triangularis (t.s., 42 mm); 7: Stereisporites antiquasporites (t.s., 28 mm); 8: Stereisporites regium (t.s., 26 mm); 9: Trilobosporites cf. crassus (t.s., 37 mm); 10: Laevigatosporites sp. (monolete spore [m.s.], 62 mm); 11: Laevigatosporites sp. w/perine sp. (m.s., 43 mm); 12: Polypodiisporites amplus (m.s., 41 mm); 13: Cycadopites sp. (gymnosperm pollen [g.p.], 32 mm); 14: Laricoidites magnus (g.p., 69 mm); 15: Taxodiaceaepollenites hiatus (g.p., 27 mm); 16: Liliacidites sp. (angiosperm pollen [a.p.], 27.5 mm); 17: Liliacidites variegatus (a.p., 32 mm); 18: Cranwellia striata (a.p., 40 mm); 19: Loranthacites pilatus (a.p., 38 mm); 20: Tricolpites microreticulatus (a.p., 20 mm); 21: Tricolpopollenites parvulus (a.p., 16 mm); 22: Azonia cf. parva (a.p., 32 mm); 23: Azonia cribrata (a.p., 31 mm); 24: Azonia pulchella (a.p., 27 mm); 25: Azonia strictiparva (a.p., 21 mm); 26: Aquilapollenites cf. decorus (a.p., 30 mm); 27: Aquilapollenites cf. quadrilobatus (angiosperm pollen, 65 mm); 28: Aquilapollenites fusiformis (a.p., 46 mm); 29: Callistopollenites? radiostriatus (a.p., 30 mm); 30: Faguspollenites cf. granulatus (a.p., 26 mm); 31: Integricorpus sp.1 Samoilovitch 1967 (a.p., 83 mm); 32: Integricorpus sp.1 Samoilovitch 1967 (a.p., 83 mm); 33: Mancicorpus trapeziforme (a.p., 21 mm); 34: Quercoidites genustriatus (a.p., 30 mm); 35: Tricolporate pollen 1 (a.p., 20 mm); 36: Betulapollenites (a.p., 28 mm); 37: Trudopollis? cf. meekeri (a.p., 22.5 mm); 38: Erdtmanipollis cf. procumbentiformis (a.p., 26 mm); 39: Lecaniella cf. Zygogonium cf. tunetanum (freshwater chlorophyte, 27 mm); 40: Botryococcus sp. (freshwater alga, 82 mm). FIGURE 30.4.

462 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

Measured section for the Prince Creek Formation that includes the Kikak-Tegoseak Quarry. ‘‘0 m’’ corresponds to the level of the Colville River in July 2005. FIGURE 30.5.

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 463

Measured section approximately 0.5 km north of the Kikak-Tegoseak quarry. Section was sampled extensively for paleopedological and palynological analyses. FIGURE 30.6.

464 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

floodplain siltstone deposits creating stable channels no more

Kraus and Wells 1999; Kraus 2002; Kraus and Davies-Vollum

than 3 m deep and 15 m wide. Lateral accretion surfaces are

2004).

less common within these shallow sand bodies than in larger tabular sand bodies (FA 1) observed elsewhere within the Prince Creek Formation, supporting the anastomosing hypothesis (Nadon 1994; McCarthy and Plint 2003). The mas-

FACIES ASSOCIATION 4: LAMINATED SILTSTONE (SHALLOW LACUSTRINE DEPOSIT)

sive siltstones overlying these sandstones represent promi-

Massive, featureless and laminated siltstone and organic-rich

nent levees that would have inhibited lateral migration of the

sediments 1–4 m thick characterize FA 4. Carbonaceous lami-

channels (Nadon 1993, 1994; Heritage et al. 2001). The splay

nations are interbedded with siltstone and silty-mudstone.

channels are isolated and have erosional bases and ribbon-like

Carbonaceous plant fragments are present but root traces are

geometries similar to splay channels observed by Smith et

generally absent.

al. (1989) in the North Saskatchewan River system. Multiple

Interpretation. FA 4 is interpreted as shallow lacustrine

channels have been observed at the same stratigraphic level

(pond) deposits that probably occupied spaces between splay

(Brandlen 2008). Smith et al. (1989) suggested low bedload

channels and topographic lows on flood plains. It typically

transport, low sedimentation rates and permanently standing

lacks pedogenic features such as root traces or burrows found

water on the flood plain favored the formation of splays like

in other facies associations. The presence of carbonaceous

these. Kraus (1987) recorded similar ribbon sand bodies in the

plant fragments suggests partially anoxic and/or neutral pH

Willwood Formation and noted that they appeared laterally

conditions during deposition of these deposits (Retallack

fixed, and exhibited low-sinuosity. We suggest that the small,

2001). The presence of coaly shale and organic-rich siltstone in

anastomosed splay channels in the Prince Creek Formation

places suggests development of palustrine depositional envi-

are distributaries to a larger, meandering trunk stream system

ronments (Ashley et al. 2004; Johnson and Graham 2004).

(FA1). Sediment would have filled these splay channels, choking them and resulting in subsequent avulsion or abandonment (Bridge 1984; Kraus 1987). The presence of root traces in some of the splay sediments indicates that there were periods of subaerial exposure (McCarthy et al. 1997a, b).

FACIES ASSOCIATION 5: INTERBEDDED TABULAR SANDSTONE AND SILTSTONE (LEVEE) Tabular sandstones up to 1 m thick, overlain by successive intervals of olive-grey and/or dark brown siltstone and sandy-

FACIES ASSOCIATION 3: TABULAR SANDSTONE AND SILTSTONE (CREVASSE SPLAY)

siltstone characterize FA 5. These siltstone intervals are typically 5–8 m thick and may be overlain by, or are interbedded with, organic-rich deposits. The lower contacts between FA 5

FA 3 consists of sandstone bodies, less than 50 cm thick, and

and anastomosed splay channel sandstones (FA 2) may be ero-

interbedded siltstones and mudstones up to 1 m thick. Sedi-

sional surfaces. Lenticular sandstones of FA 3 rarely incise this

mentary structures are not well exhibited in these deposits,

association. The siltstones display abundant evidence of pedo-

but ripple cross-laminations are preserved in some beds. Ero-

genesis such as root traces and ferruginous nodules. Evidence

sive bases are present but uncommon. Carbonaceous root

of bioturbation is common and primary sedimentary struc-

traces, plant fragments, ferruginous nodules, and mud clasts

tures are difficult to discern. Interpretation. The presence of abundant root traces and

are present. Interpretation. FA 3 is interpreted as the result of crevasse

plant fragments and weak paleosol development supports a

splay deposition (Kraus 1987). The sandstone bodies represent

levee interpretation for FA 5 (Smith 1976; Nadon 1993, 1994).

proximal splay deposits and/or vigorous flow. The siltstones

The repetitious nature of this association indicates cyclic sub-

and mudstones represent distal splay deposits and/or evi-

aerial exposure and subsequent drowning of the floodplain by

dence of waning flow. Carbonaceous root traces and ferru-

overbank flooding (Bown and Kraus 1987; Kraus 1987; Kraus

ginous nodules indicate pauses in sedimentation. The de-

and Bown 1988).

velopment of immature paleosols indicates crevasse splay propagation onto the floodplain was relatively common, with sufficient time between splays to produce weakly developed soils. Kraus (1987, 2002) suggested this arrangement is indica-

FACIES ASSOCIATION 6: SILTSTONES AND MUDSTONES (FLOODPLAIN FINES)

tive of semi-permanent splay features. In situations where tab-

FA 6 consists of thin deposits of coaly shale and organic-rich

ular sandstone bodies of FA 3 overlie lenticular sandstone

siltstone, overlying olive-grey or brown siltstone. Preserved

bodies of FA 2, they may represent overbank sedimentation

plant matter includes leaves, stems, roots, and small pieces

during the initial stages of avulsion (Bown and Kraus 1987;

of wood. These organic-rich accumulations may be laterally

Kraus 1987; Kraus and Bown 1988; Kraus and Gwinn 1997;

continuous or discontinuous and have a high clastic content.

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 465

FIGURE 30.7.

Paleopedological features. (A) Carbonaceous root trace (arrow) in plane-polarized light; (B) bioturbated groundmass showing irregular pattern of incorporation of organic matter and clay concentrations (arrow) in plane-polarized light; (C) microlaminated illuvial clay coating (arrow) in planepolarized light (note the multiple layers that are characteristic of well-developed clay coatings in modern soils; (D) clay-rich papule (arrow) in plane-polarized light; (E) Fe-oxide nodule in mudstone matrix (arrow) in plane-polarized light (note sharp boundaries with surrounding matrix); (F) Fe-depletion coating in fine-grained matrix (arrow) in cross-polarized light.

Underlying the organic-rich layers are olive-grey and/or

note better-drained or perhaps seasonally dry conditions

brown blocky, unconsolidated siltstones containing root

(Smith et al. 1989; Nadon 1994; Ashley et al. 2004; Johnson

traces, plant fragments, and yellowish-red ferruginous nod-

and Graham 2004). Evidence of bioturbation also suggests pe-

ules up to 3 mm in diameter. Root traces originating from the

riods of improved drainage (Retallack 2001). The high clastic

black organic-rich intervals commonly penetrate 3–5 cm into

content of the organic-rich layers suggests that true coal

lower beds, but are rarely seen penetrating as deep as 25 cm.

swamps were rare and that much of the organic material may

Basal contacts are typically gradational. FA 6 is differentiated

have been transported.

from FA 4 by the presence of gradational contacts between the siltstone and organic-rich deposits. Ferruginous features such as grain and void coatings and nodules are common in thin section. Evidence of bioturbation is abundant in thin sections.

FLOODPLAIN PALEOSOLS AT THE KIKAK-TEGOSEAK QUARRY

Interpretation. FA 6 is interpreted as floodplain deposits. The

Most of the floodplain sediments in the Prince Creek Forma-

drab colors and preserved plant matter suggest poorly drained

tion constitute successions of very fine-grained sandstones

conditions typical of wetland environments (Bown and Kraus

and dark grey, olive grey, and brown, laminated carbonaceous

1987; Kraus 1987; Kraus and Bown 1988; Retallack 2001; Ash-

mudstones interbedded with thin coaly shales and rooted

ley et al. 2004). However, extensively rooted horizons and

horizons interpreted as the deposits of a low-lying alluvial

abundant ferruginous features and illuvial clay coatings de-

plain. Paleosols in the Prince Creek Formation developed pri-

466 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

marily in floodplain claystones and siltstones, although many

tions and/or periodic flooding and surface ponding than with

sandstones display evidence of rooting. Individual paleosol

diagenetic development during continuous burial of the pa-

profiles are difficult to discern in the field owing to the

leosol beneath the water table (PiPujol and Buurman 1994;

cumulative nature of pedogenesis on these sedimentologi-

McCarthy et al. 1999).

cally active floodplains. Nevertheless, insights into pedogenic

Furthermore, the development of illuvial clay coatings

processes can be gained by micromorphological analyses of

within floodplain paleosols of the Prince Creek Formation

floodplain mudstones (Fig. 30.7; Table 30.2). Detailed field de-

strongly suggests alternating wetting and drying conditions

scriptions and geochemical and mineralogical trends in the

(Fedoroff et al. 1990). The presence of moderately oriented

paleosols are presented elsewhere (Brandlen 2008).

clay coatings and linear concentrations within the matrix suggests that some horizons may have been saturated for at least

PALEOENVIRONMENTAL IMPLICATIONS OF MICROMORPHOLOGICAL FEATURES FROM PALEOSOLS

part of the time (McCarthy 2002). Fragmented clay coatings and papules resulted from incorporation of translocated clay coatings into the matrix through the combined action of bioturbation and physical shrink-swell processes (i.e., wetting

The major pedogenic processes operating in floodplain paleo-

and drying; Fitzpatrick 1993; Wang et al. 1995).

sols of the Prince Creek Formation near the Kikak-Tegoseak Quarry were redoxymorphic processes, pedoturbation and

FAUNAL ANALYSIS

clay illuviation. Poorly drained conditions on the floodplains are suggested by the overall dull grey colors and abundance of

Extensive excavation of the Kikak-Tegoseak Quarry has shown

organic matter. The associated presence of lacustrine shales

that this site contains a preponderance of individuals and skel-

also indicates that permanent lakes or ponds existed on these

etal remains of Pachyrhinosaurus (Ceratopsia: Centrosaurinae)

floodplains.

and represents the northernmost locality of any ceratopsian

Evidence for fluctuating redox conditions, attributed to

bonebed. The ceratopsians recovered from the Kikak-Tegoseak

variations in groundwater levels, can be detected by analyzing

Quarry are referred to the centrosaurine genus Pachyrhino-

assemblages of micromorphological features. The presence of

saurus on the basis of the thickening and flattening of the top

void coatings, mottles and nodules of Fe-oxides and Fe-oxide

of the skull between orbits and the nares (sensu Langston

depletion zones adjacent to large voids indicates that ferru-

1967; Fig. 30.8). Achelousaurus shares a characteristically large

ginous compounds were redistributed within the paleosols

nasal boss (Sampson 1995) but the Kikak-Tegoseak specimens

(Vepraskas et al. 1994; Driese et al. 1995). Iron oxides coat-

have larger nasal bosses that extend onto the frontals, consis-

ing the surfaces of voids and channel margins indicate that

tent with the diagnosis for Pachyrhinosaurus.

the surrounding groundmass was at least periodically satu-

Other vertebrate taxa from this quarry are one type of inde-

rated, and that the soil also dried out periodically, allow-

terminate osteichthyan fish, Dromaeosaurus albertensis, Troo-

ing the dissolved Fe2+ to migrate towards the more oxidizing

don formosus, cf. Gorgosaurus libratus, an indeterminate or-

larger pores, where it was precipitated as Fe

nithomimid, and an indeterminate hadrosaur (Fiorillo and

3+

when the soil

dried out (Bouma et al. 1990; Vepraskas et al. 1994).

Gangloff 2000; Fiorillo 2006).

Iron oxide depletion coatings indicate that ponded water

Excavation of a 5 m — 4 m pit with a headwall up to 2.5 m

was present at some times and places on the floodplains

high within the bonebed yielded the skeletal remains, includ-

whereas the phreatic zone was located below the solum (Pi-

ing skulls, of at least nine individuals of Pachyrhinosaurus, as

Pujol and Buurman 1994). As ponded surface water drained

determined from the number of recovered occipital condyles.

through the vadose zone, Fe2+ was removed from the large

Because material is still being prepared, the paleontological

pores and precipitated in more oxidizing ped interiors and/or

results reported here are preliminary.

transported out of the paleosol with migrating groundwater

A study of taphonomic parameters (sensu Badgley and Be-

(Vepraskas et al. 1994; PiPujol and Buurman 1997). Although

hrensmeyer 1980; Badgley 1986 a, b; Fiorillo 1988, 1991;

similar features are known to form diagenetically during early

Rogers 1990; Eberth et al. 2007; and many others) provides a

burial (e.g., diffusion gleyans of Retallack 2001), the coexis-

means for determining the biological and sedimentological

tence of pale orange, microlaminated illuvial clay coatings

processes that influenced the formation and preservation of a

that must have formed under well-drained conditions (Fedo-

bonebed. In her examination of bonebeds through time, Be-

roff 1997; McCarthy et al. 1998) within drab-colored iron de-

hrensmeyer (2007) has recently summarized typical tapho-

pletion zones suggests variable redox conditions at different

nomic patterns and, similar to Rogers and Kidwell (2007),

times. This situation is more consistent with pedogenic de-

has discussed the various biotic and abiotic mechanisms for

velopment within a zone of fluctuating water table condi-

bonebed formation.

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 467

Table 30.2. Micromorphological Features Observed in Floodplain Palaeosols near the Kikak-Tegoseak Quarry (Prince Creek Formation) Feature

Description

Size

Interpretation

Biological features Root traces

Vertically to subhorizontal, downward tapering

3–10cm

and branching features preserved as thin carbon

long; 1–2 mm

films or as carbonaceous/coalified material (Fig.

wide

Former root zones; preserved roots and rootlets

30.7A) Bioturbation,

Burrows: subhorizontal elliptical tubules with

faunal activity

smooth or irregular boundaries; filled with fine-

0.5–2 mm

Burrows probably attributed to insects and/or earthworms

grained material similar to surrounding matrix and/or organic particles Bioturbation: disrupted, isotropic groundmass,

Fabrics similar to ‘‘excremental total fabrics’’ of bi-

disseminated organic matter (Fig. 30.7B), small

ological origin (Courty and Fedoroff, 1985); sug-

granular aggregates; may contain iron nodules or

gests prolonged surface stability

nodule fragments or small pods of coarser material than surrounding matrix Organic matter

Small, opaque fragments with rarely preserved cel-

0.5–5 mm

lular structure; typically disrupted fragments infil-

Abundance of organic material suggests abundant vegetation and/or widespread poorly drained con-

led with fine-grained matrix material

ditions

Large, carbonaceous or weakly coalified material

2–15 cm thick

Preserved roots and/or fallen logs

Microstructure

Peds: natural surfaces of weakness recognized by

0.5–5 cm

Blocky structures develop through repeated wet-

(peds and voids)

presence of clay coatings, Fe-oxide staining or by

(avg. 1–2 cm)

ting and drying (Fanning and Fanning, 1989);

Physical features

natural voids; blocky, platy and complex micro-

complex structures develop by the action of plant

structures are recognized; sandstones are generally

roots and/or soil organisms (Fitzpatrick, 1993)

apedal Voids: continuous to discontinuous linear voids

] 50–2000 mm

occur Birefringence

Striated fabrics predominate; stipple-speckled,

Variable (20–200

Striated fabrics develop from realignment of detri-

fabric (b-fabric)

mosaic-speckled, porostriated, granostriated, par-

mm)

tal clays from pressure and tensions due to alter-

allel striated and random striated fabrics are also

nate wetting and drying (Brewer, 1976); speckled

observed

fabrics develop through suspension settling or flocculation of alluvial materials (Stoops, 2003)

Sedimentary

Generally upward-fining microlamination recog-

micro-

nized by subtle grain size variation and/or colour

lamination

change; lamination typically disrupted by root

0.5–1 mm

Produced by deposition and subsequent compaction of floodplain sediments;

traces; rare in mudstones but common in sandstones Textural features 30–50 mm thick

Result from physical translocation of clay; re-

Clay coatings

(i) gray, yellow and pale orange clay infillings and

and infillings

coatings and hypocoatings of voids; (ii) pale

quires soil wetting and subsequent drying out

organge ped coatings; commonly microlaminated

(McKeague, 1983; Bullock and Thompson, 1985)

crescentic clay coatings partially or completely fill pores to form discontinuous to continuous void infillings (Fig. 30.7C); compound coatings consist of clay interlaminated with silt-size material and/or Fe-Mn oxides

468 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

Table 30.2. (Continued)

Feature Papules

Description

Size

Interpretation

Small, subrounded to rounded, clay-rich pellets

0.5–2 mm

Pedoturbation and/or erosion and short-term

with sharp boundaries containing colors and tex-

diameter

transport (Muller et al., 2004)

25–60 mm thick

Result from compression of void coatings by ped-

tures different from surrounding matrix (Fig. 30.7D) Linear and ir-

Intrapedal clay concentrations; may be straight,

regular clay

serrated or interlaced; may also contain silt-sized

oturbation; some may be intercalations that form

concentrations

grains

in saturated soils (Bullock and Murphy, 1979; Mc-

Pedorelicts

Subrounded to rounded clay and/or silty clay

Carthy, 2002) 0.05–4 mm

nodules; sharp boundaries

Low energy erosion and short-distance transport (Fielder and Sommer, 2004)

Ferruginous features Fe-oxide void

Mottles and void coatings: local redistribution of

Mottles (1–2

Segregation of Fe-oxides as result of numerous

segregations

Fe-oxides form mottles; void coatings and hypoc-

mm diameter);

wetting and drying cycles; coatings on ped sur-

oatings: thin, dark red to black coatings along

coatings and hy-

faces suggest saturation for at least part of the time

void margins and root tubules

pocoatings (0.5–

(Bouma et al., 1990)

1 mm thick) Nodules: strongly impregnated, discrete con-

0.5–3 mm

centration features with subrounded to well-

diameter

rounded shapes and sharp to distinct boundaries with surrounding matrix (Fig. 30.7E); may envelope silt and sand-size mineral grains Depletion

Surfaces of pedogenic aggregates from which iron,

coatings

clay or both have been removed (Fig. 30.7F)

[ 1 mm thick

Fe-oxides removed from surfaces of voids under conditions of free drainage (Bullock et al., 1985; Fitzpatrick, 1993)

At the Kikak-Tegoseak Quarry, the accumulation of bones ex-

entation of bones. Nevertheless, vertically oriented bones were

tends for at least 90 m along the crest of the bluff. Excavations

not observed during any phase of the excavation. That is, long

have produced more than 120 bones per cubic meter in this

axes of bones exhibit horizontal to subhorizontal orientations,

bonebed. Field observations suggest that the bones are not

and are mostly parallel to the bedding plane of the host facies.

sorted by size or type in their vertical distribution.

The large size of these bones in comparison to the fine grain

The dense concentration of bones at this quarry, in com-

size of the matrix indicates that the individual bones and

bination with the multiple individuals of ceratopsians (see

skulls were probably not carried to the site by fluvial processes.

below), suggests that there is a very high degree of skeletal

Rather, the suggestion of skeletal association, especially in the

association in this bonebed, although no articulated or di-

case of complete skulls, suggests that these fossils originally

rectly associated complete skeletal remains were observed dur-

floated into the site as whole carcasses or body parts carried by

ing field excavations. Field crews successfully retrieved one

a relatively low-energy current, and then became disarticu-

skull of Pachyrhinosaurus (DMNH 22558) in 2006 that was

lated after transport.

lying on its left side, and a second skull in 2007 that was lying

Bone modification features are limited in the small sam-

on its palate. However, although all skeletal elements seem to

ple of bones that we have so far examined from the Kikak-

be represented for this taxon—an indication that all Voorhies

Tegoseak Quarry. There is no evidence of abrasion, and bone

Groups are likely represented—a further census of skeletal ele-

weathering stages range from 0 to 2 (sensu Behrensmeyer

ments awaits final preparation. Hadrosaur and small theropod

1978; Fiorillo 1988), indicating little subaerial exposure time

remains encountered thus far are restricted to isolated teeth,

prior to burial (Figs. 30.9, 30.10).

and the remaining non-ceratopsian taxa from the site appear to be represented by isolated elements.

Minor etchings from Cretaceous roots are the most common biological modification features that we documented

The high degree of fragmentation of bone by modern weath-

within the sample of bone that we examined. Root marks (Fig.

ering at this quarry prevented meaningful analysis of the ori-

30.11) are preserved as carbonized traces along bone surfaces,

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 469

FIGURE 30.8.

Pachyrhinosaurus skull from Kikak-Tegoseak Quarry (DMNH 22558). Though the specimen is not completely prepared, the thickened and flattened boss that extends onto the frontals is sufficient to identify this taxon.

and have been found to penetrate the alveoli of some cera-

served in centrosaurine ceratopsians from Dinosaur Provin-

topsian jaws. Also present, though less common than root

cial Park, Alberta (Dodson 1990). The means of the population

traces, are circular to elliptically shaped shallow borings on

of Kikak-Tegoseak occipital condyles and those published by

the bones (Fig. 30.12). These features indicate scavenging and

Dodson are 75.6 mm and 68.9 mm, respectively. The standard

reproductive behavior of partial puparia of dermestid beetle

deviations are 4.85 and 7.08, respectively. Thus, statistically,

larvae (Coleoptera: Dermestidae) on subaerially exposed car-

the range in size of the Kikak-Tegoseak sample is consistent

casses prior to burial (sensu Hasiotis et al. 1999). Tooth-

with the variation observed across the centrosaurine taxa

marked bone is rare (Fig. 30.13), and no trample marks were

studied by Dodson. Rather than invoke the presence of multi-

found.

ple ceratopsian taxa for the Kikak-Tegoseak population, we

Based on the number of prepared ceratopsian occipital con-

suggest that the quarry population represents multiple sub-

dyles, the minimum number of individuals of Pachyrhino-

adult growth stages. Given the nearly adult size, and incom-

saurus from this quarry, so far, is nine. The condyles range in

plete fusion of cranial elements, and the modest size of the

diameter from 65.1 to 80.8 mm (Table 30.3). The size range of

condyles, all nine Pachyrhinosaurus individuals appear to be

these condyles corresponds well with those measurements ob-

subadults. It is recognized however, that definitive determina-

470 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

FIGURE 30.9. Dinosaur rib (DMNH 22330) exhibiting weathering stage 0 (sensu Fiorillo 1988). Note the lack of cracking parallel to bone fiber of the cortical bone.

FIGURE 30.11. Matrix from Kikak-Tegoseak Quarry displaying carbonized root traces. Arrow indicates the up direction.

lithic sediments that they interpreted as evidence of changing flow conditions. Root traces in growth position within the channel sediments at Kikak-Tegoseak suggest that stream flow was ephemeral, perhaps seasonally (Smith and Putnam 1980; Nadon 1994). Fluvial channels in the Prince Creek Formation near the Two small bone fragments (uncatalogued) uncovered during excavation that exhibit weathering stage 2. FIGURE 30.10.

Kikak-Tegoseak quarry were part of a laterally stable, suspended load fluvial system separated by isolated, vegetated floodplains, splays and wetlands. Topographically low interchannel areas and levees are preserved as thick intervals of floodplain alluvium separating contemporaneous channels

tion of ontogenetic age of these individuals remains uncertain

(cf. Perez-Arlucea and Smith 1999). Overall, the Prince Creek

until a detailed histologic analysis is undertaken. Regardless of

Formation near the Kikak-Tegoseak quarry is interpreted as a

the details of ontogenetic age, this assemblage of individuals is

large anastomosed splay complex adjacent to a larger mean-

the northernmost record of ceratopsian dinosaurs.

dering trunk channel (Flaig et al. 2007; McCarthy et al. 2007). Palynomorph data suggest levees and better-drained portions of the floodplain were dominated by Pinaceae, herbaceous

Discussion

shrubs and ferns, while topographic lows and shallow ponds

The facies and facies architecture of the Prince Creek Forma-

were vegetated with Taxodiaceae, and ferns.

tion near the Kikak-Tegoseak Quarry are similar to published

Overall, the micromorphological features present in flood-

accounts of anastomosed fluvial systems (Smith et al. 1989;

plain paleosols near the Kikak-Tegoseak Quarry are similar

Kirschbaum and McCabe 1992; Nadon 1994; Kraus 2002).

to those found in modern Inceptisols and Alfisols (Schaetzl

Nadon (1994) suggested that anastomosed fluvial systems

and Anderson 2005). The development of modern Inceptisols

should exist where the combination of seasonal streamflow

cannot be attributed to a unique environmental signature,

and high suspended load exists. Many of the anastomosed

however, many of the micromorphological features present

channels near the Kikak-Tegoseak Quarry contain inclined

within the paleosols do provide some paleoenvironmental in-

heterolithic stratification (IHS) in fining-upward successions.

formation for the Kikak-Tegoseak Quarry. Redoximorphic fea-

Kirschbaum and McCabe (1992) reported similar anasto-

tures form in warm to cool temperate climates with mean

mosed channels in the Dakota Formation filled with hetero-

annual temperatures in the range of 5–20\C and seasonal sat-

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 471

Table 30.3. Diameters of Ceratopsian Occipital Condyles Recovered from the Kikak-Tegoseak Quarry Diameter Specimen no.

(mm)

DMNH 22194

77.7

DMNH 22195

80.8

DMNH 22198

76.8

DMNH 22257

79.5

AK 539-V-01

65.1

AK539-V-12A

74.7

AK 539-V-28

73.5

AK 539-V-29

76.7

AK 539-V-11*

—

* Though incomplete, AK 539-V-11 is included here as a voucher for the existence of a ninth individual.

uration (Driese et al. 1995; Richardson and Vepraskas 2001). The northern limit of clay illuviation in modern soils occurs around 61\ N latitude, south of the –3.5\C mean annual temperature (MAT) isotherm with mean annual precipitation of at least 375 mm (Tarnocai 1997). Cryogenic soil features have not been observed in paleosols from the Prince Creek Formation. These features develop only under cold climate conditions with MAT of –2\C or less (Tarnocai 1997). Taken together, pedogenic features preserved in the paleosols of the Prince Creek Formation near the Kikak-Tegoseak Quarry suggest a cool temperate paleoclimate with a MAT above 5\C and a MAP above 375 mm. The paleosols further suggest periods of saturation and drying out which may have Small, semi-circular to elliptical depressions on a dinosaurian limb bone. These features are attributable to dermestid beetle activity. FIGURE 30.12.

resulted from seasonal flooding resulting from spring snowmelt in the rising Brooks Range to the south or alternatively, from intense summer thunderstorms. These paleoenvironmental interpretations are consistent with other palynological and paleobotanical data. Spicer and Parrish (1990) suggested a MAT of 6–7\C for the early Tertiary North Slope of Alaska which they considered as an upper limit for Maastrichtian MATs on the North Slope, and MAP between 500 and 1500 mm a–1. The presence of abundant charcoal also indicates that forest fires were a relatively common occurrence in this area, suggesting at least periodic dry periods (Spicer 2003). The high abundance and diversity of non-marine palynomorphs and the absolute lack of marine palynomorphs supports a non-marine depositional system. Tracheid, charcoal and structured woody clasts are rare. Small pieces of amber were also recovered. The vegetation of northern Alaska in general during this time was largely a conifer forest on the coastal plains, with a riparian broad-leafed deciduous forest. The un-

Tooth-marked bone uncovered during excavation. Mark is approximately 3 cm long. FIGURE 30.13.

derstory of the coniferous forest was likely comprised of ferns and angiosperms (Spicer 1987, 2003). Local derivation of fossil pollen at the Kikak-Tegoseak Quarry is suggested by the high relative abundance of a few

472 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

non-arboreal species (Polypodiaceae, Osmundaceae, and the

ern and ancient record, further study of the Kikak-Tegoseak

triprojectates), the common occurrence of pollen clusters, and

Quarry will provide a unique opportunity to gain important

large sheets of plant cuticle. Also at the quarry, the very low

insights into the paleobiology of horned dinosaurs in the

absolute and relative abundance of evidence of the Taxo-

ancient northern polar world.

diaceae, Sphagnaceae, other aquatic plants, and freshwater algae suggest that the local environment and sediment source

Acknowledgments

environments were well drained and contained few bodies of

The authors thank Ronald Tykoski, Shelley Hartsfield, and

standing water (ponds, oxbows, kettles, etc.) where swamp

Mark Turner for the fine preparation of the dinosaur speci-

and aquatic plants and algae could flourish.

mens that form the basis for this report. The authors also

By virtue of their paleogeographic setting, the Cretaceous

gratefully acknowledge Kent Newman, Thomas Adams, Chris-

bonebeds found in northern Alaska were formed under

topher Strganac, and Jason Petula for their efforts in exca-

unique conditions. The Brooks Range formed the southern

vating specimen DMNH 22558. David Eberth and Catherine

edge of the Colville Basin in the Cretaceous. A conservative

Forster provided useful comments on an earlier version of this

estimate for the elevation of the Brooks Range at that time is

manuscript. The authors also thank the National Science

1,500 m (Spicer 2003). Further, in contrast to estimates of mild

Foundation (OPP 04-24594 and OPP 04-25636) for financial

temperatures on the coastal plain, estimates of mean annual

support, the Barrow Arctic Science Consortium and VECO

temperature in the Brooks Range are below 0\C and, thus,

Polar Resources for logistical support in the field, and the

capable of supporting icefields and associated permafrost.

Bureau of Land Management Alaska State Office for admin-

Stream dynamics in permafrost dominated-watersheds are

istrative support. The authors also gratefully acknowledge the

known for periods of intense seasonal runoff (Woo and Win-

support of the Museum of Nature and Science, Dallas, Ameri-

ter 1993; Vandenberghe 2003) and this mechanism may be

can Airlines, Whole Earth Provision Company, Biostratigra-

responsible for the extraordinary assemblage of bones found

phy.com LLC, Arco Alaska, Inc., and Phillips Petroleum, Inc.

at the Kikak-Tegoseak Quarry. Thus, a small herd of the ceratopsian taxon Pachyrhinosaurus may have been captured by one of these hypothesized intense seasonal periods of runoff. These remains were mixed with the skeletal remains of other dinosaurs such as theropods, and a hadrosaur. Subsequent to death and entrapment, but prior to burial, the remains of the animals represented at the Kikak-Tegoseak Quarry were utilized by Cretaceous dermestids. The final stage in the biological history of this assemblage was the establishment of an incipient paleosol that supported a well-developed vegetative cover.

Summary The Kikak-Tegoseak Quarry is a new dinosaur bonebed in northern Alaska that contains numerous bones of several dinosaur taxa. The quarry accumulation is dominated by the remains of the horned dinosaur Pachyrhinosaurus, although other large-bodied dinosaurs are represented to a lesser extent at the site. Based on palynology, the age of this bonebed is likely earliest Maastrichtian. The depositional environment based on fossil pollen suggests that the pollen was locally derived and the environment was well drained. Sedimentological analyses at the site indicate that the depositional environment consisted of an anastomosing splay complex. Fossil soils in the section containing the quarry suggest that floodplains between anastomosed channels were frequently wet, but that water levels probably fluctuated seasonally from shallow, standing water to dry and subaerially exposed. As the northernmost accumulation of horned dinosaur bones in the mod-

References Cited Allen, J. R. L. 1964. Studies in fluviatile sedimentation: Six cyclothems from the lower Old Red Sandstone, Anglo-Welsh Basin. Sedimentology 3: 163–198. ———. 1965. Fining upward cycles in alluvial successions. Geological Journal 4: 229–246. ———. 1970. Studies in fluviatile sedimentation: A comparison of fining-upwards cyclothems, with special reference to coarsemember composition and interpretation. Journal of Sedimentary Petrology 40: 299–323. Ashley, G. M., J. Maitima Mworia, A. M. Musya, R. B. Owens, S. G. Driese, V. C. Hover, R. W. Renaut, M. F. Gowan, S. Mathai, and S. H. Blatt. 2004. Sedimentation and recent history of a freshwater wetland in a semi-arid environment: Lobi Swamp, Kenya, East Africa. Sedimentology 55: 1301–1321. Badgley, C. 1986a. Counting individuals in mammalian fossil assemblages from fluvial environments. Palaios 1: 328–338. ———. 1986b. Taphonomy of mammalian fossil remains from Siwalik rocks of Pakistan. Paleobiology 12: 119–142. Badgley, C., and A. K. Behrensmeyer. 1980. Paleoecology of Middle Siwalik sediment and faunas of the Potwar Plateau. Palaeogeography, Palaeoclimatology, Palaeoecology 30: 133–155. Banks, N. L. 1973. The origin and significance of some downcurrent-dipping cross-stratified sets. Journal of Sedimentary Petrology 43: 423–427. Behrensmeyer, A. K. 1978. Taphonomic and ecologic information from bone weathering. Paleobiology 4: 150–162. ———. 2007. Bonebeds through time. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 65–102. Chicago: University of Chicago Press.

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 473

Besse, J., and V. Courtillot. 1991. Revised and synthetic apparent polar wander paths of the African, Eurasian, North American and Indian plates, and true polar wander since 200 Ma. Journal of Geophysical Research 96: 4029–4050. Bouma, J., C. A. Fox, and R. Miedema. 1990. Micromorphology of hydromorphic soils: Applications for soil genesis and land evaluation. In L. A. Douglas, ed., Soil Micromorphology: A Basic and Applied Science, pp. 257–278. Amsterdam: Elsevier. Bown, T. M., and M. J. Kraus. 1987. Integration of channel and floodplain suites I. Developmental sequence and lateral relations of alluvial paleosols. Journal of Sedimentary Petrology 57: 587–601. Brandlen, E. 2008. Paleoenvironmental reconstruction of the Late Cretaceous (Maastrichtian) Prince Creek Formation near the Kikak-Tegoseak dinosaur quarry, North Slope, Alaska. M.Sc. thesis, University of Alaska, Fairbanks. Brewer, R. 1976. Fabric and Mineral Analysis of Soils. New York: Krieger. Bridge, J. S. 1984. Large scale facies sequences in alluvial overbank environments. Journal of Sedimentary Petrology 54: 583– 588. Brosge, W. P., C. L. Whittington, and R. H. Morris. 1966. Geology of the Umiat-Maybe Creek Region Alaska. U.S. Geological Survey Professional Paper 303-H: 501–638. Bullock, P., N. Fedoroff, A. Jongerius, G. Stoops, and T. Tursina. 1985. Handbook for Soil Thin Section Description. Wolverhampton: Waine Research. Bullock, P., and C. P. Murphy. 1979. Evolution of a paleo-argillic Brown earth (Palaeudalf) from Oxfordshire, England. Geoderma 22: 225–252. Bullock, P., and M. L. Thompson. 1985. Micromorphology of Alfisols. In L. A. Douglas and M. L. Thompson, eds., Soil Micromorphology and Soil Classification, pp. 17–48. Soil Science Society of America, Special Publication, 15. Chapman, R. M., R. L. Detterman, and M. D. Mangus. 1964. Geology of the Killik Etivluk Rivers region, Alaska. U. S. Geological Survey Professional Paper 303-F: 325–407. Conrad, J. E., E. H. McKee, and B. D. Turrin. 1990. Age of tephra beds at the Ocean Point Dinosaur Locality, North Slope, Alaska, based on K-Ar and 40Ar/ 39Ar Analyses. U. S. Geological Survey Bulletin 1990-C: 1–12. Courty, M. A., and N. Fedoroff. 1985. Micromorphology of recent and buried soils in a semi-arid region of northwestern India. Geoderma 35: 287–332. Davies, K. L. 1987. Duck-billed dinosaurs (Hadrosauridae: Ornithischia) from the North Slope of Alaska. Journal of Paleontology 61: 198–200. Day, J. H. 1983. The Canadian Soil Information System (CanSIS) Manual for Describing Soils in the Field, 1982 revised. Ottawa: Expert Committee on Soil Survey, Agriculture Canada, LRRI 82–52. Dodson, P. 1990. On the status of the ceratopsids Monoclonius and Centrosaurus. In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 231–243. Cambridge: Cambridge University Press

Dowdeswell, J., and M. Hambrey. 2002. Islands of the Arctic. Cambridge: Cambridge University Press. Driese, S. G., E. L. Simpson, and K. A. Erickson. 1995. Redoxomorphic paleosols in alluvial and lacustrine deposits, 1.8 Ga Lochness Foundation, Mt. Isa, Australia: Pedogenic processes and implications for paleoclimate. Journal of Sedimentary Research A65: 675–689. Eberth, D., R. R. Rogers, and A. R. Fiorillo. 2007. A practical approach to the study of bonebeds. In R. R. Rogers, D. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 265–331. Chicago: University of Chicago Press. Elliott, T. 1976. The morphology, magnitude, and regime of a Carboniferous fluvial distributary channel. Journal of Sedimentary Petrology 46: 70–76. Ethridge, F. G., T. J. Jackson, and A. D. Youngberg. 1981. Floodbasin sequence of a fine-grained meander belt subsystem: The coal-bearing Lower Wasatch and Upper Fort Union Formations, southern Powder River Basin, Wyoming. In F. G. Ethridge and R. M. Flores, eds., Recent and Ancient Nonmarine Depositional Environments; Models for Exploration, pp. 191–209. Special Publication of the Society of Economic Paleontologists and Mineralogists 31. Fanning, D. S., and M. C. B. Fanning. 1989. Soil Morphology, Genesis and Classification. New York: John Wiley & Sons. Fedoroff, N. 1997. Clay illuviation in red Mediterranean soils. Catena 28: 171–189. Fedoroff, N., M. A. Courty, and M. L. Thompson. 1990. Micromorphological evidence of paleoenvironmental change in Pleistocene and Holocene paleosols. In L. A. Douglas, ed., Soil Micromorphology: A Basic and Applied Science, pp. 653–666. Amsterdam: Elsevier. Fielder, S., and M. Sommer. 2004. Water redox conditions in wetland soils-their influence on pedogenic oxides and morphology. Soil Science Society of America Journal 68: 326–335. Fiorillo, A. R. 1988. Taphonomy of Hazard Homestead Quarry (Ogallala Group), Hitchcock County, Nebraska. Contributions to Geology, University of Wyoming 26: 57–97. ———. 1991. Taphonomy and depositional setting of Careless Creek Quarry ( Judith River Formation), Wheatland County, Montana, U.S.A. Palaeogeography, Palaeoclimatology, Palaeoecology 81: 281–311. ———. 2004. The dinosaurs of arctic Alaska. Scientific American 291: 84–91. ———. 2006. Review of the Dinosaur Record of Alaska with comments regarding Korean Dinosaurs as comparable highlatitude fossil faunas. Journal of the Paleontological Society of Korea 22: 15–27. ———. 2008. On the occurrence of exceptionally large teeth of Troodon (Dinosauria: Saurischia) from the Late Cretaceous of northern Alaska. Palaios 23: 322–328. ———. In press. Cretaceous dinosaurs of Alaska: Implications for the origins of Beringia. In R. B. Blodgett and G. Stanley, eds., The Terrane Puzzle: New Perspectives on Paleontology and Stratigraphy from the North American Cordillera. Geological Society of America Special Paper 442.

474 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

Fiorillo, A. R., and R. A. Gangloff. 2000. Theropod teeth from the Prince Creek Formation (Cretaceous) of northern Alaska, with speculations on arctic dinosaur paleoecology. Journal of Vertebrate Paleontology 20: 675–682. ———. 2001. The caribou migration model for Arctic hadrosaurs (Ornithischia: Dinosauria): A reassessment. Historical Biology 15: 323–334. ———. 2003. Preliminary note on the taphonomic and paleoecologic setting of a Pachyrhinosaurus bonebed in northern Alaska. Journal of Vertebrate Paleontology 23(3, Suppl.): 50A. Fitzpatrick, E. A. 1993. Soil Microscopy and Micromorphology. Chichester: John Wiley & Sons. Flaig, P. P., P. J. McCarthy, A. R. Fiorillo, and E. Brandlen. 2007. Alluvial facies and facies associations within the Late Cretaceous Prince Creek Formation, North Slope, Alaska. Geological Society of America Abstracts with Programs 39: 629. Frederiksen, N. O. 1990. Pollen zonation and correlation of Maastrichtian marine beds and associated strata, Ocean Point dinosaur locality, North Slope, Alaska. U.S. Geological Survey Bulletin 1990-E: 1–24. Frederiksen, N. O., T. A. Ager, and L. E. Edwards. 1988. Palynology of Maastrichtian and Paleocene rocks, lower Colville River region, North Slope of Alaska. Canadian Journal of Earth Sciences 25: 512–527. Frederiksen, N. O., and D. J. McIntyre. 2000. Palynomorph biostratigraphy of mid(?)-Campanian to upper Maastrichtian strata along the Colville River, North Slope of Alaska. U.S. Geological Survey Open-File Report 00-493. Gangloff, R. A., A. R. Fiorillo, and D. W. Norton. 2005. The first Pachycephalosaurine (Dinosauria) from the Paleo-Arctic and its paleogeographic implications. Journal of Paleontology 79: 997–1001. Gradstein, F., J. Ogg, and A. Smith, eds., 2004. A Geologic Time Scale 2004. Cambridge: Cambridge University Press. Hasiotis, S. T., A. R. Fiorillo, and R. R. Hanna. 1999. Borings in Jurassic dinosaur bones: Trace fossil evidence of beetle interactions with vertebrates. In D. D. Gillette, ed., Vertebrate Paleontology in Utah, pp. 193–200. Utah Geological Survey Miscellaneous Publication 99-1. Heritage, G. L, M. E. Charlton, and S. O. Regan. 2001. Morphological classification of fluvial environments: An investigation of the continuum of channel types. Journal of Geology 109: 21–33. Johnson, C. L., and S. A. Graham. 2004. Cycles of perilacustrine facies of Late Mesozoic rift basin, southeastern Mongolia. Journal of Sedimentary Research 74: 786–804. Kirschbaum, M. A., and P. J. McCabe. 1992. Controls on the accumulation of coal and on the development of anastomosed fluvial systems in the Cretaceous Dakota Formation of southern Utah. Sedimentology 39: 581–598. Kraus, M. J. 1987. Integration of channel and floodplain suites, II. Vertical relations of alluvial paleosols. Journal of Sedimentary Petrology 57: 602–617. ———. 2002. Basin-scale changes in floodplain paleosols: Implications for interpreting alluvial architecture. Journal of Sedimentary Research 72: 500–509.

Kraus, M. J., and T. M. Bown.1988. Pedofacies analysis: A new approach to reconstructing ancient fluvial sequences. Geological Society of America Special Paper 216: 143–152. Kraus, M. J., and K. S. Davies-Vollum. 2004. Mudrock-dominated fills formed in avulsion splay channels: Examples from the Willwood Formation, Wyoming. Sedimentology 54: 1127–1144. Kraus, M. J., and B. Gwinn. 1997. Facies and facies architecture of Paleogene floodplain deposits, Willwood Formation, Bighorn Basin, Wyoming, USA. Sedimentary Geology 114: 33–54. Kraus, M. J., and T. M. Wells. 1999. Recognizing avulsion deposits in the ancient stratigraphical record. International Association of Sedimentologists, Special Publication 28: 251–268. Langston, W., Jr. 1967. The thick-headed ceratopsian dinosaur Pachyrhinosaurus (Reptilia: Ornithischia), from the Edmonton Formation near Drumheller, Canada. Canadian Journal of Earth Sciences 4: 171–186. McCarthy, P. J. 2002. Micromorphology and development of interfluve paleosols: A case study from the Cenomanian Dunvegan Formation, NE British Columbia, Canada. Bulletin of Canadian Petroleum Geology 50: 158–177. McCarthy, P. J., E. Brandlen, P. P. Flaig, and A. R. Fiorillo. 2007. Late Cretaceous high latitude paleoenvironments at the KikakTegoseak dinosaur site, Prince Creek Formation, North Slope, Alaska. Geological Society of America Abstracts with Programs 39: 506. McCarthy, P. J., U. F. Facinni, and A. G. Plint. 1999. Evolution of an ancient coastal floodplain: Paleosols, interfluves and alluvial architecture in a sequence stratigraphic framework, Cenomanian Dunvegan Formation, NE British Columbia, Canada. Sedimentology 46: 861–891. McCarthy, P. J., I. P. Martini, and D. A. Leckie. 1997a. Pedosedimentary history and floodplain dynamics of the Lower Cretaceous upper Blairmore Group, southwestern Alberta, Canada. Canadian Journal of Earth Sciences 34: 598–617. ———. 1997b. Anatomy and evolution of a Lower Cretaceous alluvial plain: Sedimentology and paleosols in the upper Blairmore Group, south-western Alberta, Canada. Sedimentology 44: 197–220. ———. 1998. Use of micromorphology for interpretation of complex alluvial paleosols: Examples from the Mill Creek Formation (Albian), southwestern Alberta, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology 143: 87–110. McCarthy, P. J., and A. G. Plint. 2003. Spatial variability of paleosols across Cretaceous interfluves in the Dunvegan Formation, NE British Columbia, Canada: Palaeohydrological, palaeogeomorphological and stratigraphic implications. Sedimentology 50: 1187–1220. McKeague, J. A. 1983. Clay skins and argillic horizons. In P. Bullock and C. P. Murphy, eds., Soil Micromorphology. Vol. 2: Soil Genesis, pp. 367–388. Berkhamstead, England: AB Academic. Moore, T. E., W. K. Wallace, K. J. Bird, S. M. Karl, C. G. Mull, and J. T. Dillon. 1994. Geology of northern Alaska. In G. Plafker and H. C. Berg, eds., The Geology of Alaska, pp. 49–140. The Geology of North America, Vol. G-1. Boulder: Geological Society of America. Mull, C. G. 1985. Cretaceous tectonics, depositional cycles, and

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 475

the Nanushuk Group, Brooks Range and Arctic Slope, Alaska. In A. C. Huffman, Jr., ed. Geology of the Nanushuk Group and Related Rocks, North Slope, Alaska, pp. 7–36. U.S. Geological Survey, Bulletin 1614. Mull, C. G., D. W. Houseknecht, and K. J. Bird. 2003. Revised Cretaceous and Tertiary Stratigraphic Nomenclature in the Colville Basin, Northern Alaska. U.S. Geological Survey Professional Paper 1673. Muller, R., L. P. Nystuen, and V. P. Wright. 2004. Pedogenic mud aggregates and paleosol development in ancient dryland systems: Criteria for interpreting alluvial mudrock origin and floodplain dynamics. Journal of Sedimentary Research 74: 537– 551. Murphy, C. P. 1986. Thin Section Preparation of Soils and Sediments. Wageningen: Pudoc. Nadon, G. C. 1993. The association of anastomosed fluvial deposits and dinosaur tracks, eggs and nests: Implications for the interpretation of floodplain environments and a possible survival strategy for Ornithopods. Palaios 8: 31–44. ———. 1994. The genesis and recognition of anastomosed fluvial deposits: Data from the St. Mary River Formation, southwestern Alberta, Canada. Journal of Sedimentary Research B64: 451–463. Nanson, G. C. 1980. Point bar and floodplain formation of the meandering Beatton River, northeastern British Columbia. Sedimentology 27: 3–29. Parrish, M. J., J. T. Parrish, J. H. Hutchinson, and R. A. Spicer. 1987. Late Cretaceous vertebrate fossils from the North Slope of Alaska and implications for dinosaur ecology. Palaios 2: 377–389. Perez-Arlucea, M., and N. D. Smith. 1999. Depositional pattern following the 1870s avulsion of the Saskatchewan River (Cumberland Marshes, Saskatchewan, Canada). Journal of Sedimentary Research 69: 62–73. Pielou, E. C. 1994. A Naturalist’s Guide to the Arctic. Chicago: University of Chicago Press. PiPujol, M. D., and P. Buurman. 1994. The distinction between ground-water gley and surface-water gley phenomena in Tertiary palaeosols of the Ebro basin, NE Spain. Palaeogeography, Palaeoclimatology, Palaeoecology 110: 103–113. ———. 1997. Dynamics of iron and calcium carbonate redistribution and palaeohydrology in middle Eocene alluvial palaeosols of the southeast Ebro Basin margin (Catalonia, northeast Spain). Palaeogeography, Palaeoclimatology, Palaeoecology 134: 87–107. Plint, A. G. 1983. Sandy fluvial point-bar sediments from the Middle Eocene of Dorset, England. In J. D. Collinson and J. Lewin, eds., Modern and Ancient Fluvial Systems, pp. 355–368. Special Publication of the International Association of Sedimentologists 6. Retallack, G. J. 2001. Soils of the Past: An Introduction to Paleopedology. London: Blackwell Science. Rich, T. H., R. A. Gangloff, and W. R. Hammer. 1997. Polar Dinosaurs. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, pp. 562–573. San Diego: Academic Press. Rich, T. H., and P. V. Rich. 1989. Polar dinosaurs and biotas of the

Early Cretaceous of southeastern Australia. National Geographic Society Research Reports 5: 15–53. Rich, T. H., P. Vickers-Rich, and R. A. Gangloff. 2002. Polar Dinosaurs. Science 295: 979–980. Richardson, J. L., and M. J. Vepraskas. 2001. Wetland Soils: Genesis, Hydrology, Landscapes and Classification. Boca Raton, Fla.: Lewis Publishers. Rogers, R. R. 1990. Taphonomy of three dinosaur bone beds in the Upper Cretaceous Two Medicine Formation of northwestern Montana: Evidence for drought-related mortality. Palaios 5: 394–413. Rogers, R. R., and S. M. Kidwell. 2007. A conceptual framework for the genesis and analysis of vertebrate skeletal concentrations. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 1–63. Chicago: University of Chicago Press. Sampson, S. D. 1995. Two new horned dinosaurs from the upper Cretaceous Two Medicine Formation of Montana; with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). Journal of Vertebrate Paleontology 15: 743–760. Schaetzl, R., and S. Anderson. 2005. Soils: Genesis and Geomorphology. Cambridge: Cambridge University Press. Smiley, C. J. 1969. Cretaceous floras of the Chandler-Colville region, Alaska: Stratigraphy and preliminary floristics. American Association of Petroleum Geologists Bulletin 53: 482–502. Smith, D. G. 1976. Effect of vegetation on lateral migration of anastomosed channels of a glacial meltwater river. Geological Society of America Bulletin 87: 857–860. ———. 1987. Meandering river point bar lithofacies models: Modern and ancient examples compared. In G. Ethridge, R. M. Flores, M. D. Harvey, and J. N. Weaver, eds., Recent Developments in Fluvial Sedimentology, pp. 83–91. Special Publication of the Society of Economic Paleontologists and Mineralogists 39. Smith, D. G., T. A. Cross, J. P. Dufficy, and S. R. Clough. 1989. Anatomy of an avulsion. Sedimentology 36: 1–23. Smith, D. G., and P. E. Putnam. 1980. Anastomosed river deposits: Modern and ancient examples in Alberta, Canada. Canadian Journal of Earth Sciences 17: 1396–1406. Spicer, R. A. 1987. Late Cretaceous floras and terrestrial environments of northern Alaska. In I. Tailleur, and P. Weimer, eds., Alaska North Slope Geology, pp. 497–512. The Pacific Section, Society of Economic Paleontologists and Mineralogists and the Alaska Geological Society, Vol. 1. ———. 2003. Changing climate and biota. In P. Skelton, R. A. Spicer, S. Kelley, and I. Gilmour, eds., The Cretaceous World, pp. 85–162.Cambridge: Cambridge University Press. Spicer, R. A., and J. T. Parrish. 1990. Late Cretaceous-early Tertiary paleoclimates of northern high latitudes: A quantitative view. Journal of the Geological Society of London 147: 329–341. Stoops, G. 2003. Guidelines for Analysis and Description of Soil and Regolith Thin Sections. Madison: Soil Science Society of America. Tarnocai, C. 1997. Palaeosols of the northern part of North America: Their features and significance as indicators of past climates. In I. P. Martini, ed., Late Glacial and Post-Glacial Environmental Changes, pp. 1–24. Oxford: Oxford University Press.

476 fiorillo, mccarthy, flaig, brandlen, norton, zippi, jacobs, & gangloff

Thomas, R. G., D. G. Smith, J. M. Wood, J. Visser, E. A. CalverlyRange, and E. H. Koster. 1987. Inclined heterolithic stratification—terminology, description, interpretation, and significance. Sedimentary Geology 53: 123–179. Vandenberghe, J. 2003. Climate forcing of fluvial system development: An evolution of ideas. Quaternary Science Reviews 22: 2053–2060. Vepraskas, M. J., L. P. Wilding, and L. R. Drees. 1994. Aquic conditions for soil taxonomy: Concepts, soil morphology and micromorphology. In A. J. Ringrose-Voase and G. S. Humphreys, eds., Soil Micromorphology: Studies in Management and Genesis, pp. 117–132. Amsterdam: Elsevier. Vickers-Rich, P., T. H. Rich, and A. Constantine. 1999. Environmental setting of the polar faunas of southeastern Australia and adaptive strategies of the dinosaurs. In Y. Tomida, T. H. Rich, and P. Vickers-Rich, eds., Proceedings of the Second Gondwanan Dinosaur Symposium, pp. 181–195. National Science Museum Monographs, Tokyo, No. 15.

Wang, C., G. R. Brewster, and K. T. Webb. 1995. Micromorphological evidence of pedogenetic pathway of a Podzolic Gray Luvisol (Falmouth series) in Nova Scotia. Canadian Journal of Soil Science 75: 491–496. Wiggins, V. D. 1976. Fossil oculata pollen from Alaska. Geoscience and Man XV: 51–76. Witte, K. W., D. B. Stone, and C. G. Mull. 1987. Paleomagnetism, paleobotany, and paleogeography of the Cretaceous, North Slope, Alaska. In I. Tailleur and P. Weimer, eds., Alaska North Slope Geology, pp. 571–579. The Pacific Section, Society of Economic Paleontologists and Mineralogists and the Alaska Geological Society, Vol. 1. Woo, M. K., and T. C. Winter. 1993. The role of permafrost and seasonal frost in the hydrology of northern wetlands in North America. Journal of Hydrology 141: 5–31.

Paleontology and Paleoenvironmental Interpretation of the Kikak-Tegoseak Quarry 477

31 Taphonomy of Horned Dinosaurs (Ornithischia: Ceratopsidae) from the Late Campanian Kaiparowits Formation, Grand Staircase–Escalante National Monument, Utah M I C H A E L A . G E T T Y, M A R K A . L O E W E N , E R I C R O B E R T S , ALAN L. TITUS, AND SCOTT D. SAMPSON

a collaborative paleontological survey conducted

of these taxa with others found in coeval formations in

by the Utah Museum of Natural History and the Bureau

Alberta and Montana.

of Land Management in the Upper Cretaceous Kaiparowits Formation, in Grand Staircase–Escalante National Monument, southern Utah, has yielded remains of

Introduction

at least three new ceratopsid dinosaur taxa from multiple

Since 2000, the Utah Museum of Natural History (UMNH) and

localities. Significant ceratopsid specimens excavated

the Bureau of Land Management (BLM) have been conducting

and collected from these localities provide the basis of

an extensive paleontological survey of the late Campanian

a taphonomic study of ceratopsian deposition in the

Kaiparowits Formation in Grand Staircase–Escalante National

Kaiparowits Formation. Most ceratopsian remains in the

Monument (GSENM). This survey has revealed a unique dino-

formation represent isolated specimens preserved by

saur fauna that includes three new taxa of ceratopsians (Samp-

rapid burial within fluvial channels and are likely to

son and Loewen this volume). Previous to the joint survey,

be preserved as associated or partially articulated speci-

ceratopsian remains from the Kaiparowits Formation had

mens. Rare specimens preserved in fine-grained over-

been limited to isolated teeth sampled primarily through

bank or pond sediments are much more likely to be

screening for microvertebrate remains (Parrish and Eaton

disarticulated and decomposed than those in fluvial

1991; Eaton et al 1999). In the six seasons of the survey,

channel settings. The Kaiparowits Formation preserves

UMNH crews have identified a range of macrovertebrate lo-

no evidence of mass monodominant centrosaurine

calities preserving significant, diagnostic remains of cera-

bonebeds like those found in the coeval Dinosaur Park

topsid dinosaurs including a number of mostly complete

and Two Medicine Formations; however, a single mono-

skulls. Although some sites have required only minor surface

dominant chasmosaurine bonebed at UMNH locality

collections, several ceratopsian specimens were fairly com-

942 may be due to behavioral or paleoenvironmental in-

plete and required large excavations. The largest excavation

fluences, and may represent a link with other Campa-

produced bones across 29 square m and required three field

nian chasmosaurine localities in the southern Western

seasons of work. These new localities provide significant in-

Interior Basin. Trends in taphonomic modes and signa-

sight into both the taphonomy and diversity of ceratopsian

tures from ceratopsian localities in the Kaiparowits For-

remains in the Kaiparowits Formation. This report assesses

mation provide useful means of comparing preservation

the taphonomic and sedimentologic data for each major cera-

478

topsian locality (n = 12) documented by the UMNH during an

tyrannosaurid and maniraptoran theropods, hypsilophodont

8-year fossil survey of the formation. This documentation en-

ornithopods, nodosaurine ankylosaurids, lambeosaurine and

ables us to examine trends in taphonomic modes, faunal asso-

hadrosaurine hadrosaurids, and centrosaurine and chasmo-

ciations, bone modification features, and other important en-

saurine ceratopsids. Preservation of some specimens is excep-

vironmental indicators associated with ceratopsid remains in

tional, including over two dozen partial skeletons and skulls

the Kaiparowits Formation. Sampson and Loewen (this vol-

(e.g., hadrosaurs, ceratopsids, tyrannosaurids), some of which

ume) provide a brief systematic and stratigraphic discussion of

include integumentary impressions (hadrosaurs and ceratop-

these new taxa. Detailed systematic descriptions will be pub-

sids). From its inception, the KBP has also attempted to bet-

lished in future work.

ter refine the stratigraphic framework for the target formations so that the newly documented fossil sites could be placed

KAIPAROWITS BASIN PROJECT OVERVIEW

into a well-constrained stratigraphic and paleoenvironmental context. Thus, geologic/stratigraphic research has paralleled

In 2000, the Kaiparowits Basin Project (KBP) was initiated for

the paleontological work, providing key insights into local

the purpose of collecting and researching Late Cretaceous ver-

and regional chronostratigraphic and paleoenvironmental

tebrate fossils from the Kaiparowits Basin of GSENM. In par-

relationships. One of the most important milestones of geo-

ticular, the KBP has focused on the Campanian-aged non-

logic research relevant to the region’s vertebrate paleontology

marine Wahweap and Kaiparowits formations, which crop

was the discovery of multiple volcanic ash (bentonite) hori-

out extensively within the Kaiparowits Plateau (Fig. 31.1A).

zons at multiple stratigraphic levels within the Kaiparowits

To further the aims of the KBP as well as glean crucial man-

and Wahweap Formations. Radiometric analysis (Roberts et al.

agement data on the spatial and stratigraphic distribution of

2003; Roberts 2005; Jinnah et al. 2007) has provided the tight

significant vertebrate fossils sites, in 2001 the UMNH estab-

chronostratigraphic controls indicating that the Kaiparowits

lished a unique collaborative agreement with the Bureau of

Formation was deposited over approximately a 2-million-year

Land Management, the federal agency managing GSENM. In

interval, between 76 and 74 Ma, while the bulk of the Wah-

the first 5 years of the project (2001–2006), UMNH teams

weap Formation is apparently younger than 81 Ma. The Kai-

have logged a total of 392 days of fieldwork on this project,

parowits Formation thus correlates with the fossiliferous inter-

amounting to 2,635 person days (with one person day equal-

vals of other, more well-known upper Judithian dinosaur-

ing an 8-hour work day for one person), for a grand total of

bearing formations to north (e.g., the Dinosaur Park, Two

about 21,080 person hours in fieldwork alone. During the

Medicine, and Judith River Formations) and the Wahweap

same period BLM crews logged slightly less than one-third of

Formation is a Foremost-Oldman equivalent. Therefore, the

that amount and the area surveyed in both the Wahweap and

significance of a new dinosaur fauna from southern Utah is

Kaiparowits Formations by both teams totaled 37,000 acres. A

heightened by the fact that it can be compared directly with

total of over 1,000 vertebrate localities have been discovered,

penecontemporaneous macrovertebrate faunas that include a

mapped, and placed within a rigid stratigraphic framework

diversity of dinosaurs.

tied to dated ash beds, and many macrofossil sites have now been excavated. Preliminary results of the KBP have been abundant and spectacular, resulting in the identification of an essentially

GEOLOGICAL AND PALEOENVIRONMENTAL SETTING

new North American Late Cretaceous vertebrate macrofauna

The Kaiparowits Formation is characterized by 850 m of light

(Titus et al. 2001, 2005; Kirkland et al. 2002; Sampson et al.

grey, clastic strata that crop out in badland-style exposures

2002, 2004; Getty et al. 2003; Gates 2004; Wiersma et al. 2004;

across much of the Kaiparowits Plateau. These sediments were

Zanno and Sampson 2005; Roberts 2005, 2007; Roberts et al.

deposited as part of a proximal, prograding clastic wedge

2005, 2007; Roberts and Tapanilla 2006; Gates and Sampson

within the syn-evolving Sevier foreland basin (Goldstrand

2007; Gates et al. in press; Zanno et al. in press). By far the most

1992). Sandstone provenance in the Kaiparowits Formation

significant and stunning results to date relate to dinosaurs. A

(Goldstrand 1992) reveals that sediments were derived from

1997 review of fossil vertebrates from the Kaiparowits Forma-

the foreland fold and thrust belt at a point source located 300–

tion (Eaton et al. 1999) listed the presence of eight different

500 km to the southwest of present outcrops during the Late

dinosaur taxa, all known only from fragmentary remains that

Cretaceous. The formation was named by Gregory and Moore

prohibited confident genus and species identifications. Since

(1931), but a type section and detailed description of the for-

2000, the KBP has documented the occurrence of 16 nonavian

mation was never presented. Subsequent geologic investi-

dinosaur taxa in this unit, doubling the previous estimate in

gations by Lohrengel (1969), Eaton (1991), Little (1995), and

less than a decade. Over 70 partial to nearly complete dino-

Roberts (2007) demonstrate that the formation was depos-

saur skeletons have now been recovered, including remains of

ited across a broad, flat alluvial plain by fluvial and flood-

Taphonomy of Horned Dinosaurs from the Late Campanian Kaiparowits Formation 479

FIGURE 31.1.

Location map and stratigraphic column pertaining to the Kaiparowits Formation, GSENM, southern Utah. (A) Location map of the Kaiparowits Formation within GSENM; (B) stratigraphic column of the Kaiparowits Formation showing the distribution of UMNH ceratopsian localities and dated ash beds.

plain processes. Although the formation appears to be mud-

sandstone with interbedded mudstone; (4) siltstone and mud-

dominated, in part because of its slope-forming expression,

stone; and (5) carbonaceous mudstone.

the sandstone to mudstone ratio is 60/40.

Pebble conglomerates (1) are composed almost exclusively

Five of the most common fossil-bearing lithofacies docu-

of intra-formational siltstone and mudstone pebbles, with

mented in the Kaiparowits Formation are (1) pebble conglom-

minor bone pebbles, invertebrate shells (esp. unionid bi-

erate; (2) major lenticular and tabular sandstones; (3) minor

valves), and pedogenic carbonate nodules. Beds are com-

480 getty, loewen, roberts, titus, & sampson

monly lenticular, ranging in thickness from 0.1 to 3 m and they are generally less than 25 m in lateral extent. Pebble conglomerates are most commonly interpreted as thalwag deposits in fluvial channels. Major sandstones (2) are typically fine-to-medium grained, composed of texturally and compositionally immature sand. Isolated bone, pebble, and wood (including large trees with entire root balls attached) are common in these deposits. Beds range from tabular to lenticular in shape and commonly extend laterally for 50–100 m or greater. Individual bed thicknesses within the Kaiparowits Formation range from 1.5 to ]20 m, and lateral accretion, trough and tabular crossstratification, and planar and convoluted bedding are all common. Major sandstone deposits are interpreted to represent meandering to anastomosing fluvial channel deposits. Minor sandstones with interbedded mudstone (3) facies are also common, typically ranging from 2 to 20 m thick and tens of meters to kilometers in lateral extent. Planar and ripple laminations are common, along with evidence of moderate to intense bioturbation including rhizoturbation. The minor sandstone with interbedded mudstone facies is interpreted to represent a variety of environments including levee deposits, crevasse splay and channel deposits, and channel fill deposits. Exceedingly poor induration of these deposits makes precise paleoenvironmental interpretations difficult in many cases. The siltstone and mudstone (4) facies are typically composed of siltstone, silty mudstone, sandy mudstone, or muddy sandstone beds, which range between 0.3 and 7 m thick and often exceed tens to hundreds of meters in lateral extent. Beds are interpreted as floodbasin deposits including pond, lake, paleosol, channel-fill, and altered ash deposits. Paleosols observed in the formation are poorly developed, with minor incipient caliche, slickensides, gleying, and very weak color banding. These features, in addition to consistent drab, grey-green colors, are indicative of hydromorphic soils, suggestive of a relatively wet, subtropical environment (Roberts 2007). Climatic and environmental interpretations based on paleosols are supported by sedimentary evidence for abundant floodplain pond deposits and a preponderance of aquatic/ semi-aquatic vertebrate and invertebrate taxa recorded in the formation. Carbonaceous mudstones (5) are less common than the

FIGURE 31.2. Pie charts depicting (A) relative frequencies of taphonomic modes for 523 vertebrate localities in the Kaiparowits Formation, and (B) relative abundance of major vertebrate clades identified from 90 associated and articulated localities within the Kaiparowits Formation.

other major facies; however, they provide significant insights into the depositional and paleoenvironmental history of the Kaiparowits Formation This facies is typically characterized by

Analysis of data collected during the KBP fossil survey indi-

finely laminated (occasionally massive) dark brown, carbona-

cates that 23% of fossil localities (n = 523) in the Kaiparowits

ceous claystone beds, which range in thickness from 0.3 to

Formation contain aquatic/semi-aquatic taxa (e.g., turtles,

0.75 m and are typically between 10 and 300 m in lateral

fish, crocodiles, Fig. 31.2B). Initial analysis of paleobotanical

extent. Individual beds vary significantly with regard to iden-

data also demonstrates a rich and diverse flora dominated by

tifiable macroscopic organic material. Some beds are com-

angiosperms, including taxa that are also suggestive of a

pletely devoid of macrofossils, whereas others contain abun-

humid climate (e.g., ferns; R. Barclay pers. com.).

dant carbonized plant fragments, freshwater mollusks, amber, and rare vertebrate fossils (fish scales and teeth).

The formation is subdivided into three informal stratigraphic packages, reflective of gross trends in sedimentology

Taphonomy of Horned Dinosaurs from the Late Campanian Kaiparowits Formation 481

and fossil preservation (Roberts 2007). The lower unit (0–120

taphonomic signature of a locality therefore requires a de-

m) is characterized by an abundance (]60%) of major tabular

tailed examination of all specimens associated with the lo-

sandstones. Less abundant, but still common are minor sand-

cality, which can only be conducted on fully prepared ma-

stones with interbedded mudstones and the siltstone and

terial. As a result, we have focused our efforts to describe

mudstone facies. The middle unit (170–550 m level) is consid-

taphonomic signatures primarily on the material recovered

erably more mud-rich than either the lower or upper units,

from UMNH VP locality 145. Many of the other localities re-

characterized by abundant siltstone and mudstone, minor

ported herein were collected during the 2006 and 2007 field

sandstone with interbedded mudstone, and non-coalified

season, and the resulting specimens have not been sufficiently

organic mudstone facies. Major sandstone and pebble con-

prepared for detailed taphonomic analysis. All specimens col-

glomerate facies are less common. The upper unit (550–850 m)

lected at UMNH VP locality 145 were prepared and examined

is similar in character to the lower facies with an abundance of

for bone modification and other taphonomic features, includ-

major sandstone facies; however, interbedded fines are also

ing relative degree of bone weathering and abrasion, pre-

common. The upper unit is also characterized by greater indu-

burial breakage, carnivore tooth traces, trample traces, insect

ration and steeper topography than the lower and middle

traces, and bone decomposition. This data formed the basis

units. The distribution of UMNH ceratopsian localities within

of a detailed taphonomic analysis of an individual ceratop-

the stratigraphic column shows a distinct concentration of

sian locality that will serve as a template for future analysis

localities in the lower part of the formation (Fig. 31.1B).

of the remaining ceratopsian and other significant vertebrate localities.

Methods Taphonomy

This study is based on the results of both fieldwork and laboratory analyses conducted between 2001 and 2007. Field-based

This report assesses the taphonomic and sedimentologic data

research included documentation of the stratigraphic posi-

for each major ceratopsian locality (n = 12) documented by

tion, lithofacies, and paleoenvironmental context of the most

the UMNH during an 8-year fossil survey of the formation.

significant fossil localities, and identification of the tapho-

This documentation enables us to examine trends in tapho-

nomic characteristics of vertebrate fossil remains preserved in

nomic modes, faunal associations, bone modification features

each locality. For localities requiring significant excavation,

and other important environmental indicators associated

all specimens greater than 10 cm2 were mapped using a 1 me-

with ceratopsid remains in the Kaiparowits Formation.

ter grid square. A detailed map of all elements was constructed and used to examine the spatial distribution of each skeleton. The data collected from field-based research was used as the

TAPHONOMIC MODES

basis of determining the taphonomic mode of each locality.

Five primary taphonomic modes of fossil preservation are

The concept of taphonomic modes has been used by numer-

identified in association with ceratopsian remains in the Kai-

ous researchers to characterize vertebrate fossil localities and

parowits Formation:

their paleoenvironmental context, beginning with seminal papers by Behrensmeyer (1975) and Dodson et al. (1980). The

Mode A: articulated macrosites

definition we use here for taphonomic modes is based on

Mode B: associated macrosites

Eberth and Currie (2005), which is in turn developed from

Mode C: isolated macrosites

previous work by several authors, particularly associated with

Mode D: macrofossil bonebeds

work in Dinosaur Provincial Park (Dodson 1971; Wood et al.

Mode D1: monodominant macrofossil bonebeds

1988). Taphonomic modes are defined as the recurring preser-

Mode D2: multitaxic macrofossil bonebeds

vational and taphonomic features and geological associations

Mode E: vertebrate microfossil assemblages

among fossils from a stratigraphic interval that reflect pre- and postmortem influences (biotic, environmental, and diagene-

We have characterized every vertebrate locality from which

tic), as well as basin-scale controls on sediment budget and

the UMNH collected any specimens, regardless of taxa, by pre-

accommodation.

dominant taphonomic mode. The relative frequency of each

Laboratory-based investigations were used to discern the

taphonomic mode among a total 523 vertebrate localities rep-

taphonomic signature of a given vertebrate locality. The taph-

resenting all taxa provides an estimate of the relative fre-

onomic signature of a locality has been defined as the sum of

quency of each taphonomic mode in the formation (Fig.

preservational and taphonomic features and geological asso-

31.2A). Here we present the definitions and relative frequency

ciations exhibited by each specimen from an individual lo-

of each taphonomic mode for the Kaiparowits Formation gen-

cality (Eberth and Currie 2005). A complete analysis of the

erally, as well as the relative frequency of diagnostic cera-

482 getty, loewen, roberts, titus, & sampson

Table 31.1. Taphonomic Characteristics of Associated Ceratopsian Localities in the Kaiparowits Formation

sites are variable but were found to be less singnificant than in associated and isolated macrosites. One partially articulated specimen deposited in sandstone preserved skin impressions

UMNH

Lithofacies

Taphonomic

New

Approx.

Skeletal

Loc. #

type

mode

Taxon

age

elements

945

4

A

A or B

j

sk,ax,ap

890

3

A

A

a

sk,ax,ap

450

2

A

B

sa

sk

mens are a relatively uncommon taphonomic mode in the

940

3

A

C

sa

sk,ax,ap

formation, comprising about 14% of vertebrate localities (Fig.

951

4

B

A

sa

sk

145

5

B

B

a

sk,ax,ap

960

5

B

B

a

sk

277

3

B

B

sa

sk

512

4

B

?

sa

ax,ap

925

4

B

?

j

ax

identified and excavated in the Kaiparowits Formation, indi-

662

3

C

B

a

sk

cating that they are nearly twice as abundant as articulated

684

4

C

B

a

sk

ceratopsian remains. Associated specimens are most abun-

942

1

D

B

a,sa,j

sk,ax,ap

dant in siltstone and mudstone facies deposited in a variety of

Abbreviations: a: adult; sa: subadult; j: juvenile; sk: skull; ax: axial; ap: appendicular; 1: conglomerate; 2: major sandstone; 3: minor sandstone interbedded with mudstone; 4: fine grain siltstone/mudstone; 5: organic rich mudstone.

over an articulated ceratopsid forearm, indicative of the high degree of preservation generally associated with rapid postmortem burial. Taphonomic Mode B: Associated Specimens. Associated speci-

31.2A), but represent a significant portion of materials collected by the UMNH. Associated specimens consist of the disarticulated remains of a single individual skeleton. Between 2000 and 2007, seven associated ceratopsian skeletons were

overbank environments such as floodbasin ponds and paleosols. Evidence of low to moderate bone weathering (Stage 1– 2; sensu Fiorillo 1987, 1988) and possible subaqueous decay is common. Associated macrosites commonly host a greater abundance and diversity of bone modification and taphonomic features than do other types of fossil concentrations in

topsian remains associated with each mode. The subsequent

the formation. The high degree of postmortem modification

section describes each individual ceratopsian locality based on

associated with disarticulated specimens, including the sig-

associated lithofacies, as well as taphonomic modes and signa-

nificant loss of diagnostic elements from the site through fac-

tures. The general taphonomic characteristics of the 12 most

tors such as hydraulic winnowing, scavenging, decomposi-

diagnostic ceratopsian localities are presented in Table 31.1.

tion, and skeletal reworking, has resulted in our inability to

Taphonomic Mode A: Articulated Macrosites. The rarest taphonomic mode observed in the Kaiparowits Formation is articu-

determine the taxonomic affinity of at least three partial skeletons beyond the family level.

lated macrosites, making-up only about 3% of all sites (n =

Taphonomic Mode C: Isolated Elements. Isolated elements

523). Although uncommon, they are particularly significant

consist of single bones that are not associated with skeletons

for the anatomical and taphonomic details that are preserved.

or bonebeds. They are relatively common in the formation,

They are defined here as fossil localities containing partial

but since they are rarely collected, it is difficult to accurately

to fully articulated skeletons with at least four elements in ar-

determine their relative abundance to articulated or associ-

ticulation, showing natural anatomical arrangement. From

ated remains. Several isolated ceratopsian elements have been

2000 and 2007, only four localities comprised of articulated

collected from the Kaiparowits Formation, but only one has

ceratopsian remains were discovered in the Kaiparowits For-

had sufficient morphological characteristics to be identified at

mation by UMNH crews. Three of the four articulated cera-

the species level.

topsians were discovered in sandstone facies, interpreted as

Taphonomic Mode D: Macrofossil Bonebeds. Macrofossil bone-

fluvial sandstone channel deposits, indicating that articulated

beds consist of concentrations of bone elements from multi-

remains are more commonly associated with rapid burial in

ple individuals from one or more taxa, which may be articu-

the formation. Interestingly, all of the articulated sites in

lated or disarticulated, and are dominated by specimens larger

sandstone contained less than 50% of the original skeleton.

than 5 cm in length. They may be defined as monodominant

One articulated ceratopsian was found associated within

(Mode D1), consisting of the majority of remains (]50%)

muddy siltstone to fine-grained sandstone and was identified

from a single taxon, or multitaxic (Mode D2), consisting of

in the field as more than 70% articulated, which is signifi-

the remains of many taxa, and not dominated by a single one

cantly more complete and articulated than any remains asso-

(sensu Eberth and Currie 2005; Eberth and Getty 2005).

ciated with fluvial channel deposits. Preservational quality of

Monodominant ceratopsian bonebeds appear to be excep-

articulated macrosites is typically much better than those

tionally rare in the Kaiparowits Formation, with only one

found in other taphonomic modes. The degree of bone weath-

such locality identified from 2000 to 2007. Ceratopsian re-

ering and the abundance of other bone modifications in these

mains in association with multitaxic bonebeds are also quite

Taphonomy of Horned Dinosaurs from the Late Campanian Kaiparowits Formation 483

rare and have only been identified with one locality during

each locality are summarized in Table 1. Several of the locali-

the study period. The vast majority of macrofossil assemblages

ties include highly diagnostic remains that allow identifica-

are multitaxic mixed terrestrial-aquatic bonebeds consisting

tion of three distinct ceratopsian taxa labeled here as Taxon A,

primarily of accumulations of unassociated isolated bones

B, and C (pending detailed morphological descriptions).

from various taxa reworked into paleo-channel lags, which rarely produce any material diagnostic beyond family level. Taphonomic Mode E: Vertebrate Microfossil Assemblages. Mi-

KAIPAROWITS NEW TAXON A

crosites represent concentrations of small (typically [10 cm),

This taxon has thus far been identified from two localities,

unassociated elements typically dominated by teeth, scutes,

both discovered by UMNH/BLM volunteer Scott Richardson

fish scales, other small bones and bone fragments, and are

during the 2006 and 2007 field seasons. The animal, which

commonly characterized by high taxonomic diversity. They

pertains to the ceratopsid subclade Chasmosaurinae, pos-

are the second most common taphonomic mode of preser-

sesses numerous autapomorphic characters on the skull in-

vation observed in the formation, representing 23% of total

cluding ten elongate, rostrally curving epiparietal hooks

fossil localities (n = 523). Microsites are most commonly asso-

spread over a very broad and almost completely solid parieto-

ciated with the pebble conglomerate and non-coalified, carbo-

squamosal frill.

naceous mudstones (lithofacies 4). These facies are interpreted

UMNH VP Locality 890. This specimen consists of a relatively

as hydraulically sorted, fluvial lag deposits and condensed,

complete skull missing only a small part of the left side of the

floodplain pond deposits, respectively (Roberts 2007). The

face and frill (parts of the jugal, squamosal, and parietal).

only ceratopsian remains associated with this taphonomic

These missing elements were lost to erosion prior to discovery

mode are teeth and broken tooth fragments, which cannot be

of the locality in 2006. A significant portion of the axial skele-

identified beyond the family level (Parrish 1999).

ton from neck to tail was found more or less articulated with the skull, including part of the pelvic girdle and at least part of

CERATOPSIAN LOCALITIES IN THE KAIPAROWITS FORMATION

one limb (Fig. 31.3A). The locality is characterized by taphonomic mode A. This specimen was found in a silty sandstone channel facies (lithofacies type 3) and is consistent with an

From 2000 to 2007, a total of 90 localities were identified by

individual animal carcass washed into a river channel and

UMNH crews as containing associated and articulated verte-

buried quickly. The limbs and distal tail may have been lost

brate remains diagnosed to at least the clade level recognized

either to scavenging or to rotting of the carcass prior to its

as family. Twelve of these sites have been identified as con-

deposition. As most of the postcranial material in still in prep-

taining ceratopsians, indicating that horned dinosaur re-

aration, we have not been able to examine the skeleton for

mains comprise approximatey 14% of associated and articu-

signs of predation or scavenging prior to its final burial.

lated vertebrates preserved in the formation (Fig. 31.2B). This

UMNH VP Locality 951. In 2007, a second specimen of Kai-

provides a rough estimate of the relative abundance of cera-

parowits taxon A was discovered and excavated. This speci-

topsian remains in relation to other macrovertebrate remains

men consists primarily of a disarticulated skull of a subadult

in the Kaiparowits fauna, yet may also reflect preservational

individual found scattered over an area of about 3 square m

biases inherent in the data. For example, the fact that the two

(Fig. 31.3B). The site is characterized as taphonomic mode B.

most abundant groups in the formation (hadrosaurs and cera-

Most of the skull was recovered with the exception of the

topsians) also include the most robust fossil elements may

premaxilla, rostral and predentary. In addition to a high de-

result in their remains being relatively over-represented in the

gree of disarticulation, many parts of the skull were found to

fossil record.

be severely broken predepositionally, indicating that the spec-

All of the ceratopsian localities identified in the Kaiparowits

imen was completely skeletonized and decomposed prior to

Formation to date have been excavated, mapped, and col-

its final burial in silty mudstone facies (lithofacies type 4). A

lected. The geological context of each locality has been estab-

more extensive examination of taphonomic modification will

lished including stratigraphic position and sedimentological

be conducted once the specimen is fully prepared.

characteristics. Here we present the general taphonomic characteristics of each ceratopsian locality containing diagnosable skeletal remains, organized by taxa. These descriptions pro-

KAIPAROWITS NEW TAXON B

vide our best assessment of taphonomic mode and signature

This taxon was first discovered in fall of 2000 and has sub-

of these localities observed to date, but most of the localties

sequently been found to be the most abundant ceratopsian

will require more preparation before the taphonomic signa-

found in the monument, currently represented in a total of 6

ture can be studied in detail. The predominant lithofacies is

localities. Kaiparowits new taxon B represents a large chas-

also described and the general taphonomic characteristics of

mosaurine with short, laterally projecting orbital horns and

484 getty, loewen, roberts, titus, & sampson

FIGURE 31.3.

Localities representing Kaiparowits new taxon A. Each square represents 1 square meter. (A) Partial articulated skeleton and skull from UMNH VP locality 890; (B) disarticulated skull of a subadult individual from UMNH VP locality 951.

an elongate frill with a deep, recurved embayment. One or

disarticulated microvertebrate remains characteristic of taph-

more diagnostic elements of this taxon have been identified in

onomic mode E. Since no macrovertebrate remains were re-

the following localities.

covered other than the one ceratopsian specimen, we elected

UMNH VP Locality 145 (the Blues Ceratopsian, UMNH VP

for the purposes of this discussion to characterize it as tapho-

12198). This locality is the site of the initial discovery of this

nomic mode B; however, the fossil accumulation is more com-

taxon, which was collected and prepared from 2000 to 2004.

plex than this simple diagnosis indicates.

This locality represents the most extensive ceratopsian excava-

Once prepared and identified, all of the recognizable ele-

tion conducted thus far in the Kaiparowits Formation. Active

ments collected from the locality were tallied and compared

quarrying for several weeks over three seasons resulted in the

to the expected number from a complete skeleton in order to

recovery of more than 280 elements and fragments from one

determine the relative abundance of elements expected from

associated chasmosaurine individual spread over an area of

each part of the body (Figs. 31.4A, B, 31.5B). Identifiable ele-

approximately 29 square m (Fig. 31.4A). This locality preserves

ments were categorized according to four skeletal compo-

the skeletal remains of one disarticulated, individual ceratop-

nents (cranial, axial, appendicular and girdle) to approximate

sian skeleton, and is best characterized as taphonomic mode B.

the Voorhies transport groups used in previous taphonomic

However, the locality also produced a significant quantity of

analyses (Voorhies 1969; Lehman 1982; Ryan 1992). Axial and

Taphonomy of Horned Dinosaurs from the Late Campanian Kaiparowits Formation 485

FIGURE 31.4.

Localities representing Kaiparowits new taxon B. Each square represents 1 square meter. (A) Partial, disarticulated adult skeleton from UMNH VP locality 145; (B) skeletal elements recovered from UMNH VP locality 145. Elements shaded based on assignment to one of four skeletal classes (cranial, axial, appendicular, and girdle); (C) UMNH VP locality 942 monodominant bonebed with MNI of three; (D) partial disarticulated skull from UMNH VP locality 960.

appendicular elements were further divided into proximal

tebrae, and chevrons, indicating that these parts of the body

and distal to differentiate small podial and caudal elements

were absent, most likely due to scavenging or hydraulic re-

from large limbs and vertebrae. In total, we identified 45% of

moval (winnowing) prior to final deposition. Further analysis

the larger postcranial elements and 60% of the cranial ele-

of predepositional bone modifications and stratigraphic data

ments. The majority of missing elements were smaller-sized

revealed a complex taphonomic history of the deposition of

appendicular and axial elements such as distal phalanges, ver-

this skeleton.

486 getty, loewen, roberts, titus, & sampson

FIGURE 31.5. Taphonomic modifications assessed by skeletal region for UMNH VP 12198. (A) Skeletal-region abundances of postmortem decay and element breakage expressed as percentages; (B) relative amounts of all taphonomic modifications from different skeletal regions expressed as percentages; (C) skeletal region abundances of specific bone-surface taphonomic modifications expressed as percentages. app.: appendicular.

All elements of UMNH VP 12198 were prepared and exam-

abraded as it was scavenged and broken by trampling. The

ined macroscopically, and in some cases microscopically, for

axial portions of the skeleton were found to be the most modi-

evidence of decay and breakage (31.5A) and bone surface

fied, including the presence of insect borings (see Roberts

modification (31.5B, C). The relative amounts of postmortem,

2007) and bone weathering typical of subaerial exposure,

bone surface modification by skeletal region (31.5B) and the

which indicate that at least some portion of the axial skeleton

distribution of different kinds of modification (bone weather-

(vertebrae and ribs) were exposed to open air prior to being

ing and abrasion, carnivore tooth traces, trample traces, and

fully immersed in a standing body of water. The subaqueous

insect traces) by skeletal region (31.5C) were also tabulated.

burial interpretation is supported by the sedimentology (in-

Collectively the above data reveal a great deal about the

flux of fine grained pond sediments; lithofacies type 5). Grain

predepositional history of this specimen. A considerable num-

size decreases from muddy sandstone basally to a muddy silt-

ber of tooth-marked bones, in combination with a number of

stone (encasing the bones) and a capping silty claystone.

shed theropod teeth, indicate that the specimen was predated

Moreover, aquatic gastropod and small, thin shelled (quiet

upon and/or scavenged prior to decomposition. Very limited

water) bivalve numbers increase dramatically within the up-

evidence of subaerial weathering was observed and only on

per portion of the unit, with thousands of mollusks and abun-

axial elements, whereas the prevalence of bone decomposi-

dant fish scales preserved in the upper silty claystone interval.

tion on the majority of the skeleton is indicative of significant

Considered together, the taphonomic and sedimentological

subaqueous exposure of the skeleton in a pond environment

evidence suggest that the carcass may have been exposed sub-

prior to burial. The taphonomic signature of bone weathering

aerially along the margin of a floodbasin pond for some ex-

is typified by cracking and flaking of the bone surface caused

tended period of time and subjected to disarticulation, minor

by bones being exposed to the open air prior to burial. This is

bone-weathering, trampling and both carnivore and insect

very distinct from bone decomposition, which is typified by

scavenging, prior to slow burial in a subaqueous environ-

the disintegration of cortical bone, exposing internal trabecu-

ment. It is unclear whether a low-energy flow, perhaps flood-

lar bone, with no sign of surface weathering on the intact

waters associated with a crevasse-splay event, transported the

bone surfaces. Decomposition is prevalent where bones are

remaining carcass a short distance into the pond, or what

exposed in water prior to burial, and is usually very pro-

seems more likely, a regional flooding event occurred that

nounced on the thin walled articular surfaces. The large per-

filled the floodbasin pond and slowly buried the carcass in

centage of skeletal elements exhibiting signs of breakage and

situ. The limited extent of abrasion of the bones and the lack

trample marks further indicates that the bones were weakened

of directional orientation of skeletal elements observable on

by subaqueous decomposition, as well as being trampled and

the quarry map (Fig. 31.4A) supports the latter hypothesis,

broken by dinosaurs or other vertebrates scavenging or other-

although both remain plausible.

wise stepping on the skeleton. The significant breakage of cra-

UMNH VP Locality 942 (Blue Wash Ceratopsian Bonebed). This

nial elements, with few trample marks and some tooth marks,

site preserves partial remains of at least three individuals of

suggests that the skull was likely deposited with much of the

Kaiparowits new taxon B, recognized by distinct size classes of

integument intact, which kept the bone material from being

represented elements. This is the only monodominant cera-

Taphonomy of Horned Dinosaurs from the Late Campanian Kaiparowits Formation 487

topsian bonebed (taphonomic mode D1) identified to date in

formation. UMNH VP 14523 was found in major fluvial chan-

the Kaiparowits Formation The specimens were buried in a lag

nel sandstone (lithofacies type 2) and represents the lowest

deposit at the base of a sandy channel, in a thin, pebbly con-

stratigraphic occurrence of Kaiparowits new taxon B within

glomerate (lithofacies type 1). The skeletons were completely

the Kaiparowits Formation Although the skull appears to have

disarticulated and demonstrate characteristics consistent with

been predepositionally separated from the postcranial skele-

considerable predepositional transport, including winnowing

ton, it represents the only articulated skull known for this

of most small elements and breakage and surface abrasion of

taxon. Its excellent preservation, combined with lack of pre-

preserved elements (Fig. 31.4C). A number of skull elements

depositional breakage, suggest that UMNH VP 14523 repre-

were preserved including a nearly complete parietal of a large

sents the remains of an individual animal washed into a river

adult individual. The proximal portion of the squamosal was

and subject to rapid burial.

also recovered, demonstrating characteristics consistent with

UMNH VP Locality 277 (Associated Orbits and Jugal UMNH VP

Kaiparowits new taxon B. Subadult skull elements recovered

12225). This site preserves one nearly complete left orbit (in-

from the site including two dentaries, one maxilla, two pre-

cluding fused palpebral, postorbital, complete supraorbital

maxillae, and a rostral, are consistent with Kaiparowits new

horncore and partial jugal), and fragmented remains of the

taxon B and may well belong to a single individual. A partial

right supraorbital horncore. UMNH VP 12225 may have ini-

dentary, dorsal vertebrae, and pubis of a much smaller juve-

tially included a much larger portion of the skeleton, as the

nile individual were also recovered, indicating a minimum

recovered material was found on a highly eroded ridge top.

number of three individuals preserved in the site. This locality

The bone elements recovered from this locality were clearly

represents the first and only known ceratopsian bonebed from

disarticulated and characteristic of taphonomic mode B; yet,

the Kaiparowits.

with so much material potentially lost to erosion, it is also

UMNH VP Locality 960 (Adult Partial Skull UMNH VP16784).

possible that the specimen included some articulated ele-

This locality, discovered in 2007, preserves a partial disarticu-

ments. The material represents a single individual and the

lated adult skull of Kaiparowits new taxon B, characteristic of

morphology of the postorbital horncore is consistent with the

taphonomic mode B. Although less than half the skull was

type specimen. The material was deposited in a fine-grained

recovered, a number of diagnostic elements were collected in-

channel sandstone (lithofacies type 3) and was well preserved,

cluding a parietal, squamosal, nasal, orbital series, and brain-

with no crushing or distortion commonly associated with

case (Fig. 31.4C). The skull had become disarticulated prior to

specimens in mudstone.

burial, yet there was minimal predepositional breakage. As a

UMNH VP Locality 662 (Isolated Orbital Horn Core UMNH VP

result, this specimen preserves the most complete diagnostic

16404). This isolated specimen (UMNH VP 16404) consists of

elements found to date for this taxon. The specimen was de-

a single horncore that is also consistent with the type speci-

posited in an organic rich silty mudstone characteristic of

men of Kaiparowits new taxon B, both in terms of morphol-

lithofacies type 5. UMNH VP16784 was initially very well pre-

ogy and orientation of the horncore relative to the orbit. No

served, with almost no indication of abrasion or weathering,

additional materials were found with this specimen, a pattern

yet it has been highly affected by recent erosion and weather-

that is characteristic of taphonomic mode C. The specimen

ing that may ultimately impair our ability to assess the tapho-

was preserved in a very fine-grained channel sandstone typi-

nomic signature of the locality.

cal of lithofacies type 3.

UMNH VP Locality 450 (Subadult Partial Skull UMNH VP

UMNH VP Locality 684 (Partial Parietal UMNH VP 16785).

14523). This locality included an articulated skull of a sub-

This locality preserves the remains of a partial parietal of a

adult individual (UMNH VP 14523), with no associated post-

subadult-sized individual including the right side of the distal

cranial material. The skull pertains to a subadult individual in

parietal bar from the articulation with the squamosal to the

whom individual elements had not completely fused prior to

start of the embayment and epiparietals 2, 3, and 4. Although

death; nevertheless, we regard this specimen as characteristic

most of the embayment has been broken off, the remaining

of taphonomic mode A. If the specimen was a completely

morphological characteristics are very consistent with Kai-

fused adult individual an argument could be made that it was

parowits new taxon B, including evidence for large parietal

merely one element from a disarticulated specimen. Unfortu-

fenestra in the frill, rugosity of the parietal surface, and the

nately, the specimen eroded significantly prior to its discovery

shape and distribution of the epoccipitals. The specimen ap-

and about two-thirds of the skull was lost completely; only the

pears to be one isolated element characteristic of taphonomic

right side of the skull remains, which is mostly intact from the

mode C. The specimen was preserved in mudstone (lithofacies

rostrum to distal parietal. Although the most diagnostic parts

type 4) and appears to have been predepositionally well pre-

of the specimen are absent, the preserved portion of the skull

served and subsequently significantly eroded prior to its dis-

appears morphologically more consistent with Kaiparowits

covery. More material may have been initially present but was

new taxon B than with the other two taxa recognized in the

eroded from the site prior to its discovery.

488 getty, loewen, roberts, titus, & sampson

FIGURE 31.6. UMNH VP locality 940 yielded a nearly complete and partially articulated skull of Kaiparowits new taxon C, as well as an articulated forelimb and associated postcranial elements. Each square represents 1 square meter.

KAIPAROWITS NEW TAXON C

tion. Most of the parietal is also preserved and incorporates a number of fused and unfused epoccipitals. The epiparietals

This taxon is represented by a single specimen from UMNH VP

are small and triangular and do not form any large horns,

locality 940, discovered in the 2006 field season by UMNH

hooks, or other projections associated with other centro-

graduate student and technician Eric Lund. The animal, which

saurine taxa. The narial region including both premaxilla and

is a member of the ceratopsid subclade Centrosaurinae, pos-

most of the nasals is also preserved in articulation, although it

sesses numerous autapomorphic characters on the skull in-

was found broken and displaced from the upper portion of the

cluding laterally directed rostrally curving postorbital horns

skull. The narial region is typical of centrosaurines with the

and a midline epiparietal.

exception of what appears to be a very small nasal horncore.

UMNH VP Locality 940 (UMNH VP 16800). The specimen

Kaiparowits new taxon C represents the first clear evidence

from UMNH VP locality 940 consists of a mostly complete

of a centrosaurine ceratopsian from the Kaiparowits Forma-

partially articulated skull, as well as an articulated forelimb

tion. Although it shares some morphological characteristics

and a small number of axial elements including several ver-

with the recently described Albertaceratops (Ryan 2007) we are

tebrae and ribs (Fig. 31.6). The locality is characterized by

confident that this specimen represents a previously unde-

taphonomic mode A. Upon discovery, a large portion of the

scribed genus and species of Centrosaurinae. The specimen

skull was intact including the midline of the skull from the

was found preserved in a channel sandstone facies (lithofacies

orbits to the distal portion of the frill. The lower portion of

type 3). It represents a single individual and the specimen may

the frill including a large portion of both squamosals was

have been considerably more complete at the time of deposi-

eroded away and lost. The intact portion of the skull contains

tion. Although the skull and forelimb were found articulated

both postorbital horncores, which are distinct from any other

and in close association, most of the preserved postcranial

known centrosaurine in terms of size, curvature, and orienta-

skeleton was disarticulated and displaced prior to its deposi-

Taphonomy of Horned Dinosaurs from the Late Campanian Kaiparowits Formation 489

FIGURE 31.7. Quarry map depicting unidentified ceratopsian skeleton from UMNH VP locality 945 and an associated crocodile skeleton. Each square represents 1 square meter.

tion as indicated by the relatively random (although in situ)

topsian skin impressions. Taphonomically it seems reasonable

distribution of these elements around the skull. Remarkably,

to conclude the carcass was at least partially articulated with

the sandstone entombing the articulated arm preserves sev-

flesh and skin at the time it was washed into a stream bed.

eral patches of skin impressions in direct association with

Some rotting and disarticulation occurred in the river chan-

bony fossils. Skin impressions are relatively common finds in

nel, resulting in winnowing and displacement of most of the

association with articulated hadrosaurs in the formation, but

skeleton and leaving just the forelimb and skull in articulation

this specimen provides the first and only example of cera-

at the time of burial.

490 getty, loewen, roberts, titus, & sampson

NON-DIAGNOSED CERATOPSIAN LOCALITIES

our southern centrosaurine is much less affected by mass mortality events; or ecological differences, by which centrosau-

Three ceratopsian localities excavated by the UMNH in 2006

rines are relatively much less abundant in the Kaiparowits For-

and 2007 have yet to be diagnosed as to the taxa represented.

mation than farther north (supported by the general paucity

Two of these localities (UMNH VP localities 512 and 925; not

of centrosaurine remains in Campanian-aged faunas of the

figured) consist only of postcranial remains, which will be dif-

southern WIB). Chasmosaurine localities in the Kaiparowits

ficult to identify beyond family level. Both localities were

Formation can be compared with both northern and south-

found within a very fine-grained siltstone (lithofacies type 4)

ern formations in the WIB. Whereas most northern taxa are

and regarded as taphonomic mode B, consisting of highly dis-

found solely as isolated specimens, Kaiparowits Taxon B was

articulated specimens with serious taphonomic modifica-

found to occur in at least one monodominant bonebed with a

tions. These specimens are not sufficiently prepared for fur-

small number of individuals. This taphonomic mode com-

ther analysis at this time. Finally, UMNH VP locality 945

pares closely with that of Agujaceratops mariscalensis from the

preserves the remains of a subadult skeleton in nearly com-

Campanian Aguja Formation of Texas, which was found to

plete articulation (taphonomic mode A). The skeleton was

occur in similarly small, monodominant accumulations (Leh-

found preserved together with the remains of a mostly com-

man 1982, 1990). This pattern may indicate shared behavioral

plete, partially articulated crocodile skeleton about 1 m in

characteristics among southern chasmosaurine taxa that con-

length, which appears to have been deposited on top of the

trast with those of their northern counterparts. Other possible

ceratopsian carcass prior to burial (Fig. 31.7). Based on the

considerations include paleoenvironmental variations that

occurrence of two articulated skeletons, both characteristic of

resulted in some southern chasmosaurs being caught in small

taphonomic mode A, the site could also be considered as rep-

group mortality events more frequently. Although it is diffi-

resentative of taphonomic mode D2, a mulitaxic bonebed,

cult to be conclusive in determining the causal mechanisms of

once again demonstrating that assessments of taphonomic

the observed taphonomic differences at this point, this re-

modes are not always straightforward. Although this skeleton

search nevertheless provides a baseline of taphonomic charac-

has not yet been prepared sufficiently for taxonomic identi-

teristics that can lead to future, more detailed analyses.

fication, it appears to preserve sufficient skull material for a

The spatial, stratigraphic, and taphonomic context of sig-

species-level assessment. This specimen represents the most

nificant specimens also significantly aids our ability to inter-

complete articulated skeleton of any animal, found to date, in

pret the paleoenvironmental circumstances by which verte-

the Kaiparowits Formation in a fine-grained facies (lithofacies

brate specimens are preserved. It is interesting to note that

type 4) interpreted as floodplain environment. Generally, ar-

significant specimens have been associated with all tapho-

ticulated specimens in the Kaiparowits Formation are asso-

nomic modes and all lithofacies. The best preserved cera-

ciated with rapid burial in channel sandstone facies, so the

topsian specimens in the Kaiparowits Formation come from

taphonomic signature of this exceptionally well-preserved

fluvial channels associated with lithofacies 3, where deposi-

specimen in mudstone will be very interesting to examine in

tion and burial were presumably rapid. A rapid burial shortens

the future.

the duration that animal remains are exposed to both biological and physical decomposition. Thus, specimens discovered in sandy channel facies are more likely to be articulated and

Discussion

preserve skin impressions. In contrast, specimens preserved in

Working in the relatively unexplored Kaiparowits Formation

the fine-grained floodplain environments, such as Ceratop-

has provided a unique opportunity to examine the taphon-

sian taxon B from UMNH locality 145, suggest a more ex-

omy of nearly all known ceratopsian sites within the for-

tended period of disarticulation and decomposition in wet,

mation. This data is particularly useful for comparing the

muddy conditions. Channel lags and pebble conglomerates

taphonomic contexts of specimens found in the Kaiparowits

associated with lithofacies 1 seem to rarely preserve speci-

Formation with similar aged formations in the Western Inte-

mens, but have been found to produce at least one articulated

rior Basin (WIB). For example, the lack of evidence for cen-

skull. The occurrence of a single monodominant bonebed (Lo-

trosaurine bonebeds contrasts starkly with other late Campa-

cality 942) in a coarse pebble conglomerate indicates that this

nian units in the northern portion of WIB such as the Two

material may be reworked from a previous, possibly larger de-

Medicine and Dinosaur Park formations. Several possible ex-

positional event.

planations may be put forth to account for such a significant

Overall, the association of taphonomic modes with pre-

difference; these include behavioral differences, such as Kai-

dominant lithofacies indicates that articulated remains are

parowits Taxon C being much less gregarious than its north-

much more likely to be associated with major or minor sand-

ern counterparts; paleoenvironmental differences, by which

stones. Moreover, minor sandstones associated with litho-

Taphonomy of Horned Dinosaurs from the Late Campanian Kaiparowits Formation 491

facies type 3 have thus far yielded the best preserved, articu-

rine localities in the southern WIB. Although these results

lated ceratopsian remains in the formation. Ceratopsian speci-

should be regarded as preliminary, the authors plan further

mens associated with finer-grained mudstones and siltstones,

studies of vertebrate taphonomy within the Kaiparowits and

characteristic of lithofacies types 4 and 5, tend to be largely

Wahweap formations.

disarticulated and decomposed. The articulated skeleton from Locality 945 proves to be an exception to this trend, however,

Acknowledgments

reminding us that there is potential for exceptional preserva-

For assistance with fieldwork, we sincerely thank all of the

tion in fine-grained sediments as well as channel sands. Al-

participants of the UMNH field crews in GSENM from 2000 to

most all diagnostic ceratopsian remains were found in associa-

2007, especially Jelle Wiersma, Joe Gentry, Sue Beardesmore,

tion with articulated and associated specimens (taphonomic

and Scott Richardson, who delivered untold weeks of hard

modes A and B). Occasionally, however, diagnostic isolated

labor under difficult conditions. For specimen preparation, we

elements such as uniquely ornamented cranial specimens

thank all volunteers who worked in the UMNH fossil prepara-

were found in both fine- and coarse-grained lithofacies. The

tion lab from 2000 to 2007, in particular Jerry Golden, Sharon

ratio of articulated and disarticulated ceratopsian specimens

Walkington, Jay Green, Walt Elkington, and Elaine Jones. We

was found to be roughly equal in the Kaiparowits Formation,

are grateful for funding of this project, primarily provided

in contrast to taphonomic trends reported for the approx-

by the Bureau of Land Management (Assistance Agreements

imately coeval Dinosaur Park Formation, where Eberth and

JSA015003; JSA071004), as well as the University of Utah. We

Currie (2005) reported a much greater abundance of articu-

also thank the management of Grand Staircase–Escalante

lated versus disarticulated specimens.

National Monument, especially Dave Hunsaker, Harry Barber, Marietta Eaton, and Doug Powell, for ongoing support of the

Summary and Conclusions This study summarizes the current state of taphonomic inves-

project; and Kanab BLM field office manager Rex Smart and Zion National Park Helitack Lead Mike Reid for their support airlifting a number of significant specimens.

tigations of ceratopsian localities in the Kaiparowits Formation Nevertheless, it is important to underscore the fact that this work is ongoing, with several specimens yet to be fully

References Cited

prepared let alone assessed taphonomically. Moreover, if the

Behrensmeyer, A. K. 1975. The taphonomy and paleoecology of plio-pleistocene vertebrate assemblages east of Lake Rudolf, Kenya. Bulletin of the Museum of Comparative Zoology 146: 474–578. Dodson, P. 1971. Sedimentology and taphonomy of the Oldman Formation (Campanian), Dinosaur Provincial Park, Alberta, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology 10: 21–74. Dodson, P., A. K. Behrensmeyer, R. T. Bakker, and J. S. McIntosh. 1980. Taphonomy and paleoecology of the dinosaur beds of the Jurassic Morrison Formation. Paleobiology 6: 208– 232. Eaton, J. G. 1991. Biostratigraphic framework for the Upper Cretaceous rocks of the Kaiparowits Plateau, southern Utah. In J. D. Nations and J. G. Eaton, eds., Stratigraphy, Depositional Environments, and Sedimentary Tectonics of the Western Margin, Cretaceous Western Interior Seaway, pp. 47–61. Geological Society of America Special Paper 260. Eaton, J. G., R. L. Cifelli, J. H. Hutchison, J. I. Kirkland, and J. M. Parrish. 1999. Cretaceous vertebrate faunas from the Kaiparowits Plateau, South Central Utah. In D. Gillette, ed., Vertebrate Paleontology of Utah, pp. 319–321. Utah Geological Survey Miscellaneous Publication 99-1. Eberth, D. A., and P. J. Currie. 2005. Vertebrate taphonomy and taphonomic trends. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 453–477. Bloomington: Indiana University Press. Eberth, D. A., and M. A. Getty. 2005. Ceratopsian bonebeds:

past two seasons are any indication, additional specimens will be recovered in the coming years, resulting in further data for analysis. While it remains difficult in the vast majority of instances to assess cause of death for dinosaur specimens, careful collection and analysis of taphonomic data associated with each locality can reveal a great deal about pre- and postburial events. This information, in turn, leads to better interpretations of the paleoenvironmental context. With regard to the ceratopsian remains thus far collected and studied from the Kaiparowits Formation, it appears that at least three genera and species are represented: two chasmosaurines and one centrosaurine. Most of the remains appear to pertain to isolated specimens preserved by rapid burial within fluvial channels. In contrast to some other formations, ceratopsians within the Kaiparowits Formation appear equally likely to be preserved as associated or partially articulated specimens. However, specimens preserved in fine-grained overbank or ponds are much more likely to be disarticulated and decomposed than those in sandstone. Whereas there is no evidence of mass monodominant centrosaurine bonebeds like those found in the roughly coeval Dinosaur Park and Two Medicine formations, the small monodominant bonebed at UMNH locality 942 may be a behavioral or paleoenvironmental link between Kaiparowits new taxon B and other Campanian chasmosau-

492 getty, loewen, roberts, titus, & sampson

Occurrence, origins and significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 513–536. Bloomington: Indiana University Press. Fiorillo, A. R. 1987. Taphonomy of Hazard Homestead Quarry (Ogallala Group), Hitchcock County, Nebraska. M.A. thesis, University of Nebraska, Lincoln. ———. 1988. Taphonomy of Hazard Homestead Quarry (Ogallala Group), Hitchcock County, Nebraska. University of Wyoming Contributions to Geology 26: 57–97. Gates, T. A. 2004. Hadrosaurian dinosaur diversity from the Upper Campanian Kaiparowits Formation, southern Utah. Journal of Vertebrate Paleontology 24(3, Suppl.): 63A. Gates, T. A., E. K. Lund, S. D. Sampson, J. I. Kirkland, M. A. Getty, D. DeBlieux, and A. Titus. In press. Ornithopod Dinosaurs from the Late Cretaceous Kaiparowits Plateau, Grand Staircase–Escalante National Monument. Learning from the Land, Vol. 2. Bureau of Land Management. Gates, T. A., and S. D. Sampson. 2007. A new species of Gryposaurus (Dinosauria: Hadrosauridae) from the Upper Campanian Kaiparowits Formation of Utah. Zoological Journal of the Linnean Society 151: 351–376. Getty, M. A., E. M. Roberts, M. A. Loewen, J. A. Smith, T. A. Gates, and S. D. Sampson. 2003. Taphonomy of a chasmosaurine ceratopsian skeleton from the Campanian Kaiparowits Formation, Grand Staircase–Escalante National Monument, Utah. Journal of Vertebrate Paleontology 23(3, Suppl.): 54A–55A. Goldstrand, P. M. 1992. Evolution of the Late Cretaceous and Early Tertiary basins of southwest Utah based on clastic petrology. Journal of Sedimentary Petrology 62: 495–507. Gregory, H. E., and R.C. Moore. 1931. The Kaiparowits Region. U.S. Geological Survey Professional Paper 164. Jinnah, Z., A. Deino, T. A. Gates, and E. M. Roberts. 2007. The first AR/AR Age dates from the Wahweap Formation (Late Cretaceous of Utah): Implications for faunal correlations. Journal of Vertebrate Paleontology 27(3, Suppl.): 96A. Kirkland, J. I., D. Deblieux, J. A. Smith, and S. D. Sampson. 2002. New ceratopsid remains from the lower Campanian Wahweap Formation, Grand Staircase–Escalante National Monument, Utah. Journal of Vertebrate Paleontology 22(3, Suppl.): 74A. Lehman, T. M. 1982. A ceratopsian bonebed from the Aguja Formation (Upper Cretaceous), Big Bend National Park, Texas. M.A. thesis, University of Texas, Austin. ———. 1990. The ceratopsian subfamily Chasmosaurinae: Sexual dimorphism and systematics. In P. J. Currie and K. Carpenter, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 211– 229. Cambridge: Cambridge University Press. Little, W. W. 1995. The influence of tectonics and eustacy on alluvial architecture, Middle Coniacian through Campanian strata of the Kaiparowits Basin, Utah. Ph.D. diss., University of Colorado, Boulder. Lohrengel, C. F. 1969. Palynology of the Kaiparowits Formation, Garfield County, Utah. Brigham Young University Geology Studies 6: 61–180. Parrish, J. M. 1999. Dinosaur teeth from the Upper Cretaceous (Turonian-Judithian) of southern Utah. In D. Gillette, ed., Ver-

tebrate Paleontology of Utah, pp. 319–321. Utah Geological Survey Miscellaneous Publication 99-1. Parrish, J. M., and J. G. Eaton. 1991. Diversity and evolution of dinosaurs in the Cretaceous of the Kaiparowits Plateau, Utah. Journal of Paleontology 11(3, Suppl.): 50A. Roberts, E. M. 2005. Stratigraphic, taphonomic and paleoenvironmental analysis of the Upper Cretaceous Kaiparowits Formation, Grand Staircase–Escalante National Monument, Southern Utah. Ph.D. diss., University of Utah, Salt Lake City. ———. 2007. Facies architecture and depositional environments of the Upper Cretaceous Kaiparowits Formation, southern Utah. Sedimentary Geology 197: 207–233. Roberts, E. M., A. L. Deino, and M. A. Chan. 2005. 40Ar/ 39Ar age of the Kaiparowits Formation, southern Utah and correlation of contemporaneous Champanian strata and vertebrate faunas along the margin of the Western Interior Basin. Cretaceous Research 26: 307–318. Roberts, E. M., R. R. Rogers, and B. Z. Forman. 2003. An experimental approach to identifying and interpreting dermestid (Insecta: Coleoptera) bone modification. Journal of Vertebrate Paleontology 23(3, Suppl.): 89A. ———. 2007. Continental insect borings in dinosaur bone: Examples from the Late Cretaceous of Madagascar and Utah. Journal of Paleontology 81(1): 201–208. Roberts, E. M., and L. Tapanila. 2006. A new social insect nest trace from the Late Cretaceous Kaiparowits Formation of southern Utah. Journal of Paleontology 80: 768–774. Ryan, M. J. 1992. The taphonomy of a Centrosaurus bone bed (Campanian), Dinosaur Provincial Park, Alberta, Canada. M.Sc. thesis, University of Calgary, Calgary. ———. 2007. A new basal centrosaurine ceratopsid from the Oldman Formation, southeastern Alberta. Journal of Paleontology 81: 376–396. Sampson, S. D., and M. A. Loewen. 2010. Unraveling a radiation: A review of the diversity, stratigraphic distribution, biogeography, and evolution of horned dinosaurs (Ornithischia: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 405–427. Bloomington: Indiana University Press. Sampson, S. D., M. A. Loewen, T. A. Gates, M. A. Getty, and L. E. Zanno. 2002. New evidence of dinosaurs and other vertebrates from the Upper Cretaceous Wahweap and Kaiparowits Formations, Grand Staircase–Escalante National Monument. Geological Society of America Special Publication, Rocky Mountain Sectional Meeting 35: 5. Sampson, S. D., M. A. Loewen, E. M. Roberts, J. A. Smith, L. E. Zanno, and T. A. Gates. 2004. Provincialism in Late Cretaceous terrestrial faunas: new evidence from the Campanian Kaiparowits Formation of Utah. Journal of Vertebrate Paleontology 24(3, Suppl.): 108A. Titus, A. L., J. D. Powell, E. M. Roberts, S. D. Sampson, S.L. Pollock, J. I. Kirkland, and L. B. Albright. 2005. Late Cretaceous stratigraphy, depositional environments, and macrovertebrate paleontology of the Kaiparowits Plateau, Grand Staircase– Escalante National Monument, Utah. In J. Pederson, and C. M.

Taphonomy of Horned Dinosaurs from the Late Campanian Kaiparowits Formation 493

Dehler, eds., Interior Western United States, pp. 101–128. Geological Society of America Field Guide 6. Titus, A. L., S. D. Sampson, D. D. Gillette, and J. I. Kirkland. 2001. Specialist-driven long-term interdisciplinary efforts in Grand Staircase–Escalante National Monument: A model for resource inventory. Grand Junction: 6th Conference on Fossil Resources. Voorhies, M. R. 1969. Taphonomy and Population Dynamics of an Early Pliocene Vertebrate Fauna, Knox County, Nebraska. University of Wyoming, Contributions to Geology, Special Paper 1. Wiersma, J., H. Hutchison, and T. A. Gates. 2004. Crocodilian diversity in the Upper Cretaceous Kaiparowits Formation (Upper Campanian), Utah. Journal of Vertebrate Paleontology 24(3, Suppl.): 129A.

494 getty, loewen, roberts, titus, & sampson

Wood, J. M., R. G. Thomas, and J. Visser. 1988. Fluvial processes and vertebrate taphonomy: The Upper Cretaceous Judith River Formation, south-central Dinosaur Provincial Park, Alberta, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology 66: 127–143. Zanno, L., and S. D. Sampson. 2005. A new oviraptosaur (Theropoda, Maniraptora) from the Late Cretaceous (Campanian) of Utah. Journal of Vertebrate Paleontology 25: 897–904. Zanno, L., J. P. Wiersma, M. A. Loewen, M. A. Getty, and S. D. Sampson. In press. A preliminary report on the theropod dinosaur fauna of the late Campanian Kaiparowits Formation, Grand Staircase–Escalante National Monument. Learning from the Land, Vol. 2. Bureau of Land Management.

32 A Centrosaurine Mega-Bonebed from the Upper Cretaceous of Southern Alberta: Implications for Behavior and Death Events D AV I D A . E B E R T H , D O N A L D B . B R I N K M A N , A N D VA I A B A R K A S

the hilda area of southern Alberta preserves the re-

Dinosaur Provincial Park, suggest that the mega-bonebed

mains of at least 14 discrete monodominant Centrosaurus

assemblage was eventually buried during a subsequent

apertus bonebeds that occur in a single, organic-rich

coastal-plain flooding event.

mudstone bed deposited in a coastal-plain interfluve.

Data from the Hilda mega-bonebed are compatible

The identical and traceable stratigraphic positions of

with a previously proposed hypothesis that some cen-

the 14 bonebeds allow us to group and interpret them as

trosaurines exhibited east-west seasonal migratory be-

a single mega-bonebed, covering an estimated area of

haviors. The existence of the Hilda mega-bonebed

2.3 km2 and containing a cumulative minimum number

suggests that many of the stratigraphically close cen-

of individuals (MNI) estimated to be in the very low

trosaurine bonebeds at Dinosaur Park may also be parts

thousands. The Hilda mega-bonebed is a rare example of

of two or more mega-bonebeds in that area. Testing of

a multi-kilometer-scale, macrovertebrate bonebed that is

this hypothesis may be possible using geochemical

inferred to have formed, in large part, due to intrinsic

means.

biogenic means: gregarious behavior in a group of Centrosaurus apertus. We infer that formation of the Hilda mega-bonebed

Introduction

was initiated as more than one thousand gregarious cera-

Alberta’s centrosaurine bonebeds, studied since the 1980s,

topsians, and other large terrestrial vertebrates, drowned

have provided abundant evidence for gregarious behavior,

during a major coastal-plain flooding event. As flood-

mass mortalities and patterns of taphonomic modification

waters receded, individuals and groups of drowned Cen-

within ceratopsians (Currie and Dodson 1984; Visser 1986;

trosaurus apertus became pooled across the floodplain.

Ryan et al. 2001; Eberth and Getty 2005). Although numerous

Exposed skeletons experienced large degrees of dis-

areas within the province are known to produce monotaxic-

articulation and component elements were subject to

to-monodominant centrosaurine bonebeds (e.g., Dinosaur

breakage, varying amounts of abrasion, and rare tooth

Provincial Park, Grande Prairie, South Saskatchewan River,

scarring as a result of month-to-multiyear exposure

Drumheller, Scabby Butte), the majority of studies on these

times, high rates of soft tissue rotting, and reworking due

kinds of bonebeds has taken place at Dinosaur Provincial

to scavenging and trampling. Stratigraphic and sedimen-

Park (DPP; Currie and Dodson 1984; Visser 1986; Ryan et

tologic data, and comparisons with similar bonebeds at

al. 2001; Eberth and Getty 2005). There, at least 20 docu-

495

mented centrosaurine bonebeds occur in a 44 m thick stratigraphic zone in the upper one-half of the Belly River Group (Eberth and Getty 2005: 501). Among these, 17 bonebeds contain Centrosaurus apertus and occur within the lowest 14 m of the Dinosaur Park Formation (DPFm; Eberth and Getty 2005). The centrosaur-bonebed rich lower 14 m of the DPFm at DPP is exposed over an area of approximately 50 km2. The occurrence of 17 ceratopsian bonebeds in this thin stratigraphic slice, and within such a limited area, has suggested to some bonebed researchers the possibility that some of these bonebeds in the Park formed during the same event(s) (e.g., Eberth and Getty 2005). Testing this hypothesis, however, has been difficult. The three-dimensional geomorphology of the badlands at DPP significantly limits one’s ability to trace individual bonebeds and their host beds over more than a few hundreds of meters. More importantly, bonebeds are frequently truncated by surfaces that mark paleochannel incision and other Cretaceous erosion/nondeposition events (Koster and Currie 1987; Wood 1989; Eberth and Hamblin 1993). Thus, attempts at physically correlating these previously known centrosaurine bonebeds have been largely unsuccessful. To date, it has been confirmed only that the Centrosaurus apertus bonebeds at DPP formed during at least two separate events (i.e., BB91a can be shown to stratigraphically overlie BB91; Eberth and Getty 2005). Because of the limitations on assessing the stratigraphic relationships among the centrosaurine bonebeds at DPP, a 1997 field study was conducted by the Royal Tyrrell Museum at ‘‘Hilda,’’ 80 km to the southeast along the South Saskatchewan River, 25 km west of the village of Hilda. Field notes from Wann Langston, Jr. (1959) and data from a small bonebed collection amassed under the direction of Don Taylor (Provincial Museum of Alberta, now the Royal Alberta Museum, 1964–1966) suggested that at least two Centrosaurus apertus bonebeds were present in the area. Geomorphologically, the area is a steep-walled glacial valley through which the South Saskatchewan River flows northward, and where badlands have a limited extent (Fig. 32.1). Because of the absence of extensive badlands, laterally extensive beds can be traced in a north-south direction along the valley walls without significant interruption. Furthermore, at any given time during the late Campanian, Hilda was paleogeographically downdip from DPP, and thus closer to the Western Interior Seaway (Fig. 32.2). This situation may have allowed for greater accommodation of sediments, and theoretically, a reduced frequency of truncated beds, especially in the middle and upper portions of the section. During a brief reconnaissance of the area in 1996, we relocated one of the bonebeds described by Langston and Taylor, but also noted the presence of other ceratopsian specimens and centrosaurine-dominated bonebeds in the area. A

496 eberth, brinkman, & barkas

FIGURE 32.1. Location of 14 Centrosaurus apertus bonebeds (see Table 32.1) along the South Saskatchewan River, southeastern Alberta. Sites whose stratigraphic sections and correlations are shown in Figs. 32.3 (H97-02 and 04) and 32.4 (H97-02, 04, 06, 08, 11) are in bold type. Shaded area indicates extent of the stacked mudstone zone that includes the bonebed host bed (see Figs. 32.3 and 32.4). Section, township, and range locations west of the fourth meridian (W4) indicated in small print. Elevation lines are 650 and 700 m above sea level. Locations of photomosaic-constrained cross sections (pm1 and pm2; Fig. 32.5) are indicated by dashed and double-headed arrows. In this area, the South Saskatchewan River flows to the north. Inset indicates location of field area and Dinosaur Provincial Park (DPP) in Alberta (AB).

FIGURE 32.2. Schematic west-east cross section of the Belly River Group clastic wedge. The three-fold formational subdivision (Foremost, Oldman, and Dinosaur Park formations) is after Eberth and Hamblin (1993) and Eberth (2005). The Hilda mega-bonebed (asterisk) occurs near the top of the wedge, in the lower one-third of the Dinosaur Park Formation. Based on distance above the Oldman–Dinosaur Park contact, the Hilda Centrosaurus mega-bonebed is stratigraphically higher (10 m) than the Centrosaurusbonebed-rich horizon at Dinosaur Provincial Park (square). Inset shows the approximate location of the cross section through Alberta (AB) and Saskatchewan (SK). Age of the Belly River Group clastic wedge ranges from 79.1 Ma to 74.5 Ma. Age of the Hilda mega-bonebed and coeval sites at Dinosaur Provincial Park is estimated at 76 Ma. Stratigraphic thicknesses at Hilda and Dinosaur Provincial Park are 60 m and 70 m, respectively.

focussed effort in 1997 resulted in the documentation of a

Geologic Setting and Depositional Context

total of 14 discrete centrosaurine bonebeds occurring in the same laterally continuous mudstone bed (7 km in observed

The Hilda ceratopsian bonebeds all occur in the lower one-half

extent) and cropping out, as a group, across a total geographic

of the DPFm, 20–25 m above the contact with the Oldman

distance of 3.7 km in the north-south oriented valley walls of

Formation (Figs. 32.3–32.5). In the Hilda area, we estimate the

the South Saskatchewan River (Fig. 32.1; Table 32.1). The work

complete thickness of the DPFm (defined by its contacts with

also resulted in the compilation of five measured stratigraphic

the underlying Oldman and overlying Bearpaw formations;

sections (Figs. 32.3, 32.4) and two photomosaic-controlled

Eberth 2005) as close to 60 m, based on our comparisons with

cross sections (Fig. 32.5), and the systematic excavation of

exposed sections farther north at Sandy Point, and farther

one of the ceratopsian bonebeds (H97-04; Figs. 32.6, 32.7).

west at Iddesleigh (adjacent to Dinosaur Provincial Park).

Here we present those data and propose that the Hilda centro-

However, the preserved thickness of the DPFm in our field area

saurine bonebeds represent the remains of a single mega-

does not exceed 43 m, due to glacial erosion and slumping

bonebed that covered an area of at least 2.3 km2. The data

(Figs. 32.3, 32.4).

further suggest that the mega-bonebed formed as the result

The Dinosaur Park and Oldman formations (sensu Eberth

of a mass death event involving low thousands of gregari-

and Hamblin 1993) comprise the upper one-half of the Belly

ous centrosaurines, most likely Centrosaurus apertus, during

River Group clastic wedge in southern Alberta (sensu Jerzykie-

a coastal-plain flood. Because coastal-plain flood-induced

wicz and Norris 1994). The wedge accumulated east of a broad

mass deaths of large dinosaurs have been well documented at

zone of uplift and overthrusting in central and eastern British

DPP (Currie and Dodson 1984; Ryan et al. 2001; Eberth and

Columbia that, in turn, reflected regional-scale and oblique

Getty 2005; Eberth and Currie 2005), we propose that it is very

accretionary tectonics along the western margin of Canada

likely that the 17 known Centrosaurus apertus bonebeds at

(Cant and Stockmal 1989). The Belly River clastic wedge thins

the Park probably comprised portions of similar types of

to the east and consists of nonmarine-to-paralic facies that

mega-bonebeds whose true paleogeographic limits are not yet

were deposited along the western margin of the Western Inte-

documented.

rior Basin during mid-to-late Campanian time (Fig. 32.2). The

A Centrosaurine Mega-Bonebed from the Upper Cretaceous of Southern Alberta 497

FIGURE 32.3. Measured sections through two Centrosaurus bonebeds that, in part, comprise the Hilda mega-bonebed. H97-04 was excavated by our team in 1997. H97-02 was excavated by teams from the Provincial Museum of Alberta (now the Royal Alberta Museum) in the 1960s. Bedrock exposures in this area are missing the uppermost 15–25 m of the Dinosaur Park Formation.

498 eberth, brinkman, & barkas

FIGURE 32.4. Delineation of the Hilda Centrosaurus mega-bonebed (dashed lines) using measured sections through five of the individual bonebeds (BB) that, in part, comprise the mega-bonebed. The location of each measured section and bonebed is shown in Fig. 32.1. Shaded area indicates the stacked mudstone zone that includes the mega-bonebed host bed. Note its variable thickness and variable relationship to the mega-bonebed host bed. Note the distances between sites.

wedge is bounded by and interfingers with marine shales of

contains a relatively high percentage of stacked, sandy paleo-

the underlying Pakowki Formation and overlying Bearpaw

channel fills relative to interfluve mudstones. This interval

Formation.

largely reflects deposition in straight to meandering paleo-

The DPFm reflects the overall transgressive phase of the

channels that were typically up to 5 m deep, up to 200 m wide,

Belly River Group. Upward across the Oldman-Dinosaur Park

and flowed toward the east-southeast. The abundance of het-

formational contact, changes in the Belly River sediment

erolithic paleochannel deposits and variety of intraclast de-

package record a north-to-south shift in active tectonism and,

posits throughout the DPFm demonstrate that paleochannel

thus, increased subsidence in the foredeep of the southern

flow varied from standing water to high-energy and erosive.

Canadian Cordillera (Eberth and Hamblin 1993). The south-

Carbonaceous to sandy rhythmites characterize many of the

ward shift in tectonism and increase in subsidence apparently

formation’s inclined heterolithic strata (IHS), and were likely

resulted in a southeastward tilt in the basin (Eberth and

a product of daily-to-seasonal rainfall and floodplain sweep

Hamblin 1993). In response, areas of the foredeep in the cen-

rather than daily variations in tidal energy (Eberth 2005).

tral and northern Canadian Cordillera, which previously had

Laterally extensive and smectite-clay-rich, grey-green to or-

trapped clastics, began to rebound, shedding coarse clastics

ganic rich facies in the lower DPFm represent deposition in a

southeastward into the basin, and depositing the Dinosaur

floodbasin setting that was occasionally the site of hydromor-

Park Formation (Eberth and Hamblin 1993). At the same time,

phic soil formation and iron-carbonate precipitation (Eberth

the tilt may have allowed for deeper westward penetration of

2005). A vast amount of palynological data has been derived

the Western Interior Seaway into southern Alberta.

from these mudrocks (Braman and Koppelhus 2005) indicat-

The Dinosaur Park Formation consists of alluvial, estuarine

ing that southern Alberta’s climate during the mid-to-late

and paralic facies that were deposited across a broad, but very-

Campanian was warm temperate with abundant, but sea-

low-gradient, alluvial-to-coastal plain similar to modern day

sonal, rainfall. Similarly, Béland and Russell (1978) and Koster

Bangladesh (Eberth 1998; Eberth and Getty 2005). The lower

et al. (1987) synthesized geological and paleobotanical data

one-half of the formation is generally coarser grained, and

and reviewed published evidence for moderate to high mean

A Centrosaurine Mega-Bonebed from the Upper Cretaceous of Southern Alberta 499

FIGURE 32.5. Two photomosaic-based cross sections (pm1 and pm2) extending from Hilda Centrosaurus bonebed H97-02 south to Hilda Centrosaurus bonebed H97-08. Location of photomosaics is indicated in Fig. 32.1. Total distance is approximately 3.3 km. Note the overlap of cross sections at H97-06. Curved aspect of both cross sections is a result of compounded photographic distortion. Note the consistent expression of the basal sandstone horizon, stacked mudstone horizon, and uppermost coal.

annual paleotemperatures, and high annual rainfall (]1,200

as part of a stacked mudstone succession; Fig. 32.1), the host

mm) that was seasonally variable.

and the stacked mudstone succession are erosionally truncated by a deep paleochannel incision a few hundred meters

Mega-Bonebed Host Bed

south of bonebed H97-13 (Fig. 32.1). The stacked mudstone succession typically consists of 3–15

Fig. 32.1 shows the geographic distribution of the 14 centro-

mudstone beds with a combined thickness that ranges from 3

saur bonebeds in the Hilda area. Each bonebed is hosted by the

to 10 m. In our five measured sections (Figs. 32.3, 32.4), the

same laterally continuous bed of brown-grey, massive, car-

bonebed host mudstone bed occurs at the base of the mud-

bonaceous siltstone/claystone (Figs. 32.3, 32.4). We examined

stone succession in two sections (H97-08, 11), and in the mid-

host bed stratigraphy and sedimentology at each bonebed,

dle of the succession in the remaining three sections (H97-02,

and host bed lateral continuity between the bonebeds.

04, 06).

The host bed is part of a prominent and laterally traceable,

The host bed ranges in thickness from 25 cm to 1 m, and typ-

multimeter-thick mudstone succession in the lower one-third

ically fines up in its uppermost few centimeters. It often shows

of the DPFm at Hilda (Figs. 32.1, 32.3–32.5). The multimeter

evidence for soft sediment deformation and contortion, espe-

thick mudstone succession can be traced continuously along

cially close to its lower and upper contacts. Throughout its

the South Saskatchewan River valley walls for more than 7 km

lateral extent, the host bed contains abundant fragments of

(the bonebeds are distributed across only one-half of this ex-

coal, carbonaceous plant, and wood debris, and vertically to

tent; Fig. 32.1). Although the northern limit of the mega-

horizontally oriented fine root traces. Lumpy, red-brown con-

bonebed host bed is unknown (it may extend north of H97-02

cretionary ironstone masses occur locally, especially where

500 eberth, brinkman, & barkas

Table 32.1. Names and Descriptions of 14 Centrosaurine (Centrosaurus apertus) Bonebeds and Associated Fossil Sites from the South Saskatchewan River, Southern Alberta

Maximum Field # H97-02

H97-03

Name Langston’s

Lithology normally graded bones in a

Bone

Estimated size of

concentration

thanatocoenose

Thickness

extent of

Estimated

(per linear m

(order of

of BB

BB

BB area

of exposure)

magnitude)

25 cm

50 m

1963m2

5/m

low 100s

1–2/m

10s

10/m for

100s to 1000

Ceratopsian BB

carbonaceous sandy siltstone

Centrosaur skull

carbonaceous sandy siltstone

25 cm

10 m

79m2

normally graded bones in a tan

25 cm

150 m e-w

17671m2

& BB H97-04

Centrosaurus BB

colored, carbonaceous silty

50m; 1/m for

claystone H97-05

100m

nasal horncore

normally graded bones in a

BB

brown, carbonaceous, silty

25 cm

40 m (n-s)

1257m2

1–2/m

10s–low 100s

25 cm

50 m (n-s)

1963m2

1–2/m

10s–low 100s

50 cm

40 m (n-s)

1257m2

1–2/m

10s–low 100s

10 m

79m2

1–2/m

[10

706m2

5/m

high 10s–very

claystone H97-06

Jason Lavigne’s

normally graded bones in a

BB

brown, carbonaceous, sandy siltstone

H97-07

Microsite BB

normally graded bones in a massive carbonaceous sandy and clayey siltstone

H97-08

Southeasternmost BB

H97-10

Westside 1 BB

red-brown, silty carbonaceous

30 cm

shale

(e-w)

very fine grained sandstone with

25–30 cm

30 m (n-s)

intraformational pebbles; paleo-

low 100s

channel lag deposit H97-11

Westside 2 BB

grey-green carbonaceous clayey

30 cm

siltstone H97-13

Southwestern-

Low-density

?

?

Southeast of

5/m

high 10s

[30 m

706m2

1–2/m

10s

1963m2

[1/m

10s

(n-s) sandy, silty carbonaceous shale

60 cm

ceratopsian BB 1 H97-17

314m2

(e-w)

most BB H97-16

[20 m

50 m (e-w)

carbonaceous silty claystone

?

30 m

706m2

1–2/m

10s

interbedded sandstone and

1m+

[10 m

79m2

[1/m

[10

?

30 m

706m2

[1/m

10s

Langston BB H97-19

ceratopsian tail BB

siltstone (splay?) incised into bonebed mudstone

H97-20

Low-density

carbonaceous sandy siltstone

ceratopsian BB 2 H97-09

Aquatic micro-

carbonaceous clayey siltstone

site H97-12

Mesovertebrate

fine sandstone; paleochannel fill

site

with crocodile and turtle pieces

A Centrosaurine Mega-Bonebed from the Upper Cretaceous of Southern Alberta 501

bone concentrations are high. At each bonebed, bones occur

rare and were observed only at H97-03 (one partial Centro-

most frequently at or near the base of the host bed (e.g.,

saurus skull) and H97-19 (one partial ceratopsian tail).

Fig. 32.3).

Large bones are typically concentrated along a single bed-

The host bed varies in its contacts with other beds between

ding plane at the base of each bonebed, whereas fragments,

sites and across the field area. At some localities, the contacts

when present, exhibit normal grading upward through the

between the host bed and the over- and underlying mud-

bonebed, becoming matrix-supported in its upper portions

stones are sharp. At other localities, paleochannel-fill sand-

(Figs. 32.6, 32.7). At all sites, most bones appear to be oriented

stones erode into, or interfinger with the host bed (Fig. 32.5).

with their long axes parallel to the horizontal plane. However, a few small fragments were observed having subvertical in situ

Bonebed Taxonomic Composition, Bone Concentration, and Preliminary Taphonomy

orientations during the excavation at H97-04. Although this study does not include a quantified or detailed taphonomic analysis of collected elements, our examinations of bone surface modifications at H97-04 as elements

Although the vast majority of the ceratopsian material that we

were collected and among the H97-02 specimens in the Royal

encountered is taxonomically unidentifiable below the level

Tyrrell Museum collections indicate that varying degrees of

of Centrosaurinae, all well preserved adult parietals, nasal and

abrasion are common, and that scratch and tooth marks are

supraorbital horncores have a form characteristic of Centro-

present, but very rare.

saurus apertus (Ryan 2003), notably possessing a pair of well

Two vertebrate microfossil sites (H97-09 and H97-12) were

developed, anteroventrally oriented parietal processes (P1), a

also noted within the host bed stratigraphic horizon, al-

pair of medially recurved parietal processes (P2), an erect nasal

though neither qualifies as a ceratopsian bonebed (Table

horncore, and paired, pyramid-shaped orbital horncores. The

32.1). H97-09 occurs in the host facies (carbonaceous clayey

parietal features are well developed in Royal Tyrrell Museum

siltstone) and contains the remains of aquatic vertebrates

specimen TMP 1964.5.190, and the nasal and orbital horn-

(fish, turtles, crocodiles, and champsosaurs). H97-12 occurs

core features are exhibited by TMP 1965.23.25 and TMP

within a sandstone lens that cuts down into the host bed,

1965.23.10, respectively. Accordingly, we have assumed that

postdating the latter. It contains bone fragments from croco-

all the centrosaurine material from all of the bonebeds repre-

diles and turtles.

sents a monodominant assemblage (sensu Eberth et al. 2005)

Estimates of bone abundance within bonebeds can aid in

of Centrosaurus apertus. Whereas we remain entirely open

estimating the size of the original death assemblage (thanato-

to the possibility that future taxonomic assessments of some

coenose). Given that the ceratopsian bonebeds in southern

of this material may alter its generic or specific status—diag-

Alberta figure importantly in all discussions about ceratopsian

nostic material is actually quite rare—we regard that possibil-

herding and other gregarious behaviors (Brinkman et al. 1998;

ity as having no significant impact on the primary conclu-

Eberth and Getty 2005), we attempted to assess bone abun-

sions presented here regarding the presence and origin of the

dances within the Hilda bonebeds by quantifying the bone

mega-bonebed.

concentrations and projecting those results over the esti-

We examined exposed-bone quality, element/specimen

mated size of each bonebed (cf. Eberth and Getty 2005). Maxi-

abundances, and the extent of each bonebed at the 14 bone-

mum bone concentrations were estimated at each site by re-

bed sites documented here (Table 32.1). Although we exca-

cording the largest number of identifiable elements per linear

vated and collected in situ material and preliminary tapho-

meter of bonebed exposure (Table 32.1). This method ignores

nomic data from only one of these sites (H97-04), additional

bone fragments and depends instead on positive identifica-

taphonomic data were also retrieved from the ceratopsian ele-

tions of elements.

ments collected from H97-02 by the Provincial Museum of Al-

Because of the relatively small exposed lateral extents (tens

berta in 1964–1966 (now the Royal Alberta Museum), which

of meters) of almost all of the Hilda bonebeds, a single ‘‘con-

are now part of the Royal Tyrrell Museum collections.

centration value’’ was assessed for each bonebed (Table 32.1).

Highly localized exposures of concentrated ceratopsian

Whereas most sites have relatively low bone concentrations

bone in the host mudstone define the individual bonebeds.

of 1–2 bones/linear meter, H97-10 and 11 exhibit signifi-

The largest exposure appears to be H97-04, where variably

cantly higher bone concentrations (5 bones/linear meter). At

concentrated bone can be traced laterally for approximately

H97-04, however, we noted that although there was a central

150 m. All other sites appear to be considerably smaller than

zone of highly concentrated bone (10 bones/linear meter in

H97-04, having notable bone concentrations that can be

the middle 50 m of the exposure), bone concentration was

traced for 50 m or less.

significantly lower (1 bone/linear meter) for approximately

Articulated or associated skeletal elements are extremely

502 eberth, brinkman, & barkas

50 m on either side of the central zone (Table 32.1).

FIGURE 32.6.

Graphic representation of all specimens encountered in a 1 m2 (Grid A4) in Excavation A, H97-04 (location shown in Fig. 32.1). Excavation A covered a total area of 4 m2. The positions of the specimens relative to four stratigraphic levels (total thickness of 50 cm) are indicated by differences in shading. Darker shades are lower, lighter shades are higher. Note that smaller specimens were collected from the uppermost 20 cm of the excavation (white specimens). Dashed line indicates the erosional edge of the bonebed.

Our excavations at H97-04 (Figs. 32.6, 32.7) revealed that bone concentrations and breakage are highly variable within

individuals of different size, as well as a fish scale and two theropod teeth.

this site. For example, the quarry maps and specimen mani-

The unmapped ceratopsian elements from H97-02 that

fests from two separate excavations at H97-04 (excavations A

were collected by the Provincial Museum of Alberta (now the

and B) indicate that Excavation A, a total of 4 m,2 yielded 602

Royal Alberta Museum) comprise a total of 118 specimens,

specimens, most of which (79%) are broken and fragmentary

most of which are fragmentary partial elements of centro-

to such a degree as to be unidentifiable (Fig. 32.6; Table 32.2).

saurines. The fragmentary nature of this material suggests

In contrast, Excavation B (a total of 8 m;2 Fig. 32.7) yielded

that, like H97-04, specimens at this site are largely broken and

65 specimens, 48 of which are identifiable to skeletal part. The

fragmentary. We have documented a minimum of six cen-

intensity of the breakage at Excavation A was notably high,

trosaurs within this collection, based on nasal horncores. In

as indicated by the presence of 475 small and fragmentary

line with our observations at H97-04, this material also shows

specimens that were simply classified as bone fragments,

bone modification features that include variable degrees of

trabecular-bone pebbles and bone chips (Table 32.2). Identi-

abrasion, and very rare instances of tooth marking.

fiable specimens at Excavation A included cranial and postcranial elements from at least two centrosaurs, shed teeth from a large and small theropod and a crocodile, champsosaur

Interpretations

vertebrae, a fish scale, a pisidiid clam, and two small coprolites

DEPOSITIONAL HISTORY

of uncertain origin. A few meters to the southeast, Excavation B yielded significantly more identifiable centrosaur elements

Geological and paleontological data presented here and else-

of large size, including the skeletal remains of at least two

where (e.g., Eberth and Getty 2005) indicate that the cera-

A Centrosaurine Mega-Bonebed from the Upper Cretaceous of Southern Alberta 503

topsian bonebeds in the Dinosaur Park Formation of southern Alberta (at both DPP and Hilda) were deposited across a vast and wet alluvial-coastal lowland that was subjected to seasonal-storm-induced flooding events. The organic-rich host facies containing the ceratopsian bonebeds in the DPFm at DPP has been previously interpreted as resulting from the deposition of suspended sediments and organic debris in an interfluve coastal-plain setting, and modified postdepositionally by trampling, differential compaction, rooting, and diagenesis (Koster 1984; Wood 1989; Eberth 2005). At Hilda, the great lateral continuity along depositional strike of a single mudstone bed that contains 14 medium-tosmall-size Centrosaurus apertus bonebeds indicates that, on the basis of parsimony, these bonebeds should be regarded as a single, patchy mega-bonebed that formed in an extensive interfluve wetland, kilometers in extent. The minimum size of the mega-bonebed is calculated at 2.3 km2, based on the maximum north-south and east-west distributions of the 14 bonebeds (Fig. 32.1). Because these bonebeds occur on both sides of the South Saskatchewan River valley, the areal estimate proposed here includes a large volume of bedrock that was removed during valley incision at the end of the Pleistocene. The stratigraphic and sedimentologic data presented here, and comparisons with similar strata at DPP indicate that the mega-bonebed host bed at Hilda can be interpreted as having been deposited during a single flooding event, and that it was subsequently modified by scavenging, trampling, rooting, and diagenesis. The variable lithology and stratigraphic relationships of the host bed indicate that there were variations in the energy and extent of the flooding event, substrate relief, and biological productivity across the flooded area before, during, and after bonebed formation. These variations likely reflect the lateral coexistence of a variety of interfluve subenvironments (e.g., splays, ponds, paleosols, etc.), each of which probably extended for less than a kilometer along depositional strike. The local presence of meter-wide and decimeter thick, lenticular, coarse-grained fills, in and incising the host facies indicates that the interfluve also contained small drainages, similar to modern wetland and tidal creeks. Graphic representation of all specimens encountered in the total 8 m2 area of Excavation B, H97-04 (location of H97-04 shown in Fig. 32.1). All specimens were collected from the same stratigraphic horizon. Note the absence of small bone fragments that predominate in Excavation A (Fig. 32.6). Excavation B was situated approximately 3 m southeast of Excavation A. FIGURE 32.7.

The abundance of plant debris and rooted horizons in the host bed and in the over- and underlying beds clearly indicates that the area was organically productive and likely consisted of plants and plant communities that were identical or very similar to those documented in the lower DPFm at DPP (e.g., Braman and Koppelhus 2005). Locally developed, ‘‘lumpy’’ ironstone-rich mudrock suggests syndepositional precipitation of reduced iron, probably as a function of organic decay (bacterially mediated) in the presence of oxygenated, alkaline water (Allison and Briggs 1991; Tsujita 1995). Eberth and Getty (2005) and Eberth and Currie (2005) hypothesized that the centrosaurine bonebeds at DPP originated

504 eberth, brinkman, & barkas

Table 32.2. Paleontology Collection from Hilda Bonebed H97A (1997)

relatively higher base level, also indicates that coastal-plain flooding events were very common, and may have dominated the depositional history of the area. As in modern low-

Number of Specimens/elements

elements/specimens

All elements Unidentifiable fragments, chips, and

megafauna, and/or flood-related disease events. In either case, the signature of interfluve-hosted macrofossil remains is the

127

same: bonebeds in the case of gregarious animals and single occurrences of complete-to-fragmentary skeletons in the case

Centrosaurus apertus # of identifiable elements

(Eberth and Getty 2005), severe floods in the Hilda region would have resulted in either the drowning of the terrestrial

475

pebbles Identifiable elements

gradient depositional coastal environments like Bangladesh

117

MNI of C. apertus (based on 3 fibulae, 3

2

of isolated or solitary animals. Drought events have also been invoked to explain the occurrence of some Campanian-age ceratopsian bonebeds (Rog-

ischia)

ers 1990). The presence of subadult and juvenile bones at Large theropod (tooth)

1

H97-04 and H97-02 (a characteristic of some drought assem-

Troodontid theropod (tooth)

1

blages [cf. Gates 2005; Shipman 1975]) raises the question of

Crocodile (tooth)

1

whether drought may have been responsible for bonebed for-

Champsosaur (vertebrae)

3

Fish (scale)

1

Pisidiid clam

1

Coprolite

2

mation at Hilda. Although we cannot exclude the possibility that a short term, severe drought resulted in the death of the Hilda centrosaurs, such an interpretation seems highly unlikely in the absence of any direct evidence for climatic aridity, and the presence of geologic and paleontologic data (pre-

as the result of coastal-plain flooding events that frequently

sented above) that indicate that (1) the climate was wet sub-

inundated southeastern Alberta during the late Campanian.

tropical to warm temperate, (2) paleoenvironments were

They proposed that these floods repeatedly mass-drowned the

characterized by perennially flowing rivers and hydromor-

standing terrestrial macrofauna, including gregarious assem-

phic soils, and (3) the area was home to rich aquatic plant and

blages of Centrosaurus and Styracosaurus, when present. As

animal communities.

floodwaters receded, individual dinosaur carcasses and groups of drowned centrosaurines would pool across the floodplain. In some cases, carcasses and disarticulated remains were fur-

DEATH ASSEMBLAGE SIZE ESTIMATES

ther concentrated in reestablished river channels. After tapho-

Estimating the original size of a bonebed death assemblage is

nomic modification of exposed carcasses, subsequent floods

very tricky business. There are usually large numbers of un-

eventually buried the remains. Over tens, hundreds and, pos-

knowns relating to the original extent of a bonebed and the

sibly, thousands of years, paleochannels reworked the flood-

heterogeneity of its bone concentrations and preservational

plain, exhuming and secondarily reworking some ceratopsian

patterns, all of which result from the complex interplay of

bonebeds into new paleochannel-hosted bonebeds.

geologic and taphonomic influences (Lyman 1994; Eberth et

The taphonomic data and depositional history interpreta-

al. 2007). In attempting to estimate the size of the individ-

tions of Eberth and Getty (2005) and Eberth and Currie (2005)

ual and combined death assemblages for the Hilda bonebeds,

for DPP apply particularly well to the Hilda mega-bonebed.

we have relied on previous studies at DPP to help us infer

The geologic and taphonomic features described from Hilda

the geometry of the bonebeds. We have also used our esti-

are nearly identical to those identified in the overbank cera-

mates of bone concentration at all of the sites, our excavation

topsian bonebeds farther to the west at DPP (Eberth and Getty

data from H97-04, and our MNI estimates from H97-04 and

2005), thus suggesting that Hilda’s skeletal assemblages also

H97-02. Nonetheless, given the inherent uncertainty due to

experienced large degrees of disarticulation and breakage, and

unknowns, we present our results in the form of order of mag-

variable degrees of abrasion and tooth marking that resulted

nitude estimates (Table 32.1).

from month-to-multi-year exposure times, high rates of soft

We estimated the original areal extent for each of the 14

tissue rotting, and abundant bone reworking due to scaveng-

bonebeds using the assumption—drawn from previous studies

ing, trampling, and minor hydraulic action (cf. Ryan et al.

at DPP—that ceratopsian bonebeds preserved in overbank

2001).

mudstones are roughly circular-to-oval in plan view (Eberth

The greater abundance of mudstone beds in the measured

and Getty 2005). Among the 14 bonebeds examined here, 13

sections at Hilda, while likely indicating a greater degree of

show a maximum exposed linear extent of 50 m or less. Only

accommodation associated with proximity to shoreline and a

H97-04 shows a significantly greater linear exposure (150 m).

A Centrosaurine Mega-Bonebed from the Upper Cretaceous of Southern Alberta 505

The fact that there are no bonebeds whose exposed extent is

Using an order of magnitude approach for these bonebeds,

intermediate between 50 m and 150 m, and that only one has

and given that H97-02 and H97-04 are among the richest and

an exposed extent larger than 50 m (H97-04) suggests to us

largest of the 14 bonebeds, we extrapolate that the size of the

that most of the sites are indeed quite small (in comparison to

death assemblages that contributed to each of these bonebeds

the ceratopsian bonebeds at DPP [Eberth and Getty 2005]).

probably did not exceed hundreds of individuals, and were

Accordingly, we conservatively assume that, with the excep-

more likely composed of tens of individual ceratopsians.

tion of H97-04, all of the Hilda bonebeds have a maximum

Based on the preceding calculations and in the context of

diameter of 50 m, and that they originally covered less than

order of magnitude estimates, we regard the Hilda mega-

2,000 m2 [⌸(50 m/2)2]. Assuming that H97-04 has a maxi-

bonebed as most likely containing the remains of thousands

mum linear dimension of 150 m, we estimate the area of that

of centrosaurs. This conclusion follows from combining the

bonebed to be 17,671 m2.

data from the 14 bonebeds and considering that if bonebeds

Next, we sought to estimate the size of the death assem-

are preserved on either side of the South Saskatchewan River,

blages from H97-04 and H97-02 by projecting the MNI data

then it is very likely that some additional bonebed sites origi-

from the excavation areas at those bonebeds across their esti-

nally occupied the area that was subsequently removed dur-

mated original area. H97-04 yielded a minimum of 2 individ-

ing valley incision and erosion. Whereas we recognize that the

uals collected from 4 m2 in Excavation A, and a minimum of 2

speculative nature of all of these estimates, additional excava-

individuals collected from 8 m2 in Excavation B (Fig. 32.7).

tions at these sites will allow for testing of these numbers.

Both excavations were carried out along the portion of the bonebed exposure exhibiting the greatest bone concentra-

Discussion

tion, which, in turn, reflects an estimated central core area of approximately 2,000 m2 (see above). The remainder of the

The Hilda mega-bonebed is a rare example of a multi-kilometer-

bonebed (estimated at approximately 15,600 m2 area) shows a

scale, macrovertebrate bonebed that formed, in part, due to

bone concentration that is about 10% that of the central core

intrinsic biogenic means (i.e., the gregarious behaviors of the

area (Table 32.1). The average MNI from Excavations A and B

animals entombed; cf. Rogers and Kidwell 2007). Commonly

(average of 2 individuals in 4 m2 and 2 individuals in 8 m2,

known examples of macrofossil bonebeds that occur on this

respectively) is 2 individuals in 6 m . Using this estimate, we

kind of paleogeographic scale result more often from attri-

calculated the size of the death assemblage in the central core

tional accumulation of skeletal remains in a depositional set-

area at 667 centrosaurs (2 individuals for every 6 m2, across

ting that has very low or no sediment supply (hiatal surfaces,

an area of 2,000 m2). Consideration of the remainder of the

paleosols), or as the result of significant erosion events (scour

bonebed suggests that an additional 500–600 centrosaurs

or deflation lags). In either case, such bonebeds are artifacts of

were present (67 individuals for every additional 2,000 m2

geologic history (see numerous examples in Rogers and Kid-

over an area of 15,671 m2). In combination, these calculations

well 2007).

2

suggest that H97-04 resulted from a death assemblage of hundreds to approximately 1,500 individual centrosaurs.

Given the intrinsic biogenic origin of the Hilda megabonebed, and assuming that the mass death event was severe

Our examination of the H97-02 assemblage in the RTMP

enough to kill most or all of the terrestrial macrovertebrates in

collections indicated that a minimum of 6 individuals (see

the area, we propose that the Hilda mega-bonebed probably

above) were preserved in an excavation area of 20 m2 (area

reflects, at least in terms of order of magnitude, the popula-

estimated by us during the 1996 field season). Projecting these

tion of Centrosaurus apertus that was present in the Hilda area

data across the entire bonebed (1,963 m2 or less, Table 32.1),

at the time of the mass-kill event. Accordingly, we envisage a

we estimate a death assemblage of 667 centrosaurs. Again,

loosely assembled ‘‘herd’’ of C. apertus comprising thousands

because it is reasonable to assume that the excavation area was

of individuals occupying the general area. As indicated in the

also likely to have been selected because it was also the area of

discussion to follow, we regard this assemblage as representing

highest bone concentration, we consider it very likely that

a migratory group rather than a stationary group subsisting

this number over-represents the size of the death assemblage.

indefinitely on the locally available plant resources.

In the context of an order-of-magnitude estimate, we consider

Studies of the ceratopsian bonebeds at DPP suggest that

H97-02’s death assemblage most likely to have comprised low

those bonebeds commonly cover areas in the range of 10,000

hundreds of individuals.

m2 and originated from death assemblages that ranged in size

Table 32.1 shows that the other 12 centrosaurine bonebeds

from hundreds to low thousands (Currie and Dodson 1984;

(H97-03, 05–08, 10–11, 13, 16–17, 19–20) range in maximum

Ryan et al. 2001; Eberth and Getty 2005). However, the data

exposed extent from 10 to 50 m, and mostly contain low fossil

from Hilda indicate that, with the exception of H97-04, each

concentrations (typically lower than H97-02 and H97-04).

localized C. apertus bonebed is much smaller in area (ⱕ1,967

506 eberth, brinkman, & barkas

m2) and inferred death assemblage size (tens to low thou-

of the mega-bonebed hypothesis at DPP may be possible using

sands). Even when considered together as a mega-bonebed

trace-element and isotope geochemical analyses of the cera-

covering an area of 2.3 km2, these sites contain a cumulative

topsian bones from the bonebeds at Hilda and DPP (cf. True-

death assemblage estimated in the low thousands. We suggest

man 2007; Fricke 2007).

that this pattern of mostly smaller bonebeds (area and death Acknowledgments

assemblage) at Hilda has an important paleobiological and paleoecological basis.

We thank all the participants of the Tyrrell Museum’s 1997

Brinkman et al. (1998) used subsurface geological data and

Field Experience Program. DAE thanks the collections staff at

vertebrate macro- and microfossils to document that ceratop-

the Royal Tyrrell Museum, and François Therrien and Michael

sians in southern Alberta and western Saskatchewan were

Ryan for helpful and insightful reviews. As always, any errors

most abundant (relative to other ornithischians) closer to

are DAE’s alone.

shoreline during upper Dinosaur Park Formation ‘‘time’’ (Late Campanian). However, they also noted that time equivalent

References Cited

shoreline deposits at Onefour (Alberta) and Unity (Saskatche-

Allison, P. A., and D. E. G. Briggs. 1991. Taphonomy of nonmineralized tissues. In P. A. Allison and D. E. G. Briggs, eds., Taphonomy: Releasing the Data Locked in the Fossil Record, pp. 25–70. New York: Plenum Press. Béland P., and D. A. Russell. 1978. Paleoecology of Dinosaur Provincial Park (Cretaceous), Alberta, interpreted from the distribution of articulated vertebrate remains. Canadian Journal of Earth Sciences 15: 1012–1024. Braman, D. R., and E. Koppelhus. 2005. Campanian palynomorphs of Dinosaur Provincial Park area, Alberta Canada. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 101–130. Bloomington: Indiana University Press. Brinkman, D. B., M. J. Ryan, and D. A. Eberth. 1998. The paleogeographic and stratigraphic distribution of ceratopsids (Ornithischia) in the Upper Judith River Group of Western Canada. Palaios 13: 160–169. Cant, D. J., and G. S. Stockmal. 1989. The Alberta foreland basin: Relationship between stratigraphy and Cordilleran terraneaccretion events. Canadian Journal of Earth Sciences 26: 1964– 1975. Currie, P. J., and P. Dodson. 1984. Mass death of a herd of ceratopsian dinosaurs. In W. E. Reif and F. Westphal, eds., Third Symposium of Mesozoic Terrestrial Ecosystems, pp. 52–60. Tubingen: Attempto Verlag, Tubingen. Eberth, D. A. 1998. Clustered ceratopsian bonebeds, southern Alberta, Canada: Primary evidence for the size of ceratopsianherd death assemblages. The Dinofest Symposium, p. 13. Philadelphia: Academy of Natural Sciences. ———. 2005. The Geology. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 54–82. Bloomington: Indiana University Press. Eberth, D. A., and P. J. Currie. 2005. Vertebrate taphonomy and taphonomic modes. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 453–477. Bloomington: Indiana University Press. Eberth, D. A., and M. A. Getty. 2005. Ceratopsian bonebeds: Occurrence, origins, and significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 501–536. Bloomington: Indiana University Press.

wan) do not preserve monotaxic or monodominant ceratopsian bonebeds. In contrast, they documented that although ceratopsians were significantly less abundant up-dip (200–400 km west) from these shoreline locations in lower Dinosaur Park Formation at DPP, monodominant ceratopsian bonebeds were very common. To understand this pattern of greater ceratopsian abundance but no bonebeds in more shoreward areas (i.e., DPFm at Onefour and Unity) versus lesser ceratopsian abundance but more bonebeds in more upland areas (i.e., lower DPFm at DPP), they proposed that centrosaurines probably exhibited some form of east-west, seasonal migratory behavior. In their hypothesis, these ceratopsians hatched and raised their young in small family groups near shoreline and then, later in the year, migrated westward, gradually assembling themselves into larger ‘‘herds,’’ possibly for protection from predators. Although the westward limit of the migration was not identified, Brinkman et al. (1998) proposed that the migration may have occurred seasonally to avoid annual storms or some other form of seasonal stress, or to take advantage of seasonally available resources in another region. Our data are compatible with the hypothesis of Brinkman et al. (1998). The Hilda bonebeds described here are roughly coeval and 80 km down-dip from those at DPP (Fig. 32.2). Furthermore, the fact that there is 40 m of transgressive DPFm section above the Hilda mega-bonebed horizon indicates that this location was still a significant distance up-dip from shoreline during bonebed formation. Accordingly, within the context of the lower DPFm, the Hilda area is appropriately considered as a paleogeographic intermediate between Dinosaur Provincial Park’s lower Dinosaur Park Formation beds and the coeval (approximately 76 Ma) shoreline deposits in Saskatchewan and southeastern Alberta. Finally, the interpretation that the 14 Hilda centrosaurine bonebeds represent the remains of a single mega-bonebed adds additional weight to the concept that many of the centrosaurine bonebeds in the lower 14 m of DPFm exposure at DPP are likely parts of single mega-bonebeds. Further testing

A Centrosaurine Mega-Bonebed from the Upper Cretaceous of Southern Alberta 507

Eberth, D. A., and A. P. Hamblin. 1993. Tectonic, stratigraphic, and sedimentologic significance of a regional discontinuity in the upper Judith River Group (Belly River wedge) of southern Alberta, Saskatchewan, and northern Montana. Canadian Journal of Earth Sciences 30: 174–200. Eberth, D. A., R. R. Rogers, and A. R. Fiorillo. 2007. A practical approach to the study of bonebeds. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 265–331. Chicago: University of Chicago Press. Eberth, D. A., M. Shannon, and B. G. Noland. 2007. A bonebed database: Classification, biases, and patterns of occurrence. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 103–221. Chicago: University of Chicago Press. Fricke, H. 2007. Stable isotope geochemistry of bonebed fossils: Reconstructing paleoenvironments, paleoecology, and paleobiology. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 437–490. Chicago: University of Chicago Press. Gates, T. A. 2005. The Late Jurassic Cleveland-Lloyd dinosaur quarry as a drought-induced assemblage. Palaios 20: 363–375. Jerzykiewicz, T., and D. K. Norris. 1994. Stratigraphy, structure and syntectonic sedimentation of the Campanian ‘‘Belly River’’ clastic wedge in the southern Canadian Cordillera. Cretaceous Research 15: 367–399. Koster, E. H. 1984. Sedimentology of a foreland coastal plain: Upper Cretaceous Judith River Formation at Dinosaur Provincial Park. Field Trip Guidebook. Calgary: Canadian Society of Petroleum Geologists. Koster, E. H., and P. J. Currie. 1987. Upper Cretaceous coastal plain sediments at Dinosaur Provincial Park, southeast Alberta. In S. S. Beus, ed., Rocky Mountain Section, Centennial Field Guide, pp. 9–14. Denver: Geological Society of America. Koster, E. H., P. J. Currie, D. Eberth, D. Brinkman, P. Johnston, and D. Braman. 1987. Sedimentology and palaeontology of the Upper Cretaceous Judith River/Bearpaw Formations at Dinosaur Provincial Park, Alberta. Joint Annual Meeting,

508 eberth, brinkman, & barkas

Saskatoon, Saskatchewan, Field Trip #10. Toronto: Geological Association of Canada. Lyman, R. L. 1994. Vertebrate Taphonomy. Cambridge Manuals in Archaeology. Cambridge: Cambridge University Press. Rogers, R. R. 1990. Taphonomy of three dinosaur bone beds in the Upper Cretaceous Two Medicine Formation of northwestern Montana: Evidence for drought-related mortality. Palaios 5: 394–413. Rogers, R. R., and S. M. Kidwell. 2007. A conceptual framework for the genesis and analysis of vertebrate skeletal concentrations. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 1–63. Chicago: University of Chicago Press. Ryan, M. J. 2003. Taxonomy, systematics and evolution of centrosaurine ceratopsids of the Campanian Western Interior of North America. Ph.D. diss., University of Calgary, Calgary. Ryan, M. J., A. P. Russell, D. A. Eberth, and P. J. Currie. 2001. The taphonomy of a Centrosaurus (Ornithischia: Certopsidae [sic]) bone bed from the Dinosaur Park Formation (Upper Campanian), Alberta, Canada, with comments on cranial ontogeny. Palaios 16: 482–506. Shipman, P. 1975. Implications of drought for vertebrate fossil assemblages. Nature 257: 667–668. Trueman, C. 2007. Trace element geochemistry of bonebeds. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 397–435. Chicago: University of Chicago Press. Tsujita, C. 1995. Origin of concretion-hosted shell clusters in the Late Cretaceous Bearpaw Formation, Southern Alberta, Canada. Palaios 10: 408–423. Visser, J. 1986. Sedimentology and taphonomy of a Styracosaurus bonebed in the Late Cretaceous Judith River Formation, Dinosaur Provincial Park, Alberta. M.Sc. thesis, University of Calgary, Calgary. Wood, J. M. 1989. Alluvial architecture of the Upper Cretaceous Judith River Formation, Dinosaur Provincial Park, Alberta, Canada. Bulletin of Canadian Petroleum Geology 37: 169–181.

33 Insect Trace Fossils Associated with Protoceratops Carcasses in the Djadokhta Formation (Upper Cretaceous), Mongolia JAMES I. KIRKLAND AND KENNETH BADER

Protoceratops skeletons preserved in the Upper Cretaceous

the Upper Cretaceous Djadokhta Formation at Bayan Zag

Djadokhta Formation and its correlatives in central Asia

(Flaming Cliffs) and Tugrugiin Shireh (Tugrik; Fig. 33.1). (The

are often associated with trace fossils such as borings and

transliteration of Mongolian place names and stratigraphic

diagenetically enhanced burrows in the surrounding

terms is taken from Benton [2000], such that Bayn Dzak and

rock. An articulated skeleton of a Protoceratops, uncovered

Tugrikin Shireh of Fastovsky et al. [1997] become Bayan Zag

at Tugrugiin Shireh, Mongolia, documents the associa-

and Tugrugiin Shireh.) Bored skeletons were observed in both

tion of both insect borings and casts of insect pupation

field areas, as were examples in the preparation lab and ex-

chambers with a dinosaur skeleton for the first time. Doc-

hibits of the Mongolian Academy of Sciences in Ulan Bator.

umentation of the prevalence of borings associated with

The excavation of a large Protoceratops skeleton discovered by

many of the articulated skeletons in these rocks indicates

participant Ed Fox (Fox Protoceratops), preserving both borings

a significant ecological relationship between dinosaurs

and associated insect pupation chambers, provided definitive

and necrophagous insects in an eolian setting.

evidence that insects were scavenging dinosaur carcasses buried in the Djadokhta sediments during the Late Cretaceous.

Introduction

Here, we provide the field observations of bone modification from this Protoceratops skeleton and compare these obser-

Insect modification of dinosaur bone has been reported from

vations to bone modifications observed in other Djadokhta

the Upper Jurassic and Cretaceous of Montana, Wyoming,

fossils and by those produced by modern insects. Identifica-

Utah, South Korea, and Madagascar (Rogers 1992; Jerzykiewicz

tion of these Cretaceous bone-modifying organisms betters

et al. 1993; Fastovsky et al. 1997; Hasiotis et al. 1999; Paik 2000;

our understanding of the taphonomy of dinosaur skeletons

Roberts et al. 2007; Britt et al. 2008). Examples of insect-

and the paleoecology of the Djadokhta Formation.

modified dinosaur bones from the Upper Cretaceous Djadokhta Formation in the Gobi Desert have been reported (Jerzykiewicz et al. 1993; Fastovsky et al. 1997) but not formally described. In 1997, the Dinamation International Society (DIS) with

Methods FIELD

Nomadic Expeditions, in cooperation with the Mongolian Academy of Sciences, under the direction of JIK, led an expedi-

The articulated skeleton of the Fox Protoceratops was dis-

tion into the Gobi Desert of southern Mongolia to investigate

covered on the lower slope to the south of the large mesa,

509

The Protoceratops skeleton was protected by a plaster jacket, carefully reburied and hidden, as the site had been visited by people during excavation (Fig. 33.2). Aware of the specimen’s importance, Anadin Chimedtseren of the Mongolian Technical University, planned on returning in a few days to complete excavation of the specimen. The initial plan was to have the fossil prepared such that additional detailed observations could be made of the pupal cases and the borings. Three days later, he returned with John Horner of the Museum of the Rockies and found that the skeleton had been uncovered by vandals, resulting in damage to the skull (Fig. 33.2). Subsequently, the specimen was not collected and many critical observations such as the description of mandibular marks, thin sections, and more detailed descriptions of the pupal cases could not be made. Therefore, the photographs presented here comprise the only known data salvaged from this specimen. Dozens of other dinosaur skeletons from the Djadokhta Formation have been subsequently examined by JIK at Mongolia Academy of Sciences in Ulan Bator (Fig. 33.4), Field Museum in Chicago, American Museum of Natural History in New York, and in other collections, which include specimens in private hands. Extensive insect damage in the form of borings 0.5 cm in diameter and damaged joint areas, as described below, are exhibited by the majority of all intact skeletons. Many of the Locality maps. Omnigov, Mongolia, and surrounding regions, showing localities where Protoceratops have been recovered. (A) Inset of central Asia; (B) topographic map of Tugrugiin-Shireh. Note active barchan dune in southeastern part of locality. Silhouette of Protoceratops indicates approximate locality of the Fox Protoceratops site with borings and pupae reported here. All modified after Fastovsky et al. (1997). FIGURE 33.1.

dinosaur skeletons on exhibit in various museums have had the borings and damaged joint areas repaired, such that the repairs are visible on close inspection. Although no statistics were kept, of the many dozens of skeletons we have observed over the past decade, only a few did not preserve evidence of being bored.

LABORATORY about 2 km west of the large modern barchan dune at Tugru-

Natural modification of the Protoceratops skeleton, docu-

giin Shireh (Tugrik; Figs. 33.1, 33.2). The specimen was exca-

mented during the excavation, was compared to published

vated using standard paleontological field techniques. Early in

reports of modern and fossil bone modification by insects. To

the excavation, small cylindrical nodules were tossed away as

further our understanding of modern (and, thus, ancient) der-

superfluous. However, it soon was observed that they were

mestid beetle bone modification, modern bones were placed

consistent in size and shape. Although more elongate, they ap-

into aquariums in the University of Kansas dermestid colony.

peared to compare favorably to the isolated pupal chambers we

Experiments were conducted in a mature colony of Der-

had previously identified by comparison with the figures in

mestes maculates, used by the University of Kansas Natural His-

Johnston et al. (1996). It was thereby hypothesized that they

tory Museum and Biodiversity Research Center to clean soft

may represent another form of insect pupation chamber. They

tissue from skeletons for scientific study. Red-legged Ham bee-

were, from that point, left in association with the skeleton. The

tles, Necrobia rufipes (Cleridae), spider beetles (Ptinidae), and

skeleton was exposed in dorsal and left lateral view, and the dis-

small species of dermestids are present in the dermestid col-

tribution of borings and possible pupation chambers were then

ony in minor numbers. Larvae of the American Carrion beetle,

noted. None of the small cylindrical nodules were observed

Necrophila americana (Coleoptera: Silphidae), were found on

more than a few centimeters away from the skeleton. No bur-

discarded bones outside the dermestid colony. The colony is

rows were observed in the sediment around the skeleton. Addi-

maintained in a dark room at 29\C year-round, and is divided

tional burrows not associated with the skeleton were observed

between several standard 10-gallon aquariums and an approx-

by JIK only in areas of the Djadokhta Formation exhibiting

imately 3-foot by 6-foot crate. Screen tops with wood frames

secondary diagenetic enhancement of iron oxide (Fig. 33.3).

cover the aquariums to keep the beetles inside and to exclude

510 kirkland & bader

FIGURE 33.2. Fox Protoceratops site. (A) Fox Protoceratops; (B) Fox Protoceratops site following plaster jacketing and back-filling the excavation; (C) Fox Protoceratops site as documented by Jack Horner 3 days later. The site had been reexcavated and the top jacket had been ripped off, breaking parts of skeleton.

predatory spiders. The aquariums are sprayed with water once

the regional depositional model proposed by Eberth (1993),

a week. A 5-gallon bucket of water is also maintained to keep

this suggests that Bayan Zag and Tugrugiin Shireh represent

the humidity high in the dermestid room.

more distal, basinal aeolian environments.

A variety of lamb, cow, chicken, and turkey bones with at-

Protoceratops is the most common vertebrate encountered in

tached flesh are fed to the dermestids when the museum is not

the Djadokhta Formation. It is consistently found articulated,

actively cleaning skeletons. These bones are removed from the

and many have been observed in a standing posture at a high

colonies after they are thoroughly cleaned of soft tissues. Each

angle to the horizontal, referred to as ‘‘tail standing’’ ( Jerzykie-

bone was examined for evidence of dermestid modification.

wicz et al. 1993; Jerzykiewicz 2000). The taphonomic basis for

To test the burrowing ability of dermestid beetles, an aquar-

the preservation of undisturbed dinosaur nests and articu-

ium with approximately 10 cm of dry coarse sand was added

lated skeletons is still a matter of some debate (Eberth, pers.

to the colony. A bone with adhering flesh was buried to a

com. 2008), but have been interpreted as the result of dune

depth of 2.5 cm in the sand before adding dermestid beetles.

collapse (Loope et al. 1998, 1999) or sandstorms ( Jerzykiewicz et al. 1993; Fastovsky et al. 1997), such that the cause of death resulted from the rapid burial of the animal (obrution).

Previous Work GEOLOGICAL SETTING

TRACE FOSSILS

The Upper Cretaceous (Campanian) Djadokhta Formation of

Previously described invertebrate trace fossils from the Dja-

southern Mongolia is comprised of cross-bedded and struc-

dokhta Formation include three burrow morphotypes: Type

tureless eolian sandstones, sheet sandstones, coarse-grained

1; tube-shaped burrow with meniscate outer surface, Type 2;

interdune deposits, paleosols with caliche horizons, and al-

burrows with pustulate surface textures, and Type 3; complex

luvial fan deposits (Lefeld 1971; Jerzykiewicz et al. 1993;

burrows with radiating shafts (Fastovsky et al. 1997: fig. 10), in

Jerzykiewicz 2000; Dashzeveg et al. 2005). Dashzeveg et al.

addition to insect borings in bone. Throughout the forma-

(2005) divided the Djadokhta Formation into a lower ‘‘Bayn

tion, the best-preserved burrows and root casts are diagen-

Dzak Member’’ dominated by reddish-orange sandstones and

tically enhanced by iron-oxide (Fig. 33.3). JIK observed that

lesser amounts of mudstone and an upper ‘‘Tugrugyin Mem-

burrow fills are not preserved in all horizons, and borings in

ber’’ dominated pale-orange to light-grey sandstones. The de-

bone are not always associated with visible evidence of bur-

positional environment is interpreted as a dune field with pe-

rowing (Fastovsky et al. 1997).

rennial ponds and streams with climate fluctuating between

Johnston (pers. com. 1998) noted that, fine bone material

arid and semiarid ( Jerzykiewicz et al. 1993; Fastovsky et al.

may line a burrow for a short distance after passing through

1997; Eberth 1993, this volume). Caliche horizons were not

the bone. Similar bone-debris filled burrows are associated

observed in the Tugrugyin Member at Tugrugiin Shireh and

with ancient borings reported from Korea (Paik 2000) and Wy-

are not common at in the Bayn Dzak Member at Byan Dzak as

oming (Britt et al. 2008). JIK has observed similar burrows

compared to the famous American Museum of Natural His-

filled with bone meal associated with damaged joint areas in a

tory localities at Ukhaa Tolgod (Loope et al. 1998). Following

partially prepared Protoceratops specimen (Fig. 33.3C).

Insect Trace Fossils Associated with Protoceratops Carcasses in the Djadokhta Formation 511

FIGURE 33.3. Examples of trace fossils observed in the Djadokhta Formation. (A) Left side of a bored Pinacosaurus skull and neck encased in burrows from Bayan Zag (white arrows indicate borings; specimen at the Mongolia Academy of Sciences); (B) right side of (A); (C) example of iron oxide or perhaps iron carbonate cemented burrows and roots at Tugrugiin-Shireh.

512 kirkland & bader

FIGURE 33.4. Borings associated with dinosaur skeletons recovered from the Djadokhta Formation. (A) Protoceratops with borings on exhibit at Mongolia Academy of Sciences; (B) theropod skeleton in preparation lab of the Mongolia Academy of Sciences with damage from borings; (C) Protoceratops foot with burrow back-filled with bone meal extending away from damaged wrist area. White arrows indicate borings. Black arrowhead indicates damaged joint areas. White arrowhead indicates burrow back-filled with bone fragments.

FIGURE 33.5.

Diagram of Protoceratops skeleton indicating positions of borings, damaged joints, and pupae.

Subspherical pupation chambers are unusual fossils that are

rectly through bones whereas others enter larger bones and

commonly encountered in the Djadokhta Formation. Pre-

there is no readily apparent exit (Figs. 33.6, 33.7). Addition-

viously interpreted as small vertebrate eggs or inorganic nod-

ally, semicircular shaped notches at the edge of bones were

ules (Sabath 1991; Jerzykiewicz et al. 1993; Mikhailov et al.

observed that are morphologically similar to borings passing

1994). Johnston et al. (1996) were the first to identify them

through the edges of bone (Fig. 33.6C). The third type of mod-

as beetle pupation chambers constructed by beetles (Coele-

ification is the destruction of articular surfaces, and is most

optera: Curculionidae, Scarabaeidae, or Tenebruionidae) and

abundant at the joints of the limb bones (Fig. 33.7). The de-

assigned them to the new ichnogenus and ichnospecies Ficto-

struction of the ankle and wrist areas is so extensive that there

vichnus gobiensis and F. parvus.

is little remaining of the carpal and tarsal elements. Although Fastovsky et al. (1997) reported Type III burrows penetrating bone surfaces at other sites at Tugrugiin Shireh,

Results

no burrows of any kind were recognized in association with

FOX PROTOCERATOPS

the Fox Protoceratops skeleton, perhaps due to the absence of diagenetic enhancement. Additionally, no traces of bone de-

Excavation of the Fox Protoceratops revealed an articulated skel-

bris were recognized in association with borings in the Fox

eton of a mature ‘‘male’’ Protoceratops, based on its large-size,

Protoceratops, although the light color of the enclosing matrix

erect frill, and large nasal ‘‘horn’’ (Brown and Schlaikjer 1940;

may have obscured the debris.

Dodson 1976), found in a ‘‘standing’’ posture in a poorly indurated, well-sorted, structureless sandstone (Fig. 33.2).

Cylindrical to elliptical sandstone casts that fit the description of beetle pupation chambers ( Johnston et al. 1996), were

Modifications and Associations. Close examination of the

found singly and in clusters in the matrix immediately sur-

specimen in the field revealed a large number of borings and

rounding the right and left jugal, pelvic girdle, and along the

the associated casts of pupation chambers (Fig. 33.5). Three

tail in close proximity to borings in the Protoceratops skele-

types of bone modification were recognized on the Proto-

ton (Figs. 33.6–33.8). A particularly large cluster of pupation

ceratops skeleton. The first type of modification is a nearly

chambers were observed on the lower part of the jugal above

circular boring, 6–10 mm in diameter. Some of these pass di-

and caudal of the articulation of the lower jaw (Fig. 33.8).

Insect Trace Fossils Associated with Protoceratops Carcasses in the Djadokhta Formation 513

FIGURE 33.6. Borings and pupae. (A) Right side of pelvis with numerous borings; (B) left side of pelvis; (C) left side of tail with borings, glancing across neural spines and pupa under caudal rib; (D) close-up of pupa under caudal centra. White arrows indicate borings. Black arrows indicate pupae. Black arrowhead indicates damaged joint areas.

FIGURE 33.7. Skeletal borings and damage. (A) Left rear margin of skull and forelimb with borings and extensive damage to joint areas; (B) right forelimb with extensive damage to joint areas. White arrows indicate borings. Black arrows indicate pupae. Grey arrowhead indicates damaged joint areas.

514 kirkland & bader

the inner surface of the pupation chamber separating it from the surrounding matrix.

OTHER SPECIMENS Other articulated skeletons with damaged bones from the Djadokhta Formation display borings and damage to the joint areas nearly identical to those observed in the Fox Protoceratops. None were observed to preserve pupation chambers. One isolated large ornithischian bone at Tugrugiin Shire was observed to be covered by numerous round pits approximately 3–4 mm in diameter and approximately as deep (Fig. 33.9). No other bone damage of this morphology has been observed by us from the Djadokhta Formation.

Discussion Carcasses decay in a series of stages, with each stage having a unique fauna of necrophagous insects (Payne 1965; Payne et al. 1968; Catts and Haskell 1997). Bone modifying insects (termites, tineid moths, dermestid beetles, and possibly scarab beetles) arrive during the final stage of decay and feed on dried skin, muscle, tendons, ligaments, and occasionally bone. One or more of these insects, or an unidentified extinct species, is likely responsible for the bone modification reported from the Djadokhta Formation. Borings found on the Fox Protoceratops and other skeletons from the Djadokhta Formation bear no resemblance to previous reports of insect-modified bones from the Jurassic and Cretaceous, with the exception of tunnels penetrating bones (e.g., Rogers 1992; Paik 2000; Roberts et al. 2007; Britt et al. 2008). It is clear that the bone damage to the Fox Protoceratops and most other articulated skeletons in the Djadokhta Formation occurred after burial because there is no displacement of any of the bones in the more severely damaged joint areas (Fig. 33.7). Additionally, many bored skeletons have associated traces in the encasing sediment that indicate the borings were made after the carcasses were buried (Figs. 33.3, 33.4). The pattern of damage in bones from the Djadokhta FormaFIGURE 33.8. Mass of pupal chambers associated with jugal of Fox Protoceratops. (A) Right jugal area; (B) close-up of right jugal area. White arrows indicate borings. Black arrows indicate pupae. Black arrowhead indicates damaged joint areas.

tion bears no resemblance to that produced by termites. Bone modification by modern termites occurs during the construction or enlargement of nests or galleries (Derry 1911; Wood 1976; Behrensmeyer 1978; Thorne and Kimsey 1983; Watson and Abbey 1986; Wylie et al. 1987; Haynes 1991; Tappen 1994). The termites build up walls of soil, saliva, and feces (stercoral) from the soil surface to cover the underside of a

These casts are approximately 25 mm long and 10 mm wide

bone. Underneath the stercoral, termites incise small round

with rounded ends and could be distinguished from the sur-

pits in the bone forming linear trails (Tappen 1994). The pits

rounding sandstone matrix by their lighter color and greater

are expanded into branching galleries that follow the bone

induration. Many had a rough exterior, suggesting they were

surface (Haynes 1991).

internal casts of the pupation chamber, whereas others had a

Damage on the Protoceratops skeleton does not follow the

smoother surface, and may have preserved an impression of

pattern of bone modification of tineid moths. Tineid moths

Insect Trace Fossils Associated with Protoceratops Carcasses in the Djadokhta Formation 515

1963; Timm 1982). Larvae will consume bone when other food sources are not available (Hefti et al. 1980) and will bore pupation chambers into hard materials, including dried flesh, wood, and bone (Gabel 1955; Timm 1982). Dermestes maculatus produces pupation chambers in wood and bone with vertical sides and a U-shaped cross section (Martin and West 1995). The absence of pupation chambers within borings in the Fox Protoceratops skeleton suggest that dermestids did not feed on the buried carcasses from Tugugiin Shireh. Additionally, the borings in the Fox Protoceratops average nearly twice as large as those made by modern dermestids or hypothesized to have been made by dermestids on fossil bones (Hasiotis et al. 1999; Britt et al. 2008). Observations made by KB in the University of Kansas dermestid colony indicate that after consuming muscles, tendons, ligaments, and cartilage, dermestid larvae chew the periosteum off cortical bone, leaving a roughened bone surface. Bone modification was restricted to the softer, cancellous bone at the epiphyses of chicken and turkey bones. The dermestid larvae removed the cartilage and outer layer of bone before pitting and then, removing the cancellous bone exposing the pneumatic cavity (Fig. 33.10). Destruction of the joint areas in articulated skeletons from the Djadokhta Formation compares well with these observations. If damage to the joint areas occurred after burial, then it is unlikely that dermestids modified the bones. KB found that the dermestids would not dig through the sand to feed on the buried bones and flesh. Most of the beetles either died in the cage or escaped in search of an alternative food source. In contrast to this test and unlike Hasiotis et al.’s (1999) report indicating dermestid damage was limited to exposed fossil bone, Britt et al. (2008) found no significant preference in borings in the upper versus lower surfaces of fossil bones hypothesized to have been damaged by dermestids. Small, possible dermestid beetle borings preserved in a large isolated ornithischian limb bone at Tugrugiin-Shireh. FIGURE 33.9.

Scarab beetles and possibly tenebrionid beetles are likely candidates for the bone-modifying organism in skeletons from the Upper Cretaceous Djadokhta Formation. Both families of beetles are common in deserts and feed on a variety of

are keratinophagous, feeding on the skin, horns, hooves,

dried plant and animal remains. Tenebrionids feed on car-

hair, or feathers of desiccated carcasses before burial, and do

casses in caves and may damage bone. Scarab beetles feed on

not feed on buried carcasses (McCorquodale 1898; Busck

subaerially exposed and buried carcasses (Payne and King

1910; Bornemissza 1957; Behrensmeyer 1978; Coe 1978; De-

1970; Haglund 1976), constructing burrows and tunnels un-

yrup et al. 2005). Hill (1980) reported that Ceratophaga bore

derneath the carcass. The damaged joint areas are similar to

straight-sided, cylindrical pupation chambers into the as-

the report by Haglund (1976) of damage to buried skeletons

tragali of African bovids. Both tineid moth larvae construct

by scarab beetles. Haglund (1976) noted that larvae of scarab

tubes used as temporary shelter during molting. These tubes

beetles (Anaplognathus) feed on buried Aboriginal skeletons

are composed of silk, earth, and keratin that extend from the

and other organic matter in the soil. Modification of the bones

modified surface down into the soil (Busck 1910; Be-

of these skeletons showed that the soft cancellous bone at the

hrensmeyer 1975, plate 3b) and do not resemble the pupation

epiphyses was eroded, and broad and shallow pits were exca-

chambers associated with the Fox Protoceratops.

vated on the flat cortical bone of the pelvis. This damage is

Dermestid beetle larvae feed on dry carcasses and are often

similar to bone modification by termites or chemical etching

used by museums to clean soft tissue from skeletons (Hinton

by roots (Hasiotis pers. com. 2007). Haglund (1976) believes it

516 kirkland & bader

(Grimaldi and Engel 2005). Adult wasps visit wet carcasses to feed on fluids and insect larvae, but they do not lay eggs on the carcass, nor do their larvae have mouth parts capable of boring into bone (Haskell pers. com. 1999). The pupation chambers cannot be comfortably identified without further comparison with other pupation chambers from the Djadokhta or a large sample of modern insect pupation chambers. However, unlike the round pupation chambers described by Johnston et al. (1996), those associated with the Fox Protoceratops skeleton were elongate, cylindrical structures. The association of borings and pupation chambers on the Fox Protoceratops may be called into question. If these elongate pupal chambers are not found to be consistently associated with dinosaur carcasses, then they may not represent necrophagous organisms. Another cause for the borings could be a fossorial organism burrowing through the bone while moving through the sediment, incidentally penetrating the bone. This interpretation also applies to previous examples of tunnels bored through Cretaceous dinosaur bones (Rogers 1992; Paik 2000; Roberts et al. 2007). The borings on the large ornithischian bone found at Tugrugiin Shireh (Fig. 33.9) compare well with borings attributed to those made by dermestid beetles for pupation (Martin and West 1995; Hasiotis et al. 1999; Britt et al. 2008; Bader et al. 2009). Therefore, these borings may indicate that the carcass of this large dinosaur was exposed at the surface and available Dermestid beetle damage to articular ends of chicken humerus. (A) Undamaged right humerus; (B) dermestid beetle-damaged left humerus. FIGURE 33.10.

to another insect taxon, likely a dermestid beetle.

Conclusions A detailed comparison between pupation chambers associ-

was possible that termites initially modified the bones and

ated with the skeletons and pupation chambers not associated

beetle larvae were present during the excavation.

with the skeletons ( Johnston et al. 1996) is necessary before

Unlike the inflated subspherical-shaped pupation chambers

the traces associated with the Fox Protoceratops and other fos-

of Fictovichnus gobiensis and F. parvus from the Djadokhta For-

sils from the Djadokhta Formation can be properly identified

mation, the pupation chambers associated with the Fox Proto-

ichnologically. Comparisons with the pupae of modern in-

ceratops are subcylindrical with rounded ends (Fig. 33.7B,

sects are hindered by the lack of fine-detailed preservation of

33.8D), and although smaller, are more similar to pupation

the pupation chambers, perhaps owing to the grain size of the

chambers made by Australian weevils (Curculionidae) that

host sediment.

Johnston et al. (1996) illustrated from the Quarternary of Aus-

In the future, researchers should sample the sediment sev-

tralia as Fictovichnus ichnospp. However, weevils most likely

eral meters away from modified skeletons for pupation cham-

did not construct the pupation chambers found in association

bers. If the elongate pupation chambers are only found in

with the modification of the Fox Protoceratops skeleton, be-

association with carcasses, then they were likely produced by

cause modern adult weevils are entirely herbivorous.

necrophagous insects.

Hasiotis (pers. com. 1998; Kirkland et al. 1998) originally

The pupation chambers, borings, and burrows reported

thought these more elongate pupae compared best with those

here record a significant ecological relationship between dino-

of solitary wasps belonging to either the family Pompilidae or

saur carcasses and a large carrion insect, most likely a beetle.

Sphecidae. We now reject the interpretation. Modern sphecids

Further research into these and other trace fossils associated

and pompilids have either very small mandibles or long, thin

with skeletal remains preserved in the Djadokhta Formation

mandibles that are not suitable for chewing through bone (KB

may serve to put environmental constraints on sediment

pers. obs.) These wasps excavate simple burrows with their legs

moisture, substrate texture, and burial depth of the excep-

and construct a chamber to fill with paralyzed arthropods

tional dinosaur skeletons preserved in these strata.

Insect Trace Fossils Associated with Protoceratops Carcasses in the Djadokhta Formation 517

The predominance of insect modified skeletons preserved in the Djadokhta Formation is apparent from a large number of observed cases. However, there is a pressing need to specifically document the pattern of bone damage to skeletons in the field and as they are prepared in the laboratory to substantiate the limited observations presented here. Examples of invertebrate interactions with dinosaurs are rare (e.g., Chin and Gill 1996), and any direct evidence of ecological interactions between different taxa in the fossil record is particularly important to fully document. The field observations of the Fox Protoceratops, reported here, were collected in less than one day at the end of a short visit to the Gobi Desert; even so, a reasonable interpretation of their meaning can be reported here. During the Late Cretaceous, environmental conditions in the deserts of central Asia, regularly resulted in the death and burial of small to medium sized dinosaurs. We hypothesized that adults of a large beetle would detect the decaying carcasses buried below the sand and dig down to them in order to feed and lay their eggs. Perhaps some of the borings and associated burrows associated with the skeletons were made at this stage. The larvae would have fed on the carcass prior to pupating. The last generation of larvae would feed on the dried tendons and cartilage in the joint areas and, subsequently, the bone itself, prior to pupating. The adult beetles would dig their way back up to the surface, perhaps boring through bone on their way, before beginning to search for new carcasses on which to continue their life cycle. Testing of this scenario with further field observations is a critical next step and will expand on our understanding of the desert paleoecosystems during the Late Cretaceous in central Asia. Acknowledgments

Assistance in the field by the participants of the DIS/Nomadic 1997 Expedition to Mongolia is appreciated, especially Carlos Rene Delgado de Jesus (Museum of the Desert, Saltillo, Coahuila, Mexico) and Anadin Chimedtseren (Mongolian Technical University, Ulan Bator, Mongolia). This research benefited from discussions with Rinchen Barsbold, David Fastovsky, Steve Hasiotis, Neil Haskell, Jack Horner, Paul Johnston, and Conrad Labandeira. Carolyn Steahle is thanked for rendering the drawing of the Protoceratops specimen. Robert Timm provided access to the University of Kansas dermestid colony. Reviews by Jennifer Cavin, Don DeBlieux, Mike Lowe, Robert Ressetar, Dave Eberth, and Brooks Britt are appreciated. References Cited Bader, K. S., S. T. Hasiotis, and L. D. Martin. 2009. Trace fossils on dinosaur bones from a quarry in the Upper Jurassic Morrison Formation, northeastern Wyoming. Palaios 24: 140–158. Behrensmeyer, A. K. 1975. The taphonomy and paleoecology of the Plio-Pleistocene vertebrate assemblages of Lake Rudolf,

518 kirkland & bader

Kenya. Bulletin of the Museum of Comparative Zoology 146: 473– 578. ———. 1978. Taphonomic and ecologic information from bone weathering. Paleobiology 4: 150–162. Benton, M. J. 2000. Conventions in Russian and Mongolian palaeontological literature: Mongolian place names and stratigraphic terms. In M. J. Benton, M. A. Shishkin, D. M. Unwin, and E. N. Kurochkin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. xxii–xxviii. Cambridge: Cambridge University Press. Bornemissza, G. F. 1957. An analysis of arthropod succession in carrion and the effect of its decomposition on the soil fauna. Australia Journal of Zoology 5: 1–12. Britt, B. B., R. D. Scheetz, and A. Dangerfield. 2008. A suite of dermestid beetle traces on Dinosaur bone from the Upper Jurassic Morrison Formation, Wyoming, USA. Ichnos 15: 59–71. Brown, B., and E. M. Schlaikjer. 1940. The structure and relationships of Protoceratops. Annals of the New York Academy of Sciences 40: 133–266. Busck, A. 1910. Notes on a horn-feeding lepidopterous larva from Africa. Smithsonian Miscellaneous Collections 56: 1–2. Catts, E. P., and N. H. Haskell. 1997. Entomology and Death: A Procedural Guide. Clemson: Joyce’s Print Shop, Inc. Chin, K., and B. D. Gill. 1996. Dinosaurs, dung beetles, and conifers: Participants in a Cretaceous food web. Palaois 11: 280– 285. Coe, M. 1978. The decomposition of elephant carcasses in the Tsavo (East) National Park, Kenya. Journal of Arid Environments 1: 71–86. Dashzeveg, D., L. Dingus, D. B. Loop, C. C. Swisher III, T. Dulam, and M. R. Sweeney. 2005. New stratigraphic subdivision, depositional environment, and age estimate for the Upper Cretaceous Djadokhta Formation, southern Ulan Nur Basin, Mongolia. American Museum Novitates 3498: 1–31. Derry, D. E. 1911. Damage done to skulls and bones by termites. Nature 86: 245–246. Deyrup, M., N. E. Deyrup, M. Eisner, and T. Eisner. 2005. A caterpillar that eats tortoise shells. American Entomologist 51: 245– 248. Dodson, P. 1976. Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. Journal of Paleontology 50: 929–940. Eberth, D. A. 1993. Depositional environments and facies transitions of dinosaur-bearing Upper Cretaceous redbeds at Bayan Mandahu (Inner Mongolia, People’s Republic of China). Canadian Journal of Earth Sciences 30: 2196–2213. ———. 2010. A review of ceratopsian paleoenvironmental associations and taphonomy. In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 428–446. Bloomington: Indiana University Press. Fastovsky, D. E., D. Badamgarav, H. Ishimoto, M. Watabe, and D. B. Weishampel. 1997. The paleoenvironments of TugrugiinShireh (Gobi Desert, Mongolia) and aspects of the taphonomy and paleoecology of Protoceratops (Dinosauria: Ornithischia). Palaios 12: 59–70.

Gabel, H. H. 1955. Beitrag zur Kenntnis der Biologie des Speckkafers Dermestes vulpinus F. Zeitschrift fur Angewandte Entomologie 37: 153–191. Grimaldi, D., and M. S. Engel. 2005. Evolution of the Insects. Cambridge: Cambridge University Press. Haglund, L. 1976. An Archaeological Analysis of the Broadbeach Aboriginal Burial Ground. St. Lucia: University of Queensland Press. Hasiotis, S. T., A. R. Fiorillo, and R. R. Hanna. 1999. Preliminary report on borings in Jurassic dinosaur bones: Evidence for invertebrate-vertebrate interactions. In D. D. Gillette, ed., Vertebrate Paleontology in Utah, pp. 193–200. Miscellaneous Publication, Utah Geological Survey, 99-1. Haynes, G. 1991. Mammoths, Mastodonts, and Elephants: Biology, Behavior, and the Fossil Record. Cambridge: Cambridge University Press. Hefti, E., U. Trechsel, H. Rufenacht, and H. Fleisch. 1980. Use of dermestid beetles for cleaning bones. Calcified Tissue International 31: 45–47. Hill, A. P. 1980. Early postmortem damage to the remains of some contemporary east African mammals. In A. K. Behrensmeyer and A. P. Hill, eds., Fossils in the Making: Vertebrate Taphonomy and Paleoecology, pp. 131–152. Chicago: University of Chicago Press. Hinton, H. E. 1963. A monograph of the beetles associated with stored products. British Museum of Natural History (reprinted 1945 edition) 1: 1–443. Jerzykiewicz, T. 2000. Lithostratigraphy and sedimentary settings of the Cretaceous dinosaur beds of Mongolia. In M. J. Benton, M. A.Shishkin, D. M. Unwin, and E. N. Kurochkin, eds., The Age of Dinosaurs in Russia and Mongolia, pp. 279–296. Cambridge: Cambridge University Press. Jerzykiewicz, T., P. J. Currie, D. A. Eberth, P. A. Johnston, E. H. Koster, and J. Zheng. 1993. Djadokhta Formation correlative strata in Chinese Inner Mongolia: An overview of the stratigraphy, sedimentary geology, and paleontology and comparisons with the type locality in the pre-Altai Gobi. Canadian Journal of Earth Sciences 30: 2180–2195. Johnston, P. A., D. A. Eberth, and P. K. Anderson. 1996. Alleged vertebrate eggs from Upper Cretaceous redbeds, Gobi Desert, are fossil insect (Coleoptera) pupal chambers: Fictovichus new ichnogenus. Canadian Journal of Earth Sciences 33: 511–525. Kirkland, J. I., C. R. Delgado, A. Chimedtseren, S. T. Hasiotis, and E. J. Fox. 1998. Bored dinosaur skeletons and associated pupae from the Djadokhta Fm. (Cretaceous, Campanian), Mongolia. Journal of Vertebrate Paleontology 18(3, Suppl.): 56A. Lefeld, J. 1971. Geology of the Djadokhta Formation at Bayn Dzak (Mongolia). Palaeontologia Polonica 25: 101–127. Loope, D. B., L. Dingus, C. C. Swisher III, and C. Minjin. 1998. Life and death in a Late Cretaceous dunefield, Nemegt Basin, Mongolia. Geology 26: 27–30.

Loope, D. B., J. A. Mason, and L. Dingus. 1999. Lethal sandslides from eolian dunes. Journal of Geology 107: 707–713. Martin, L. D., and D. L. West. 1995. The recognition and use of dermestid (Insectia: Coleoptera) pupation chambers in paleoecology. Palaeogeography, Palaeoclimatology, Palaeoecology 113: 303–310. McCorquodale, W. H. 1898. Horn-feeding larvae. Nature, London 58: 140–141. Mikahailov, K. E., K. Sabath, and S. Kurzanov. 1994. Eggs and nests from the Cretaceous of Mongolia. In K. Carpenter, K. F. Hirsch, and J. R. Horner, eds., Dinosaur Eggs and Babies, pp. 88– 115. Cambridge: Cambridge University Press. Paik, I. S. 2000. Bone chip-filled burrows associated with bored dinosaur bone in floodplain paleosols of the Cretaceous Hasandong Formation, Korea. Palaeogeography, Palaeoclimatology, Palaeoecology 157: 213–225. Payne, J. A. 1965. A summer carrion study of the baby pig Sus scrofa Linnaeus. Ecology 46: 592–602. Payne, J. A., and E. W. King. 1970. Coleoptera associated with pig carrion. Entomologist’s Monthly Magazine 105: 224–232. Payne, J. A., E. W. King, and G. Beinhart. 1968. Arthropod succession and decomposition of buried pigs. Nature 219: 1180– 1181. Roberts, E. M., R. R. Rogers, and B. Z. Foreman. 2007. Continental insect borings in dinosaur bone: Examples from the Late Cretaceous of Madagascar and Utah. Journal of Paleontology 81: 201–208. Rogers, R. R. 1992. Non-marine borings in dinosaur bones from the Upper Cretaceous Two Medicine Formation, northwestern Montana. Journal of Vertebrate Paleontology 12: 528–531. Sabath, K. 1991. Upper Cretaceous amniotic eggs from the Gobi Desert. Acta Palaeontologica Polonica 36: 151–192. Tappen, M. 1994. Bone weathering in the tropical rain forest. Journal of Archaeological Science 21: 667–673. Thorne, B. L., and R. B. Kimsey. 1983. Attraction of neotropical Nasutitermes termites to carrion. Biotropica 15: 295–296. Timm, R. M. 1982. Dermestids. Field Museum of Natural History Bulletin 53: 14–18. Watson, J. A. L., and H. M. Abbey. 1986. The effects of termites (Isoptera) on bone: Some archeological implications. Sociobiology 11: 245–254. Wood, W. B. 1976. The skeletal material from the Brooloo Range and Rocky Hole Creek burial sites. Archaeology and Physical Anthropology in Oceania 11: 175–185. Wylie, F. R., G. L. Walsh, and R. A. Yule. 1987. Insect damage to aboriginal relics at burial and rock-art sites near Carnarvon in central Queensland. Journal of the Australian Entomological Society 26: 335–345.

Insect Trace Fossils Associated with Protoceratops Carcasses in the Djadokhta Formation 519

34 Faunal Composition and Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis and Other Dinosaurs, Aguja Formation (Upper Cretaceous), Big Bend, Texas J U L I A T. S A N K E Y

new sedimentologic and paleontologic information

berta. The most-complete skull of the ceratopsian Agujacera-

are presented from multiple, closely associated mixed

tops mariscalensis (originally Chasmosaurus mariscalensis [Leh-

bonebeds in the Aguja Formation (Campanian) from the

man 1989]) was collected from a mixed bonebed in the late

Big Bend area of Texas. These bonebeds appear to have

Campanian upper Aguja Formation at Rattlesnake Mountain

been deposited as component parts of channel lags dur-

in Big Bend. However, to date, only a few associated fossils

ing major flooding events. One of these bonebeds yields

have been described from this site, and no sedimentological

the most complete skull of Agujaceratops mariscalensis.

and taphonomic information have been presented. Since dis-

Correlation of paleosols in the bonebed sections with

covery of the skull, hundreds of dinosaur teeth, bones, and

those from the well-studied Dawson Creek section (Big

eggshell pieces have been collected from this site and an-

Bend) provides a critical stratigraphic context for A. mari-

other mixed bonebed that overlies the site. These sites and

scalensis and these bonebeds.

their assemblages are described here, with special emphasis

Cumulatively, all the sites yield a rich assemblage of

on the dinosaurs. The dinosaur component is particularly

plants, invertebrates, and other vertebrates. The com-

rich and is emphasized here because of the biogeographic and

bined vertebrate assemblage provides a means of assess-

biostratigraphic utility of comparing these large terrestrial

ing local terrestrial-community composition in the Late

vertebrates within North America (see Sampson and Loewen

Cretaceous in Big Bend, and for comparing Late Creta-

this volume). Most notably, these dinosaur assemblages in-

ceous southern faunas with those of comparable age

clude (1) teeth and bones from hatchling dinosaurs (abundant

from Montana and Alberta.

hadrosaurs and less common ceratopsians and ankylosaurs); (2) postcranial bones and teeth from tyrannosaurids, dro-

Introduction Big Bend National Park, Texas, is the southernmost area in the

maeosaurids, and ornithomimids; and (3) more than 300 dinosaur eggshell fragments from six different eggshell morphotypes.

United States that yields Upper Cretaceous macrofossil dino-

In combination, all of these data provide a glimpse of

saur remains, as well as rich and diverse vertebrate microfossil

the paleoenvironment and paleocommunity of which Agu-

assemblages. Although this area contains unique vertebrate

jaceratops mariscalensis was a part, and provide opportu-

assemblages, much less is known about them than those of

nities for comparisons with other Late Cretaceous fossil

similar age from regions farther north in Montana and Al-

assemblages.

520

the Big Bend fossil assemblages contain some of the south-

Background BIG BEND NATIONAL PARK, TEXAS

ernmost dinosaurs of the ‘‘Kritosaurus fauna.’’ Based on the presence of abundant soil nodules, this area had dry seasons at least as early as the late Campanian, with

Shoreline, coastal plain, and alluvial sediments were deposited

more intense and frequent dry seasons in the Maastrichtian.

in the Tornillo Basin of west Texas during the Late Cretaceous–

During the Late Cretaceous, Big Bend vertebrates were proba-

Paleocene (Figs. 34.1, 34.2). These deposits are now exposed in

bly better adapted to dry paleoenvironments than their north-

Big Bend National Park. This is a unique terrestrial record be-

ern counterparts (Sankey 1998, 2001).

cause it contains the southernmost Late Cretaceous (latest

One of the dinosaur taxa unique to Big Bend is the cera-

Campanian) through Paleocene terrestrial deposits with verte-

topsian, Agujaceratops mariscalensis (Lucas et al. 2006). In Big

brates from the United States (Lehman 1985; Standhardt 1986,

Bend, it is common in the late Campanian Aguja Formation,

1989; Schiebout et al. 1987; Sankey 1998, 2006, 2008; Nordt et

and is often associated with coastal floodplain sedimentary

al. 2003; Sankey et al. 2007a, b).

deposits. In fact, all but one specimen have been found from

Previous investigations of the latest Campanian through

coastal plain deposits in the lower part of the upper shale

Paleocene alluvial deposits and their paleosols exposed at

member of the Aguja, and A. mariscalensis may have preferred

Dawson Creek in Big Bend have resulted in the development

densely vegetated marshy areas (Lehman 2007). Associated

of a high-resolution stratigraphic and paleoclimatic frame-

skeletons and isolated bones of A. mariscalensis, the hadrosaur

work that includes temperature and rainfall estimates (Fig.

Kritosaurus, ankylosaurs, and tyrannosaurids have been col-

34.2; Nordt et al. 2003; Atchley et al. 2004). The A. mariscalen-

lected from bonebeds in the upper Aguja near Talley Moun-

sis ‘‘mixed bonebed’’ at Rattlesnake Mt. is correlated to the

tain. These bonebeds may have formed during droughts when

Dawson Creek section in this paper. Importantly, the Dawson

the dinosaurs accumulated at watering holes, the skeletons

Creek section is currently the only terrestrial section known

were exposed and trampled, and then were transported and

that records both the middle and late Maastrichtian green-

buried during subsequent floods (Lehman 1982, 1989; Davies

house episodes (Nordt et al. 2003). Additionally, the Tornillo

and Lehman, 1989).

Basin contains an excellent record of the effects of the Western Interior Seaway’s sea-level fluctuations on alluvial deposits. The distance between the shoreline and the Tornillo Basin

MICROVERTEBRATE SITES

fluctuated from 100 to 500 km during the latest Campanian

Although the large dinosaurs from Big Bend, such as A. mari-

through early Paleocene, and was less than 100 km away dur-

scalensis and Kritosaurus, have been described based on mate-

ing deposition of the Rattlesnake Mt. bonebeds during the late

rial from the Talley Mountain bonebeds, considerably less is

Campanian (Atchley et al. 2004).

known about the rarer dinosaurs in the assemblage, such as the theropods, especially compared to the northern fauna.

BIG BEND DINOSAURS

Additionally, little is known about baby dinosaurs and nesting sites in Big Bend. However, in Big Bend, screening and surface

Lehman (1997) recognized differences among the late Campa-

collection of microsites has yielded numerous teeth from baby

nian dinosaur faunas within the Western Interior of North

dinosaurs and from theropods. These sites have also produced

America and named the southern biogeographic province the

thousands of teeth and/or small bones of other vertebrates,

‘‘Kritosaurus fauna.’’ Sullivan and Lucas (2006) argued that this

such as sharks and rays, amphibians, lizards, turtles, croco-

apparent provincialism might be a result of comparing north-

dilians, pterosaurs, and mammals (Standhardt 1986; Rowe et

ern and southern faunas of slightly different ages, and they

al. 1992; Cifelli 1995; Schiebout 1997; Sankey 1998, 2001,

assigned the upper Aguja Formation vertebrate assemblages to

2005, 2006, 2008; Schiebout et al. 1998; Sankey and Gose

the recently named Kirtlandian Land Vertebrate Age (LVA),

2001; Sankey et al. 2005b; Welsh 2005; Nydam et al. 2007;

which spans the time between the Judithian and Edmon-

Welsh and Sankey 2008).

tonian LVA. However, the age of the Aguja is not well enough

For example, microvertebrate sites associated with the Tal-

constrained to test Sullivan and Lucas’s idea (2006; see Leh-

ley Mountain bonebeds have produced more than 3,000 spec-

man et al. 2006 for a detailed discussion about the age of the

imens from 38 vertebrate taxa, and clarified the age and paleo-

Aguja and Javelina Formations).

ecology of these deposits (Sankey 1998, 2005, 2008; Sankey

In terms of numbers and completeness of specimens and

and Gose 2001). The dinosaur teeth include those from baby

diversity of taxa, much less is known about the dinosaurs from

hadrosaurs, ceratopsians, pachycephalosaur, and theropods

the ‘‘Kritosaurus fauna’’ compared to its northern counterpart,

(Saurornitholestes, Richardoestia, and tyrannosaurid), have

the ‘‘Corythosaurus fauna,’’ from Alberta, Montana, Wyoming,

been described in detail, and have been used to document that

and northern Colorado. It is of considerable interest then that

nesting sites occurred in Big Bend (Sankey 2001).

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 521

FIGURE 34.1. (1) Late Cretaceous (Maastrichtian) paleogeographic reconstruction of North and South America (redrawn from Patzkowsky et al. 1991). (2) Late Cretaceous through Paleocene exposures of the Tornillo Basin, Big Bend National Park, Texas, with major field areas labeled.

One advantage of microsites is that they can yield large samples, the size of which are important because (1) typically rare

1986; Sankey 2001; Sankey et al. 2005b; Welsh 2005; Welsh and Sankey 2008).

fossils, such as mammal teeth, are present and can provide

Additionally, unlike rarer macrofossil bonebed sites, the

important biostratigraphic information (Sankey 1998; Sankey

abundance of microsites in Big Bend allows them to be sam-

and Gose 2001); (2) teeth of small sharks and rays can be

pled from numerous different stratigraphic intervals. Com-

recovered, which can document marine influences (Sankey

parison and stratigraphic analyses of Big Bend microsite as-

1998, 2005, 2008); and (3) unusual paleoecological informa-

semblages and their associated facies also provide detailed

tion can be obtained. For example, small teeth from juvenile

information about how Late Cretaceous paleoenvironments

dinosaurs and dinosaur eggshell fragments were the first evi-

and paleocommunities changed through time (Brinkman 1990;

dence that dinosaurs nested in the Big Bend area (Standhardt

Sankey et al. 2005b; Brinkman et al. 2007). For example, strati-

522 sankey

FIGURE 34.2.

(A) Late Cretaceous through Paleocene stratigraphy (redrawn from Nordt et al. 2003); (B) cross section of Aguja Formation showing stratigraphic position of vertebrate sites (modified from Lehman 2007).

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 523

graphic analyses of five microsite horizons in the upper Aguja

Mountain (Texas Memorial Museum 43098-1; Forster et al.

near Talley Mountain and from a sample of 3,349 specimens,

1993). Forster et al. (1993) described the skull as appearing to

show that up section, sharks and rays decrease in abundance,

be from a mature individual based on the level of ossification

thus reflecting the eustatic drop in sea level that occurred at the

of the cranial sutures. Lehman (2007) noted that the skull was

Campanian/Maastrichtian boundary (Sankey 1998, 2008).

from stream channel deposits, which agrees with the inter-

Another advantage of microsites is that large sample sizes

pretations presented in Sankey et al. (2007a, b) and here. In-

can be compared between geographically separate areas in

terestingly, the skull is the only specimen of an Agujaceratops

order to determine biogeographic differences (Brinkman et al.

that has been found from deposits other than coastal marsh

2007). For example, theropod teeth recovered from the Talley

and swamp deposits (Lehman 2007).

Mountain microsites are distinct from northern faunas, at

From 2002 to 2007, intensive surface collection from the

least at the species level, and theropods in Big Bend were pos-

microvertebrate sites preserved in the same deposits that con-

sibly less diverse (Sankey 2001).

tained the skull yielded thousands of vertebrate and other fos-

The A. mariscalensis skull from Rattlesnake Mountain was

sils, and is now one of the largest collections of fossils from the

collected from a ‘‘mixed bonebed’’ (defined by Eberth et al.

upper Aguja Formation in terms of numbers of specimens and

2007 as containing vertebrate microfossils and macrofossils).

diversity of taxa (Table 34.1). Fossils include snails, vertebrate

Hundreds of small teeth and bones from dinosaurs were col-

coprolites, fish (gar), salamander, lizard, turtle, crocodilian,

lected from this site and those closely associated with it. The

pterosaur, dinosaur, bird, and mammal (Sankey 2005, 2006;

dinosaurs are described here, which is important because

Gasaway et al. 2007). Study of the salamander, lizard, and

(1) the dinosaurs associated with Agujaceratops mariscalensis

crocodilians are in progress. However, the turtle and dinosaur

are not well known, and (2) from this large sample, rare speci-

eggshell specimens have been studied and described in detail,

mens were found such as eggshells, bones and teeth from

including counts of specimens, and those findings are sum-

hatchlings, and rare dinosaurs such as ankylosaurs and thero-

marized below.

pods. These data fill an important gap in our knowledge about

Turtle shell fragments are the most common fossils, and

the dinosaurs in Big Bend in general and the dinosaurs associ-

hundreds of specimens have been collected. Taxa include

ated with Agujaceratops mariscalensis in particular. Together

Baenidae indet., cf. Hoplochelys, Adocus, Basilemys, and cf.

these data represent an important baseline for future biogeo-

Helopanoplia and other trionychids. From a study of 388 turtle

graphic and comparative stratigraphic studies.

specimens collected primarily from these microsites, trionychids are the most abundant (81%), with other taxa consider-

Materials and Methods

ably less common: Adocus (6%), Basilemys (6%), baenids (5%), and kinosternoids (1%; Sankey 2006). Compared to the turtle

The measured section was trenched to unweathered rocks (ap-

diversity in the overlying, more inland, Maastrichtian-aged

proximately 0.5 m). Beds were measured using a hand level

Javelina Formation (Tomlinson 1997), turtles were more di-

and a Jacob’s Staff marked with centimeter gradations. Rock

verse in the Aguja Formation, probably reflecting a warmer

and paleosol color designations were assessed using a GSA rock

and wetter climate (Sankey 2006).

color chart (GSA 1991). Most fossils were surface collected.

In 2002, Ed Welsh recognized and collected dinosaur egg-

Some fossils were collected in situ. A sample of sediment from

shell fragments from the microsites (Welsh 2005). During

the bonebeds was wet screened and picked for small fossils,

2005 and 2006 we increased this sample to over 300 dinosaur

using a microscope. All fossils are curated in the Louisiana State

eggshell fragments; all are described by Welsh and Sankey

University Museum of Natural Science (LSUMNS) Geology

(2008). Six different eggshell morphotypes are present, in-

Collections in Baton Rouge. Locality data are also on file at that

cluding those from ornithischian and theropod dinosaurs.

institution. A complete LSUMNS catalog number contains the

The six morphotypes include dinosauroid spherulitic, dino-

locality and specimen numbers, separated by a colon (e.g.,

sauroid prismatic, cf. ornithoid prismatic, ornithoid ratite,

834:17621). Most fossils were coated with vaporized am-

and two types tentatively referred to the oogenera Continuoo-

monium chloride and photographed using a Nikon E995 with

lithus and Porituberoolithus (Welsh 2005; Welsh and Sankey

a Nikkor lens, attached to a Wild Leitz MSC microscope.

2008). Many of the eggshells have spherulitic shell units, typical of hadrosaur eggs and some eggshells have a mammilary layer and continuous layer, typical of theropods. Many of the

Results

eggshells have the angusticaniculate type of pore canal sys-

RATTLESNAKE MOUNTAIN MICROSITES

tem, common in eggs from arid environments (Mikhailov 1997; Carpenter 1999). Importantly, the eggshells do not

In 1991, the most complete skull of an Agujaceratops mari-

closely match any eggshell types from contemporaneous

scalensis was collected from the upper Aguja at Rattlesnake

northern faunas (Zelenitsky et al. 1996), further supporting

524 sankey

Table 34.1. Vertebrate Taxa from the Rattlesnake Mountain Bonebeds (Sankey 2006; Sankey et al. 2005; Welsh 2005; Welsh and Sankey 2008)

Sedimentology and Stratigraphy. The Rattlesnake Mountain microsites are hosted by grey, organic-rich silty mudstones. Sediments have abundant tiny fragments of coalified plants, large pieces of burned wood, and large and small isolated

Fish

bones. The microsites are located approximately 10 m below

Dasyatidae

the first moderately developed paleosol in the inland flood-

Lepisostidae

plain facies of the upper shale member of the Aguja in the

Amphibian Albanerpeton

Rattlesnake Mountain area (Sankey et al. 2007a, b, in prep).

Turtle

The microsites were correlated to the base of the Dawson

Baenidae

Creek section, 7 km to the north of Rattlesnake Mt., using

Kinosternoidea cf. Hoplochelys

paleosols. We were able to correlate the paleosols directly

Adocidae-Adocus

above the Rattlesnake Mt. microsites to paleosols #42 and #43

Nanhsiungchelyidae-Basilemys

at the base of the well-studied Dawson Creek section (Nordt

Plastomeninae-Helopanoplia

et al. 2003). Our correlations were based on similarities in pa-

Trionychidae-Genus et sp. indet.

leosol morphology, and carbon and oxygen isotope values

Lizard

(Sankey et al. 2007a, b, in prep). In addition to similarities in

Crocodylians

paleosol morphology and color, the carbon and oxygen ratios

Brachychampsa

from the carbonate soil nodules in the Rattlesnake Mt. pa-

Goniopholidae

leosols are similar to those from the Dawson Creek paleosols

Pterosauria

#42 and #43 (Sankey et al. 2007a, b, in prep.).

Theropod-Richardoestesia

The ‘‘Purple Hill’’ section is 11.4 m thick (Fig. 34.3) and

Aves Nodosauridae cf. Edmontonia

begins at the base of the southern tip of the small hill (locality

Hadrosauridae

VL-842). The lower 1 m contains light olive grey (5Y 5/2),

Ceratopsia cf. Agujaceratops mariscalensis

poorly sorted, moderately well-indurated, muddy sandstone

Tyrannosauridae

with abundant clay balls, pedogenic carbonate nodules, small

Ornithomimidae

plant fragments, snails, and small and large vertebrate bones

Dromaeosauridae-Saurornitholestes

(Bonebed #1; Fig. 34.3). The deposits fine and lighten in color

Mammalia, Multituberculata

upwards, with less plant and bone material in the uppermost 0.9 m. The section continues at the southern base of ‘‘Purple Hill’’ (locality VL-747). Between the 2.5 and 4.5 m level are

one of these conclusions proposed here and previously on the

medium-dark grey (N4), muddy sandstones, with abundant

basis of teeth that Big Bend dinosaurs were distinct, at least at

clay balls, pedogenic carbonate nodules, small plant frag-

the species level, from those in northern faunas (Sankey 2001;

ments, snails, and small and large vertebrate bones (Bonebed

Sankey et al 2005b). Based on the variety of eggshell morpho-

#2; Figs. 34.3, 34.4A). Between the 4.5 and 6 m level are

types in the microsites, it seems reasonable to conclude that

coarser and lighter-colored sands (10 YR 6/6), with common

a variety of dinosaurs either nested together or that their nest-

gypsum casts of large wood and bones (Fig. 34.4B, C). At the 6

ing sites were in close proximity (Welsh 2005; Welsh and

to 7.7 m level there are yellowish grey (5Y 7/2), moderately

Sankey 2008).

well-cemented, muddy, fine-grained sandstones. Plant frag-

Sankey et al. (2005b) described the theropod teeth made

ments, vertebrate fossils, and clayballs are absent. At the 7.7 to

during the initial (2002) surface collection of these micro-

10.2 m level there are mudstones, with moderately developed

sites, in addition to the theropod teeth collected from the

paleosols. The lowest palesol (Paleosol #1) is 1.5 m thick, and

Maastrichtian-aged uppermost Aguja and Javelina formations

greyish-red (5R 4/2) with small areas of green staining. This is

(Standhardt 1986), and noted differences between the Cam-

the purple stratum referred to in the name ‘‘Purple Hill,’’ and

panian and Maastrichtian theropods in Big Bend. However,

is easily recognized and useful for local stratigraphic correla-

since this initial report, hundreds of additional dinosaur teeth

tion (Fig. 34.3). Paleosol #2 (Fig. 34.3) is 1 m thick, brownish

have been collected, including numerous teeth from embry-

grey (5YR 4/1) with small areas of green staining (i.e., iron

onic and hatchling hadrosaurs, and lesser amounts from

depletion due to episodic anaeorobiosis). This paleosol is ero-

hatchling ceratopsians, ankylosaurs, and theropods. These are

sionally overlain by 1.2 m of moderately sorted, moderately

described here. Dinosaur teeth are difficult to age, so the term

indurated, yellowish grey (SY 7/2) sandstone with rip-up

hatchling is used here to refer to teeth less than 15% adult size;

clasts of the underlying green mudstone.

these are assumed to be from individuals that are less than two years old, following Lehman (2007).

Most of the fossils were collected from the surface, and some were collected in situ (Fig. 34.4B) from the two deposits of

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 525

FIGURE 34.3. ‘‘Purple Hill’’ stratigraphic section. Photograph shows Rattlesnake Mountain in the distance (view to the north). In the foreground is ‘‘Purple Hill,’’ with the distinct purple paleosol near the top. Person is working at Bonebed #2. The skull of A. mariscalensis (TMM 43098; Forster et al. 1993) is from Bonebed #1.

526 sankey

the Rattlesnake Mt. bonebeds appear to match Eberth’s high energy, in-channel flow facies. The 1.9 m thick interval comprising stacked mudstone beds near the top of the section is interpreted as an overbank floodplain setting in which muds episodically settled out from suspension. The absence of erosional contacts and the presence of numerous moderately developed paleosols through the interval indicate that an alluvial plain paleoenvironment was established for a considerable time in this area. However, the top paleosol is truncated by a disconformity overlain by a paleosol/mudstone-intraclast-rich sandstone. This stratigraphic arrangement indicates the start of a major channel incision perhaps due to channel avulsion into this area, or, possibly, a regional drop in relative sea level.

Systematic Paleontology Ankylosauria Osborn 1923 Nodosauridae Marsh 1890a cf. Edmontonia Sternberg 1928 (Figs. 34.5, 34.6)

Referred Specimens. 834:17621 and 746:17849 (osteoderms); 834:17859 (ossicle); 834:17749, 746:6271, and 746:8439 (teeth). Sedimentary rocks and fossils from the ‘‘Purple Hill’’ field area. (A) Fossiliferous muddy sandstone with abundant small plant fragments from Bonebed #2; (B) in situ dinosaur vertebra from Bonebed #2; (C) coalified tree limb (1 meter in length) from surface of Bonebed #2; (D) pedogenic carbonate nodules within sandstone overlying Bonebed #2. FIGURE 34.4.

Description. 8439 is a small tooth (6 mm anterior to posterior dimension and 6 mm top to base). The anterior denticles are large (1 mm in width). The surface is rugose, with coarse wrinkles. Both the lingual and labial sides have a shelf or cingulum. The anterior surface of the tooth, from tip to base, shows evidence of tooth wear. 6271 is a small, unworn tooth (8 mm from tip to base; 6 mm greatest FABL; Fig. 34.6B, C). Denticles are large and pointed (4 on posterior side and

poorly sorted, muddy sandstone (Bonebeds #1, #2). The A.

5 on anterior side). Specimen 17749 is an extremely worn,

mariscalensis skull site is in Bonebed #1.

probably digested, tooth. No denticles are present, but the

Depositional Environments. The ‘‘Purple Hill’’ stratigraphic

typical shape of an ankylosaur tooth remains.

section is interpreted as a succession of interbedded chan-

Specimen 17849 is a small, complete osteoderm (Fig. 34.5H,

nel and over-bank deposits (Fig. 34.3). The two main bone-

I). It is oval in outline. The exterior side has a single conical

producing deposits (Bonebeds #1, #2) are muddy sandstones

point, but no keel, and the interior side is flat to slightly con-

at the base of their own fining upward successions, and are

cave. The surface is pitted, grooved, and rugose. The edges are

interpreted as fluvial channel deposits that were deposited

scalloped shaped. Specimen 17621 is a fragment of a larger

during upper flow-regime flow. The overlying upward-fining

osteoderm. It is 55 mm in diameter. The surface has deep

successions of finer sandstone were probably deposited during

grooves and pits. The edge is fluted and rugose. Specimen

lower energy flows.

17859 is a dermal ossicle. It is oval in outline. The surface is

Eberth (1990) described two main sedimentary facies in

pitted.

which vertebrate microfossil sites accumulated in the late

Discussion. The teeth are clearly ankylosaur based on their

Campanian beds of the Oldman and Dinosaur Park forma-

distinctive morphology, specifically the conical shape in lat-

tions at Dinosaur Provincial Park, Alberta: (1) mudstone peb-

eral outline; the enlarged, bulbous base; the large, pointed

ble intraclast deposits that formed during high energy, in-

denticles (i.e., cusps), with the apical cusp the largest; and

channel flow, and (2) tabular or sheet-like, interbedded silt-

the teeth are slightly labially lingually compressed. In the

stone/sandstones that were deposited as overbank splays.

nodosaurids, teeth are generally bigger and often have a shelf

Based on the more limited sedimentary data available here,

(cingulum) at the base of the tooth crown instead of a swollen

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 527

base of the crown in ankylosaurids (Carpenter 1997). Ankylo-

bryonic dinosaur teeth have wear on their tips, possibly from

saurid teeth have smoother, less wrinkled surfaces (Carpenter

grinding their teeth while still in the egg. This is seen in hadro-

1982). (See Coombs 1990 and Carpenter 1997 for descriptions

saurs, sauropods, and Psittacosaurus. The other small tooth is

of ankylosaur teeth.)

a digested tooth; possibly evidence for predation. The baby

Both of the two unworn Big Bend teeth have a cingulum,

dinosaur teeth and the abundance of eggshell material in the

although it is larger in specimen 8439, and both have surfaces

Rattlesnake Mt. bonebeds indicate that nesting sites were

that are wrinkled and rugose. The teeth more closely resemble

nearby. Also similar to the Hell Creek site, the Rattlesnake

those of nodosaurids than of ankylosaurids.

Mt. bonebeds contains baby dinosaur teeth from a variety of

The scutes of ankylosaurs are typically flat or keeled in the

taxa. However, what is not discussed in Carpenter (1982) is

neck and shoulder regions and over the back and tail, scutes

how and why a variety of baby dinosaur fossils are preserved

are smaller and keeled (Carpenter 1997). Located between the

together in the Hell Creek site. Were multiple nesting sites

areas with large scutes are abundant small, irregular ossicles,

eroded and transported or did nesting sites include different

which gave the animals flexibility (Carpenter 1997). The small

species of dinosaurs?

(pebble-sized) dermal ossicle (17859) is similar to those illus-

With some exceptions, most North American ankylosaur

trated for Edmontonia (Carpenter 1997: fig. 21.4). The small,

fossils are typically isolated elements or fragmentary skeletons

oval, complete osteoderm (17849) with the conical point, but

(Vickaryous et al. 2004). In general, ankylosaurs were rare, a

no keel, matches the small skull dermal ossicles illustrated in

minor component of Late Cretaceous dinosaur faunas (Car-

Coombs (1978).

penter 1990). This is also true for Big Bend, where isolated

In a recent review of ankylosaurs, Vickaryous et al. (2004)

ankylosaur bones and osteoderms are usually a minor compo-

list the nodosaurids Edmontonia sp. and Edmontonia rugosidens

nent in bonebeds of hadrosaurs and ceratopsians in the upper

(= Panoplosaurus rugosidens Gilmore 1930) as present in the

Aguja. Several quarries were collected by the WPA (Works

Aguja of Big Bend. The latter species is known from a skull

Progress Administration) crews in the 1930s near Talley Mt.,

without lower jaws and from other isolated bones from Big

and in one example (Quarry #3), ankylosaurs make up 29%

Bend (Coombs 1978). Edmontonia occurs from Texas to Al-

out of 82 prepared specimens (Lehman 1982; Davies and Leh-

berta (Vickaryous et al. 2004). There are few other published

man 1989). Lehman (1997) estimates that ankylosaurs were

reports on ankylosaurs from Big Bend, and most are brief,

only 8% of the dinosaur fauna in Big Bend. However, based on

without detailed descriptions or illustrations. Lehman (1985:

extensive collection of the productive Rattlesnake Mt. bone-

263) mentioned the presence of ankylosaur osteoderms, limb

bed, where only three teeth (all juveniles) and three osteo-

bones, and vertebrae in the upper Aguja, and referred them

derms have been found, ankylosaurs are may have been an

to the nodosaurs based on the ‘‘flattened bases of the osteo-

even rarer component, at least in this particular paleoenviron-

derms, a condition present in Nodosaurinae.’’ Standhardt

mental setting. Based on North American localities of ankylo-

(1986) referred a small ankylosaur tooth (LSUMNS 1305) from

saur tracksites (McCrea et al. 2001) and the presence of aquatic

the upper Aguja of Dawson Creek (site LSUMNS 113) to the

vertebrates associated with ankylosaur bone and teeth, an-

ankylosaurid cf. Euoplocephalus, based on its resemblance to E.

kylosaurs probably preferred wet environments (Vickaryous

tutus (Lambe 1902: 57, 152), in particular the ‘‘leaf-shaped

et al. 2004). However, many Asian ankylosaurs lived in arid

crown with large serrations and the small size of the crown

or semi-arid environments (Carpenter 1997) and their rar-

relative to the root.’’ The specimen was mentioned in Stand-

ity in Big Bend and other coastal and floodplain settings

hardt (1989) and photographed in Schiebout (1997). This

may indicate that they were more common in more inland

tooth resembles those described in this paper, which are tenta-

areas that are not well documented in the North American

tively identified as Edmontonia. Rowe et al. (1992) also men-

fossil record.

tioned ankylosaur osteoderms from the Terlingua microsite in the upper Aguja Formation. Based on these reports of isolated

Ornithopoda Marsh 1881b

and fragmentary fossils, ankylosaurs were rare in Big Bend,

Hadrosauridae Cope 1869

and most are referred to the nodosaurids.

(Figs. 34.5, 34.6)

Importantly, the three small ankylosaur teeth described here are similar in size to teeth of baby ankylosaurs (i.e., em-

Referred Specimens. Hundreds of isolated small teeth and several small bones from hatchlings.

bryonic or hatchling) from the Hell Creek Formation (Carpen-

Description. These teeth are clearly from hadrosaurs because

ter 1982: 128), documenting that ankylosaurs nested in Big

they have the diagnostic ridge on the outer surface of the

Bend. Similar to the Hell Creek site, these two Big Bend teeth

tooth (Carpenter 1999). Hundreds of small hadrosaur teeth

are unworn, indicating little abrasion during transport prior

are the most abundant dinosaur fossils in the bonebeds (Fig.

to burial. Horner and Currie (1994) showed that some em-

34.6A). Tooth size matches those of hatchling dinosaurs.

528 sankey

FIGURE 34.5. Dinosaur postcranial bones from the Rattlesnake Mountain bonebeds. (A–G) Hadrosaur hatchling centrae (LSUMNS 842:8397); (H–I) ankylosaur scute (LSUMNS 834:17621); ( J–L) ornithomimid manual ungual (LSUMNS 726:8347); (M–P) ornithomimid ungual (LSUMNS 747:19948); (Q–T) dromaeosaurid pedal phalange (LSUMNS 726:6212); (U–X) cf. tyrannosaurid manual phalange (LSUMNS 746:8440); (Y–A1) dromaeosauridae metacarpal (LSUMNS 839:17756).

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 529

Dinosaur teeth from the Rattlesnake Mountain bonebeds. (A) Hadrosaur hatchling, occlusal view (LSUMNS 834:4241); (B– C) ankylosaur hatchling, side views (LSUMNS 746:6271); (D–I) tyrannosaurid juvenile (LSUMNS 746:8374); (D) labial view; (E) lingual view; (F) denticles on posterior carina; (G) anterior view; (H) basal view; (I) denticles on posterior carina; ( J–L) tyrannosaurid juvenile (LSUMNS 746:8414); ( J) labial view; (K) lingual view; (L) basal view; (M–O) tyrannosaurid juvenile premaxillary (LSUMNS 726:6218); (M) anterior view; (N) labial view; (O) basal view. FIGURE 34.6.

Bones from hadrosaur hatchlings are also present, such as

Theropoda Marsh 1881a

three small vertebral centrae with the notochordal pit present

Family Tyrannosauridae Osborn 1905

(Fig. 34.5A–G). This is a characteristic of baby (i.e., embryonic

Tyrannosauridae indeterminate

or hatchling) hadrosaurs; the pit usually disappears after the

(Figs. 34.5, 34.6)

hatchling stage (Weishampel and Horner 1990). Discussion. Hadrosaurs are the most common embryonic

Referred Specimens. LSUMNS 726:17896, 6219, 8236, 6209,

dinosaurs found, and are associated with Spheroolithus eggs

6201, 6220, 6221; 842:8375; 746:17803, 6272, 6274, 8368,

in Alberta, Montana, and Mongolia (Carpenter 1999). In the

8291, 6262, 6227, 8252; 834:8416, 8332, 8243, 8323; (tooth

Aguja, hadrosaur specimens are more common than ceratop-

fragments); 842:17876 (complete tooth); 746:6282 (frag-

sians by about 60% (Lehman 2007). Hadrosaur hatchling

ment of premaxillary tooth); 746:8371; 726:6218, (complete

teeth are the most abundant teeth in the bonebeds; many of

premaxillary teeth); 746:8374, 8414 (4 specimens), 6247;

these teeth are small and unworn. The most abundant egg-

834:8211; 726:8217, (complete teeth); 746:8440 (1 manual

shell type in the bonebeds match those of hadrosaurs (Welsh

phalanx).

and Sankey 2008).

Description. Most specimens are fragments of teeth, but are easily identified as tyrannosaurid based on distinctive tooth

Ceratopsia Marsh 1890a

and denticle shape and size. Nine complete and nearly com-

cf. Agujaceratops mariscalensis Sullivan and Lucas 2006

plete teeth range in size, from small (8 mm in length) to large

(= Chasmosaurus mariscalensis Lehman 1989)

(40 mm), with an average of 24 mm. Lengths were measured

Referred Specimens. Uncounted, fragmentary and shed teeth.

along a straight line between tip and base of tooth, where the

Description. Teeth match those referred to as ceratopsian

enamel ends. Widths vary from 6 to 18 mm, with an average

teeth in the Judith River Group of Alberta (Baszio 1997; Peng

of 13 mm. Width refers to the size of the tooth, mesially dis-

et al. 2001).

tally, and was measured at the base of the tooth. Lateral teeth

Discussion. Isolated shed ceratopsian teeth are present, but

(i.e., non-premaxillary) vary in cross-sectional shape from

they are rare compared to the common hadrosaur teeth. An

oval to flattened oval. Degree of recurvature ranges from

epioccipital process was also collected (identified by Sampson,

strongly to slightly recurved. Denticles are present on both

pers. com. 2007).

the anterior and posterior carinae (carinae are the serrated

530 sankey

ridges that extend along the anterior and posterior edges of

Some of the teeth from the Rattlesnake Mt. bonebeds are

theropod teeth). Denticles are approximately equal in size on

clearly from juveniles. For example, 746:8371 is a small, com-

both carinae, although slightly larger on the posterior. Den-

plete premaxillary tooth (6 mm in length) and 842:17876 is a

ticle size is approximately uniform from tooth base to tooth

small, complete tooth (8 mm in length). Even fragmentary

tip, although slightly smaller near the tooth base and tip.

teeth can be referred to small teeth, especially if it is a tooth tip

In most teeth, denticles are broad and wide, with bulbous

or if two sides of the tooth remain, giving a sense of tooth size.

tips. On unworn denticles, tips are also slightly pointed. In

Many other Rattlesnake Mt. teeth fragments are from small

some teeth, denticles are longer, narrower, and tightly packed

teeth, and probably also from juveniles. Some of the complete

(i.e., smaller interdenticle spaces). Denticles/mm on posterior

teeth are labial-lingually flattened and recurved, and this may

carinae range from 2 to 3/mm. (All denticle measurements

be due to variation of tooth shape along the tooth row (Sam-

were made from the complete teeth only, and from the largest

man et al. 2005). Additionally, juvenile tyrannosaurid lateral

denticles present on the posterior carinae.) In premaxillary

teeth are more bladelike (i.e., flattened labial-lingually) than

teeth, both carinae are on the anterior surface, teeth have little

in adults, where teeth are broader and can be almost as thick

recurvature, are round in cross section, and denticles are quite

(labial-lingually) as wide (mesial-distally) (Holtz 2004). So,

small. Two complete and nearly complete premaxillary teeth

these bladelike teeth may be from juveniles. Many of the ty-

are 6 and 16 mm in length.

rannosaurid teeth from the Rattlesnake Mt. bonebeds are

Additionally, there is one tyrannosaurid manual phalanx (Fig. 34.5U–X; Longrich pers. com. 2007).

from juveniles, and a few are from quite small individuals; possibly from hatchlings. Finding juvenile tyrannosaurid

Discussion. Sankey et al. (2005b) described and measured 13

teeth along with small theropod and other dinosaur teeth is

tyrannosaurid teeth and teeth fragments collected from the

common from other Late Cretaceous deposits (Currie et al.

Rattlesnake Mt. bonebeds (LSUMNS localities: 726, 746, 747,

1990).

834) from the initial fieldwork in 2002, in addition to those from other sites in the Aguja and from the Javelina Forma-

Family Ornithomimidae Marsh 1890b

tions. The subsequent years collecting from the Rattlesnake

(Fig. 34.5)

Mountain bonebed sites has increased the sample by 32 speci-

Referred

mens, including 9 complete teeth. These are described here. Tyrannosaurid teeth are easily identifiable based on charac-

Specimens.

Phalanges

(841:17764;

727:5913;

746:8326; 747:17769). Unguals (728:8347; 746:6260, 17755; 747:17768, 19948)

teristics of their distinctive tooth and denticle size and shape.

Description. The most common theropod post-cranial ele-

Tyrannosaurid teeth are described in detail in Currie et al.

ments are from ornithomimids. There are three morphotypes:

(1990), Abler (1997), Baszio (1997), Sankey et al. (2002), and

(1) a small ornithomimid, (2) a Struthiomimus-sized ornitho-

Samman et al. (2005), and these characteristics are summa-

mimid, and (3) a very large, Ornithomimis-sized ornithomi-

rized here. Typically, the teeth are rounder in cross section

mid (Longrich pers. com. 2007). An example of the latter is

(i.e., less flattened, labial-lingually) than from other thero-

747:19948 (Fig. 34.5M–P).

pods such as the dromaeosaurids. Both carinae have similar-

Discussion. Although Lehman (1985) reported fragmentary

sized denticles and their size is consistent from base to tip of

ornithomimid remains from Big Bend, this is the first report

tooth. This is different from dromaeosaurid teeth, which have

with illustrations.

considerably larger denticles on the posterior carinae and less uniformly sized denticles along the carina. Denticle shape and

Family Dromaeosauridae Mathew and Brown 1922

size is also distinctive in tyrannosaurids. Denticles usually oc-

Saurornitholestes Sues 1978

cur 3/mm (Currie et al. 1990), with clear interdenticle spaces.

Saurornitholestes indet.

Denticle tips are bulbous, and slightly pointed if unworn.

(Fig. 34.7)

Sankey (2001: figs. 3.6, 3.7, 3.13, 3.14) referred two incom-

Referred Specimens. In addition to those referred to in Sankey

plete teeth from the Talley Mt. microsites to theropod indet.,

et al. (2005b) are the following: 726:6204; 746:6270, 6280,

based partly on their long, narrow, closely spaced denticles

6281, 8254, 8327, 8368, 8449, 17750, 17751, 17754, 17803,

that did not closely resemble those from tyrannosaurids or

17783, 17786, 747:8282; 841:17752.

from small theropods. However, among the Rattlesnake Mt.

Description. Saurornitholestes are the most abundant thero-

sample are additional teeth with similar denticles. Some are

pod teeth in the bonebeds (Fig. 34.7). See Sankey et al. (2005b)

complete teeth, such as 746:8374, which is clearly a tyran-

for detailed descriptions.

nosaurid. Juvenile tyrannosaur teeth may contain denticles

Discussion. Sankey et al. (2005b) described the theropod

that are slightly different from those of adult teeth, and this

teeth from these bonebeds in detail, and distinguished three

may be the reason for the variation seen here.

distinct morphotypes of Saurornitholestes. No new taxa are

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 531

FIGURE 34.7. Theropod teeth from the Rattlesnake Mountain bonebeds. (A–E) Saurornitholestes sp. (LSUMNS 726:6204); (A) labial view; (B) lingual view; (C) denticles on posterior carina; (D) anterior view; (E) basal view; (F–J); Saurornitholestes sp. (LSUMNS 746:8449; (F) labial view; (G) lingual view; (H) denticles on posterior carina; (I) anterior view; ( J) basal view; (K–N) Saurornitholestes (LSUMNS 746:6281); (K) labial view; (L) lingual view; (M) anterior view; (N) basal view; (O–S) Saurornitholestes (LSUMNS 746:17754); (O) labial view; (P) lingual view; (Q) denticles on posterior carina; (R) anterior view; (S) basal view; (T–X) Saurornitholestes sp. (LSUMNS 746:8327); (T) labial view; (U) lingual view; (V) denticles on posterior carina; (W) anterior view; (X) basal view; (Y–A2) Saurornitholestes sp. with flat side (LSUMNS 746:8388); (Y) labial view; (Z) lingual view; (A1) denticles on posterior carina; (A2) basal view.

present in this expanded collection, and no further descrip-

teeth in the bonebeds, these bones may be from the same

tions are necessary.

theropod.

Dromaeosauridae—indet.

Discussion

(Fig. 34.5) Referred Specimens. Metacarpal I (746:17756; Fig. 34.5 Y–A1). Pedal phalange (726:6212; Fig. 34.5Q–T). Description. A metacarpal and phalange are referred to the Dromaeosauridae (Longrich pers. com. 2007). Discussion. Although Lehman (1985) mentions the presence of theropod postcranials in Big Bend, this is the first paper that illustrates them. Based on the abundance of Saurornitholestes

532 sankey

Deciphering the taphonomic history behind the Rattlesnake Mt. bonebeds is important in order to understand if the assemblage represents a local paleocommunity (i.e., animals that lived in or near the same paleoenvironment) or a mixture from various paleocommunities. The key observations about the bonebeds are:

1. There are no complete or associated vertebrate

1999). However, this may reflect a preservational bias in the

skeletons; all of the teeth and bones are isolated.

fossil record. Soil conditions are key for eggshell preservation

2. This is a mixed assemblage, with both aquatic and

because acidic soils leach away the calcium carbonate of egg-

terrestrial vertebrates. However, the majority of the

shells. Paleosols with high carbonate content, such as from

fossils are from aquatic vertebrates, for example,

carbonate nodules, leads to preservation of eggshells (Car-

gars, crocodylians, and trionychid turtles. Less

penter 1982, 1999). The carbonate from clam and gastropod

common are terrestrial vertebrates such as lizards,

shells can also provide an important buffering agent in acidic

dinosaurs, and mammals.

soils. For example, from the fossil-rich Dinosaur Park For-

3. All of the dinosaur eggshells are small fragments, less

mation of Alberta, only two eggshell sites have been found. However, both sites have numerous clam shell fragments, in-

than 10 mm in diameter. 4. Eggshells are from a variety of dinosaurs, both ornithischians and theropods. 5. Many of the small dinosaur teeth are from

dicating that the carbonate from the clam shells buffered the acidic sediment conditions and prevented leaching of the eggshells (Brinkman 1986; Brinkman et al. 1987; Tanke and Brett-

hatchlings; hadrosaurs are the most abundant.

Surman 2001). Clearly, more dinosaur eggshell sites would

6. The dinosaur teeth and bones are from a variety of

have been preserved under the right conditions. The abun-

dinosaurs, both ornithischians and theropods.

dance of carbonate soil nodules and snails in the Rattlesnake

7. All the tyrannosaurid teeth are small, shed teeth, and most are probably from juveniles. 8. Many of the dinosaur teeth are remarkably unworn. For example the denticles on many of the Saurornitholestes teeth are still sharp and unabraded. 9. Vertebrate coprolites are abundant. Their size, shape,

Mt. bonebeds allowed preservation of the eggshells. Interestingly, all of the snails in the bonebed are missing their shells, indicating dissolution of their shells within acidic sediment conditions. How did the theropod teeth enter the site? Were the theropods predating or scavenging the nesting site(s)? It is interest-

and absence of bone material, match coprolites

ing that all of the tyrannosaurid teeth are from juveniles. Per-

identified as crocodylian (Schwimmer 2006).

haps the diet of juvenile tyrannosaurids differed from that of

10. The bonebeds are within muddy sandstones, with

adult tyrannosaurids. Dietary differentiation between juve-

abundant small, carbonized plant fragments, clay

niles and adults is often seen in modern animals. Additionally,

balls, and snail steinkerns.

track sites composed entirely of juvenile dinosaur tracks is

11. All of the snail steinkerns in the bonebed are missing

evidence that juvenile dinosaurs stayed together, separate from adults (Carpenter 1999).

their shells. 12. Several large (2 m long) pieces of coalified wood are

The abundant vertebrate coprolites in the bonebeds may be

present; all are poorly fossilized and have a thick

a clue to predation and/or scavenging. The size and shape of

coating of gypsum.

the coprolites, and the absence of bone material in them, are similar to those identified as crocodylian (Schwimmer 2006).

The presence of isolated and fragmentary bones suggest

A microvertebrate site that has been attributed to predation

that many were possibly exposed, reworked and/or trans-

is from Horseshoe Canyon Formation (Maastrichtian) of Al-

ported before burial. The dinosaur eggshells and baby dino-

berta (Ryan et al. 1998). The site is within over-bank deposits

saur teeth and bones indicate that bone and eggshell from at

that formed during a flood event. In this site, 66% of the ele-

least one nesting site containing hatchlings were reworked

ments are hadrosaurs, with 10% from babies, 17% of the ele-

and deposited here. The presence of complete (i.e., not frag-

ments (teeth) are from Troodon, and the remaining 11% are

mented) snails and the rarity of sharks and rays together indi-

from other theropods, ankylosaurs, and ceratopsians. The un-

cate that the environment of deposition was fresh-water or

usually high numbers of Troodon and hadrosaur is cited as

possibly brackish. The sediments and fossils further indicate

evidence for predation by Troodon on eggs or babies (Ryan et

that the likely agent of reworking and transport was a fast

al. 1998). The abundance of teeth from juvenile tyrannosau-

flowing alluvial channel. However, this interpretation does

rids and hatchling hadrosaurs in the Rattlesnake Mt. micro-

not preclude the possibility that these fossils experienced

sites supports the idea that juvenile tyrannosaurids preyed on

other taphonomic modifications prior to final burial.

the hatchling hadrosaurs.

How far away were the nesting sites? Well-preserved sites

Eberth and Currie (2005) describe the taphonomy of the

with dinosaur eggs, eggshells, and hatchling bones, such as

fossil assemblages in the late Campanian deposits of Dino-

nesting sites in Alberta and Montana, are usually found in

saur Park, Alberta. In multitaxic bonebeds, fossils are from

paleosols within more inland and better-drained parts of

many different species. In one example (BB 47), the fossils in-

floodplain (Zelenitsky et al. 1996; Carpenter 1999; pers. obs.

clude worn and rounded bones to complete turtle shells and a

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 533

crocodile skull. The Rattlesnake Mt. bonebeds are similar; they

5. Deposition of this slurry as a channel lag

contain a variety of taxa with a range of preservation quality.

6. Subsequent deposition of the overlying finer sands

For example, the rounded and worn bone fragments and the

during the final, receding flood stage

unworn dinosaur teeth, indicate different taphonomic histories for various taxa.

Eberth and Currie (2005) proposed that the over-riding for-

Another taphonomic mode Eberth and Currie (2005) iden-

mative and preservational influences on the rich fossil depos-

tify from Dinosaur Park is vertebrate microfossil assemblages,

its at Dinosaur Provincial Park were frequent and severe floods

which they describe as concentrations of well-sorted, small

across the coastal plain. They envisioned ‘‘Bangladesh-style

bones and teeth. Because the fossils are typically small and

flooding’’ events for the region. Secondarily, they cited preda-

resistant elements (e.g., teeth, scales, and dense bone), they

tion, scavenging, and trampling as important factors that

suggest that some of these deposits may have had long, com-

modified the carcasses and bones of vertebrates as they were

plex preburial histories (Eberth and Currie 2005), such as

about to enter the fossil record. This sort of flooding and post-

(1) accumulation and scattering on a floodplain, or (2) con-

mortem events are broadly similar to the scenario I envision

centrated within feces, or (3) previously buried, before further

for the Rattlesnake Mt. bonebeds.

reworking, transport, and final deposition within the microsite (Eberth 1990). However, because the taxonomic composition of the microsites changes with stratigraphic position and

Conclusions

the sites contain easily modified mudstone clasts, Eberth

This paper has provided new sedimentologic and paleontol-

(1990) argued that they were composed of local taxa. Brink-

ogic information from two closely associated high-diversity,

man et al. (2005) also concluded that microfossil assemblages

mixed bonebeds in the Big Bend area. One of these sites

in Dinosaur Park are locally derived, sampling local paleocom-

yielded the most complete skull of Agujaceratops mariscalensis

munities. This fits my interpretation of the Rattlesnake Mt.

and, in combination, all of the sites yield a rich assemblage of

microsites. The presence of fragile elements, such as the deli-

plants, invertebrates, and other vertebrates. This composite

cate, unworn theropod teeth and dinosaur eggshell, argues

vertebrate assemblage provides an important glimpse of local

against long transport or a complex taphonomic history, at

paleocommunity structure in this region, as well as an oppor-

least for some of the taxa.

tunity to make comparisons with other Campanian-age verte-

There are few taphonomic data currently available from the

brate faunas within North America. Although the Big Bend

Rattlesnake Mt. bonebeds Although the original associations,

Cretaceous vertebrate assemblage is less well known than ver-

distributions and orientations of the bones cannot be recon-

tebrate assemblages of comparable age from Montana and Al-

structed, I am confident that most if not all are disassociated

berta, it is an important southern datum.

and isolated elements (one exception is a partially associated partial salamander skeleton, approximately 20% complete).

Acknowledgments

This preservation pattern is compatible with the interpreta-

I appreciate financial support from South Dakota School of

tion of the deposits as channel lags. Lastly, the range of fossil

Mines and Technology (1999–2002) and California State Uni-

preservation (whole to fragmentary; pristine to worn) suggest

versity, Stanislaus (2003–present). Fossils were collected un-

variable taphonomic histories prior to burial.

der research permit BIBE-2007-SCI-0001. Thanks to D. Corrick

Given the data and interpretations presented here, a likely

and V. Davilla (Science and Natural Resources Division) for

scenario as to how the Rattlesnake Mt. bonebeds formed is as

their logistical support. I thank student assistants (S. Gas-

follows:

away, G. Knauss, S. MacInnes, V. Meredith, E. Nona, N. Ortiz, R. Peltier, N. Polan, T. Pranger, D. Tovar, M. Wedel, and

1. Death of the ceratopsian A. mariscalensis and nearby hadrosaur hatchlings 2. Scavenging and scattering of the ceratopsian carcass

E. Welsh). Thanks to W. and M. Clark, who have accompanied me to Big Bend for the past 7 years. I also thank LSUMNS staff ( J. Schiebout and S. Ting) for their help with these collections.

and nearby hadrosaur hatchlings by young

L. Pond (Louisiana Geological Survey, Louisiana State Univer-

tyrannosaurids and small theropods

sity) made the figures and compiled photographic plates. I

3. A severe flood event that washed over the floodplain,

appreciate identifications by and discussions with D. Brink-

sweeping the remains of the A. mariscalensis

man (Royal Tyrrell Museum), P. Currie (University of Alberta),

skeleton(s), the theropods, and the nesting sites with

S. Hope (California Academy of Science), N. Longrich (Uni-

hatchling hadrosaurs and eggshells into the channel

versity of Calgary), and S. Sampson (University of Utah).

of a nearby and large river

W. Langston (University of Texas) made copies of field photo-

4. Mixing of elements in a slurry of sand, mud, snails, and small plant fragments

534 sankey

graphs of A. mariscalensis from Rattlesnake Mountain. Thanks to D. Brinkman for broad access to the Tyrrell Museum. I espe-

cially appreciate the assistance of geologists at Baylor University (Drs. S. Atchley, L. Nordt, S. Dworkin, and S. Driese) with stratigraphic correlations and paleosol interpretation. Thanks to the organizers and editors of the horned dinosaur symposium and volume. Thorough reviews by Dave Eberth and Don Brinkman improved the paper considerably. References Cited Abler, W. L. 1997. Tooth serrations in carnivorous dinosaurs. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, pp. 740–741. San Diego: Academic Press. Atchley, S. C., L. C., Nordt, and S. I. Dworkin. 2004. Eustatic control on alluvial sequence stratigraphy: A possible example from the Cretaceous-Tertiary transition of the Tornillo Basin, Big Bend National Park, West Texas, U.S.A. Journal of Sedimentary Research 74: 391–404. Baszio, S. 1997. Investigations on Canadian dinosaurs: Systematic palaeontology of isolated dinosaur teeth from the Latest Cretaceous of south Alberta, Canada. Courier Forschungsinstitut Senckenberg 196: 33–77. Brinkman, D. B. 1986. Microvertebrate sites: Progress and prospects. In B. G. Naylor, ed., Dinosaur Systematics Symposium, Field Trip Guidebook to Dinosaur Provincial Park, pp. 24–37. Drumheller: Tyrrell Museum of Palaeontology. ———. 1990. Paleoecology of the Judith River Formation (Campanian) of Dinosaur Provincial Park, Alberta, Canada: Evidence from vertebrate microfossil localities. Palaeogeography, Palaeoclimatology, Palaeoecology 78: 37–54. Brinkman, D. B., D. A. Eberth, and P. J. Currie. 2007. From bonebeds to paleobiology: Applications of bonebed data. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 221–263. Chicago: University of Chicago Press. Brinkman, D. B., D. A. Eberth, and P. A. Johnston. 1987. Bonebed 31: Palaeocology of the upper Cretaceous Judith River Formation at Dinosaur Provincial Park, Alberta, Canada. In D. A. Eberth, ed., Fourth Symposium on Mesozoic Terrestrial Ecosystems, Field Trip ‘‘A’’ Guidebook, pp. 12–13. Royal Tyrrell Museum of Paleontology, Occasional Paper 3. Brinkman, D. B., A. P. Russell, and J.-H. Peng. 2005. Vertebrate microfossil sites and their contribution to studies of paleoecology. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 88–98. Bloomington: Indiana University Press. Carpenter, K. 1982. Baby dinosaurs from the Late Cretaceous Lance and Hell Creek formations and a description of a new species of theropod. Contributions to Geology, University of Wyoming 20: 123–134. ———. 1990. Ankylosaur systematics: Example using Panoplosaurus and Edmontonia (Ankylosauria: Nodosauridae). In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 281–298. Cambridge: Cambridge University Press. ———. 1997. Ankylosaurs. In J. O. Farlow and M. K. Brett-Surman, eds., The Complete Dinosaur, pp. 307–316. Bloomington: Indiana University Press.

———. 1999. Eggs, Nests, and Baby Dinosaurs: A Look at Dinosaur Reproduction. Bloomington: Indiana University Press. Cifelli, R. L. 1995. Therian mammals of the Terlingua local fauna ( Judithian), Aguja Formation, Big Bend of the Rio Grande, Texas. Contributions to Geology, University of Wyoming 30: 117– 136. Coombs, W. P., Jr. 1978. The families of the ornithischian dinosaur Order Ankylosauria. Palaeontology 21: 143–170. ———. 1990. Teeth and taxonomy in ankylosaurs. In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives, pp. 269–279. Cambridge: Cambridge University Press. Cope, E. D. 1869. [Remarks on Eschrichtius polyporous, Hypsibema crassicauda, Hadrosaurus tripos, and Polydectes biturgidus]. Proceedings of Natural Academy of Science, Philadelphia 21: 192. Currie, P. J., J. K. Rigby Jr., and R. E. Sloan. 1990. Theropod teeth from the Judith River Formation of southern Alberta, Canada. In K. Carpenter and P. J. Currie, eds., Dinosaur Systematics: Approaches and Perspectives. Cambridge: Cambridge University Press. Davies, K., and T. M. Lehman. 1989. The WPA Quarries. In A. B. Busbey III and T. M. Lehman, eds., Vertebrate Paleontology, Biostratigraphy, and Depositional Environments, Latest Cretaceous and Tertiary, Big Bend Area, Texas; Guidebook, Field Trips Nos. 1a, b, c, pp. 32–42. Society of Vertebrate Paleontology 49th Annual Meeting, Austin, Texas. Eberth, D. A. 1990. Stratigraphy and sedimentology of vertebrate microfossil sites in the uppermost Judith River Formation (Campanian), Dinosaur Provincial Park, Alberta, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology 78: 1–36. Eberth, D. A., and P. J. Currie. 2005. Vertebrate taphonomy and taphonomic modes. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 453–477. Bloomington: Indiana University Press. Eberth, D. A., M. Shannon, and B. G. Noland. 2007. A bonebeds database: Classification, biases, and patterns of occurrence. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis, and Paleobiological Significance, pp. 103–219. Chicago: University of Chicago Press. Forster, C. A., P. C. Sereno, T. W. Evans, and T. Rowe. 1993. A complete skull of Chasmosaurus (Dinosauria: Ceratopsidae) from the Aguja Formation (late Campanian) of west Texas. Journal of Vertebrate Paleontology 13: 161–170. Gasaway, S., J. T. Sankey, N. Ortiz, and V. Meredith. 2007. Paleoecology of a Chasmosaurus mariscalensis bonebed, Late Cretaceous (late Campanian), Big Bend National Park, Texas. Journal of Vertebrate Paleontology 27(3, Suppl.): 79A. Geological Society of America. 1991. Rock-Color Chart. Boulder: The Geological Society of America. Gilmore, C. W. 1930. On dinosaurian reptiles from the Two Medicine Formation of Montana. Proceedings of the United States National Museum 77: 1–39. Holtz, T. R., Jr. 2004. Tyrannosauroidea. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 111–136. Berkeley: University of California Press. Horner, J. R., and P. J. Currie. 1994. Embryonic and neonatal

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 535

morphology and ontogeny of a new species of Hypacrosaurus (Ornithischia, Lambeosauridae) from Montana and Alberta. In K. Carpenter, K. F. Hirsch, and J. R. Horner, eds., Dinosaur Eggs and Babies, pp. 312–336. Cambridge: Cambridge University Press. Lambe, L. M. 1902. On vertebrata of the mid-Cretaceous of the Northwest Territory. 2. New genera and species from the Belly river Series (mid-Cretaceous). Contributions to Canadian Palaeontology 3: 25–81. Lehman, T. M. 1982. A ceratopsian bone bed from the Aguja Formation (Upper Cretaceous), Big Bend National Park, Texas. M.A. thesis, University of Texas, Austin. ———. 1985. Stratigraphy, sedimentology, and paleontology of Upper Cretaceous (Campanian-Maastrichtian) sedimentary rocks in Trans-Pecos, Texas. Ph.D. diss., University of Texas, Austin. ———. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. ———. 1997. Late Campanian dinosaur biogeography in the Western Interior of North America. In D. A. Wolberg, E. Stump, and G. D. Rosenberg, eds., Dinofest International: Proceedings of a Symposium Held at Arizona State University, pp. 223–240. Philadelphia: Academy of Natural Sciences. ———. 2007. Growth and population age structure in the Horned Dinosaur Chasmosaurus. In K. Carpenter, ed., Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs, pp. 259–317. Bloomington: Indiana University Press. Lehman, T. M., F. W. McDowell, and J. N. Connelly. 2006. First Isotopic (U-Pb) Age for the Late Cretaceous Alamosaurus Vertebrate Fauna of West Texas, and its Significance as a Link Between Two Faunal Provinces. Journal of Vertebrate Paleontology 26: 922–928. Lucas, S. G., R. M. Sullivan, and A. P. Hunt. 2006. Reevaluation of Pentaceratops and Chasmosaurus (Ornithischia: Ceratopsiadae) in the upper Cretaceous of the Western Interior. In S. G. Lucas and R. M. Sullivan, eds., Late Cretaceous Vertebrates from the Western Interior, pp. 367–370. New Mexico Museum of Natural History and Science Bulletin 35. Marsh, O. C. 1881a. Principal characters of American Jurassic dinosaurs. Part V. American Journal of Science 3: 417–423. ———. 1881b. Principal characters of American Jurassic dinosaurs. Part IV. Spinal cord, pelvis and limbs of Stegosaurus. American Journal of Science 3: 167–170. ———. 1890a. Additional characters of the Ceratopsidae, with notice of new Cretaceous dinosaurs. American Journal of Science 3: 418–426. ———. 1890b. Description of new dinosaurian reptiles. American Journal of Science 3: 81–86. Mathew, W. D., and B. Brown. 1922. The family Deinodontidae, with notice of a new genus from the Cretaceous of Alberta. Bulletin of the American Museum of Natural History 46: 367–385. McCrea, R. T., M. G. Lockley, and C. A. Meyer. 2001. Global distribution of purported ankylosaur track occurrences. In K. Carpenter, ed., The Armored Dinosaurs, pp. 413–454. Bloomington: Indiana University Press.

536 sankey

Mikhailov, K. E. 1997. Fossil and recent eggshell in amniotic vertebrates: Fine structure, comparative morphology, and classification. Special Papers in Paleontology 56: 1–80. Nordt, L., S. Atchley, and S. Dworkin. 2003. Terrestrial evidence for two greenhouse events in the latest Cretaceous. Geological Society of America Today 13: 4–9. Nydam, R. L., J. G. Eaton, and J. T. Sankey. 2007. New taxa of transversely-toothed lizards (Squamata: Scincomorpha) and new information on the evolutionary history of ‘‘Teiids.’’ Journal of Paleontology 81: 538–549. Osborn, H. F. 1905. Tyrannosaurus and other Cretaceous carnivorous dinosaurs. Bulletin of the American Museum of Natural History 21: 259–265. ———. 1923. Two Lower Cretaceous dinosaurs of Mongolia. American Museum Novitates 95: 1–10. Patzkowsky, M. E., L. H. Smith, P. J. Markwick, C. J. Engberts, and E. D. Gyllenhaal. 1991. Application of the Fujita-Ziegler paleoclimatic model: Early Permian and Late Cretaceous examples. Palaeogeography, Palaeoclimatology, Palaeoecology 86: 67–85. Peng, J., A. P. Russell, and D. B. Brinkman. 2001. Vertebrate Microsite Assemblages (Exclusive of Mammals) from the Foremost and Oldman Formation of the Judith River Group (Campanian) of Southeastern Alberta: An Illustrated Guide. Provincial Museum of Alberta, Natural History Occasional Paper No. 25. Rowe, T., R. L. Cifelli, T. M. Lehman, and A. Weil. 1992. The Campanian Terlingua local fauna, with a summary of other vertebrates from the Aguja Formation, Trans-Pecos, Texas. Journal of Vertebrate Paleontology 12: 472–493. Ryan, M. J., P. J. Currie, J. D. Gardener, M. K. Vickaryous, and J. M. LaVigne. 1998. Baby hadrosaurid material associated with an unusually high abundance of Troodon teeth from the Horseshoe Canyon Formation, Upper Cretaceous, Alberta, Canada. Gaia 15: 123–133. Samman, T., G. L. Powell, P. J. Currie, and L. V. Hills. 2005. Morphology of the teeth of western North American tyrannosaurids and its applicability to quantitative classification. Acta Palaeontologica Polonica 50: 757–776. Sampson, S. D., and M. A. Loewen. 2010. Unraveling a radiation: A review of the diversity, stratigraphic distribution, biogeography, and evolution of horned dinosaurs (Ornithischia: Ceratopsidae). In M. J. Ryan, B. J. Chinnery-Allgeier, and D. A. Eberth, eds., New Perspectives on Horned Dinosaurs: The Royal Tyrrell Museum Ceratopsian Symposium, pp. 405—427. Bloomington: Indiana University Press. Sankey, J. T. 1998. Vertebrate paleontology and magnetostratigraphy of the upper Aguja Formation (late Campanian), Talley Mountain area, Big Bend National Park, Texas. Ph.D. diss., Louisiana State University, Baton Rouge. ———. 2001. Late Campanian southern dinosaurs, Aguja Formation, Big Bend, Texas. Journal of Paleontology 75: 208–215. ———. 2005. Late Cretaceous vertebrate paleoecology Big Bend National Park, Texas. In D. R. Braman, F. Therrien, E. B. Koppelhus, and W. Taylor, eds., Dinosaur Park Symposium: Short Papers, Abstracts, and Program, pp. 98–106. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2006. Turtles of the upper Aguja Formation (late Campa-

nian), Big Bend National Park, Texas. In S. Lucas and R. Sullivan, eds., Late Cretaceous Vertebrates from the Western Interior, p. 235–243. New Mexico Museum of Natural History and Science Bulletin 35. ———. 2008. Vertebrate paleoecology from microsites, Talley Mountain, upper Aguja Formation (Late Cretaceous), Big Bend National Park, Texas, USA. In J. T. Sankey and S. Baszio, eds., The Unique Role of Vertebrate Microfossil Assemblages in Paleoecology and Paleobiogeography, pp. 61–77. Bloomington: Indiana University Press. Sankey, J. T., S. Atchley, L. Nordt, S. Dworkin, and S. Driese. 2007a. Vertebrates and paleoclimate from a Chasmosaurus mariscalensis bonebed, Late Cretaceous (late Campanian), Big Bend National Park, Texas. In D. R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 134–139. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2007b. Dinosaurs and dirt: Dinosaur paleoecology, paleosol stratigraphy, and isotope geochemistry from the upper Aguja Formation (Late Cretaceous: late Campanian–early Maastrichtian), Big Bend National Park, Texas. Journal of Vertebrate Paleontology 27(3, Suppl.): 140A. Sankey, J. T., D. B. Brinkman, R. C. Fox, and D. A. Eberth. 2005a. Patterns of distribution of mammals in the Dinosaur Park Formation and their paleobiological significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 436–449. Bloomington: Indiana University Press. Sankey, J. T., D. B. Brinkman, M. Guenther, and P. J. Currie. 2002. Small theropod and bird teeth from the Judith River Group (late Campanian), Alberta. Journal of Paleontology 76: 751–763. Sankey, J. T., and W. A. Gose. 2001. Late Cretaceous mammals and magnetostratigraphy, Big Bend, Texas. Occasional Papers of the Museum of Natural Science, Louisiana State University 77: 1–16. Sankey, J. T., B. R. Standhardt, and J. A. Schiebout. 2005b. Theropod teeth from the Upper Cretaceous (CampanianMaastrichtian), Big Bend National Park, Texas. In K. Carpenter, ed., Carnivorous Dinosaurs, pp. 127–152. Bloomington: Indiana University Press. Schiebout, J. A. 1997. Microvertebrate sites. In P. J. Currie and K. Padian, eds., Encyclopedia of Dinosaurs, pp. 437–442. San Diego: Academic Press. Schiebout, J. A., C. A. Rigsby, S. D. Rapp, J. A. Hartnell, and B. R. Standhardt. 1987. Stratigraphy of the Cretaceous-Tertiary and Paleocene-Eocene transition rocks of Big Bend National Park, Texas. Journal of Geology 95: 359–375. Schiebout, J. A., S. Ting, and J. T. Sankey. 1998. Microvertebrate concentrations in pedogenic nodule conglomerates: Recognizing the rocks and recovering and interpreting the fossils. Palaeontologia Electronica 1: 1–54. Schwimmer, D. R. 2006. King of the Crocodylians: The Paleobiology of Deinosuchus. Bloomington: Indiana University Press.

Standhardt, B. R. 1986. Vertebrate paleontology of the Cretaceous/Tertiary transition of Big Bend National Park, Texas. Ph.D. diss., Louisiana State University, Baton Rouge. ———. 1989. Dawson Creek: Late Cretaceous and Paleocene vertebrates of Big Bend National Park. In A. B. Busbey III and T. M. Lehman, eds., Vertebrate Paleontology, Biostratigraphy, and Depositional Environments, Latest Cretaceous and Tertiary, Big Bend Area, Texas. Guidebook, Field Trips Nos. 1a, b, c., pp. 27–29. Society of Vertebrate Paleontology 49th Annual Meeting, Austin, Texas. Sternberg, C. M. 1928. A new armored dinosaur from the Edmonton Formation of Alberta. Transactions of the Royal Society of Canada 3: 93–106. Sues, H.-D. 1978. A new small theropod dinosaur from the Judith River Formation (Campanian) of Alberta, Canada. Zoological Journal Linnaean Society London 62: 381–400. Sullivan, R. D., and S. G. Lucas. 2006. The Kirtlandian LandVertebrate ‘‘Age’’: Faunal composition, temporal position and biostratigraphic correlation in the nonmarine upper Cretaceous of western North America. In S. G. Lucas and R. M. Sullivan, eds., Late Cretaceous Vertebrates from the Western Interior, pp. 7–29. New Mexico Museum of Natural History Bulletin 35. Tanke, D. H., and M. K. Brett-Surman. 2001. Evidence of hatchling-size hadrosaurs (Reptilia: Ornithischia) from Dinosaur Provincial Park (Dinosaur Park Formation: Campanian), Alberta. In D. H. Tanke and K. Carpenter, eds., Mesozoic Vertebrate Life: New Research Inspired by the Paleontology of Philip J. Currie, pp. 206–218. Bloomington: Indiana University Press. Tomlinson, S. L. 1997. Late Cretaceous and Early Tertiary turtles from the Big Bend region, Brewster County, Texas. Ph.D. diss., Texas Tech University, Lubbock. Vickaryous, M. K., T. Maryanska, ´ and D. B. Weishampel. 2004. Ankylosauria. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 363–392. Berkeley: University of California Press. Weishampel, D. B., and J. R. Horner. 1990. Hadrosauridae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, pp. 534–561. Berkeley: University of California Press. Welsh, E. 2005. Eggshells and baby dinosaurs in the upper Aguja Formation of Big Bend National Park, Texas. Journal of Vertebrate Paleontology 25(3, Suppl.): 128A. Welsh, E., and J. T. Sankey. 2008. First dinosaur eggshells from Texas, USA: Aguja Formation (late Campanian), Big Bend National Park. In J. T. Sankey and S. Baszio, eds., The Unique Role of Vertebrate Microfossil Assemblages in Paleoecology and Paleobiogeography, pp. 166–177. Bloomington: Indiana University Press. Zelenitsky, D. K., L. V. Hills, and P. J. Currie. 1996. Parataxonomic classification of ornithoid eggshell fragments from the Oldman Formation ( Judith River Group; upper Cretaceous), southern Alberta. Canadian Journal of Earth Sciences 33: 1665– 1667.

Significance of High-Diversity, Mixed Bonebeds Containing Agujaceratops mariscalensis 537

PART FIVE HISTORY OF HORNED DINOSAUR COLLECTION

35 Lost in Plain Sight: Rediscovery of William E. Cutler’s Missing Eoceratops D A R R E N H . TA N K E

during the fall and winter of 1919–1920 and sum-

BMNH: collections acronym for the Natural History

mer of 1920, William E. Cutler, despite ill health and

Museum, London; CMN: Canadian Museum of Nature,

poor weather, succeeded in uncovering and collecting a

Ottawa; DPP: Dinosaur Provincial Park, Alberta; GSC:

partial ceratopsian skeleton from quarry 78 in Dinosaur

Geological Survey of Canada, Ottawa; NHM: Natural His-

Provincial Park, Alberta. At the time, Cutler was not affil-

tory Museum (formerly the British Museum Natural His-

iated with any professional institution, and the specimen

tory [BMNH], London); RTMP: Royal Tyrrell Museum of

was put into storage in Calgary awaiting a buyer. None

Palaeontology, Drumheller; TMP: collections acronym

was immediately found, and the skeleton remained in

for the Royal Tyrrell Museum.

Calgary for several years. Cutler died in Africa during fieldwork in 1925; the subsequent whereabouts of his

[The author’s comments are contained within square brackets.]

ceratopsian became unknown and the specimen was seemingly lost. Since then a number of authors have tried to relocate the specimen; all have failed, due mainly

Background and Collection History

to the confusing state of affairs related to the closing of

William Edmund Cutler (Fig. 35.1) was an independent

a museum Cutler was involved with and improper dis-

Alberta-based fossil collector whose paleontological activities

persal of its fossil collection. This article explores the tan-

in the province are incompletely known. Born in 1878, Cutler

gled history and successful 2005 relocation of the long-

was apparently a rather eccentric character and was difficult

lost skeleton. The implications of this relocation are also

to get along with. He first moved to Alberta from England

considered. Institutional Abbreviations. AB: Alberta, Canada;

in the late 1800s, quite possibly as a remittance man.1 After ranching and farming with English friend Frederick S.

AMNH: American Museum of Natural History, New York;

Presant on Threehills Creek in the Carbon, AB district for a

1. Remittance men were typically young British males who were a problem to society or especially to their families. Often they were the second-born son, who would inherit little to nothing of the family estate. In an attempt to shoo them away from home and have them safely far away for good, wealthier families enticed the

young men to make a name for themselves in Canada or elsewhere in the colonies. The young men would be given a one-way ticket for a boat ride overseas and a sum of money to get started. To keep them away, a ‘‘remittance check’’ or allowance was sent to them at regular intervals.

541

number of years (Anonymous 1986; Graf 1986), around 1912

Currie–led field crew as a volunteer in 1979. A number of

he decided to try his hand at collecting dinosaur fossils for

authors (L. S. Russell 1966; Spalding 1999; Maier 2003) were

commercial gain. This decision may have been inspired by the

able to track the specimen as far as Calgary, possibly to the

intensive collecting activities of Barnum Brown (AMNH) and

Calgary Zoo, but no farther. Godfrey and Holmes (1995: 728–

Sternberg family (GSC) about 20 km east and 33 km southeast

729) suggested it was in Buenos Aires, without substantiation.

of Cutler’s ranch in the Drumheller area. His earliest collecting

It is known to have been shipped to and stored in Calgary, as

efforts appear to have been near Drumheller, in and around

Cutler himself writes:

the mouth of Kneehills Creek (Tarbuck 1972), and probably not far from where J. B. Tyrrell collected a partial Albertosaurus skull (CMN 5600) in 1884. April 1913 saw him collecting dinosaurs in DPP for the Calgary Syndicate for Prehistoric Research (a group of prominent and well-heeled Calgarian businessmen supporting Cutler’s fieldwork), and the now defunct Calgary Public Museum, an institution curating and display-

There is at present in Calgary, stored in some 14 boxes, a skeleton of Eo-Ceratops which fairly well represents one side of the dinosaur and is, with a good skull with the rear elements present, which were so far unknown, the best or only skeleton of this form. It will not be developed in Calgary but awaits purchase by some capable Institute. (Cutler 1922)

ing an eclectic assortment of human and natural history ar-

The same year, Cutler was anxiously awaiting an opportunity

tifacts, including dinosaur fossils (Ritchie 1934). Later that

to leave Alberta and dig dinosaurs in Africa on behalf of the

summer, Cutler joined the AMNH crew, gaining several

BMNH (Spalding 1999: 225). Cutler no doubt met with senior

months’ valuable field experience with them before he was

staff at the BMNH during leave from his military service and

asked to leave. In 1914, he was back in DPP, where he collected

had discussed this possibility. Cutler was hired by them a few

the type of ‘‘Scolosaurus cutleri.’’ Unfortunately, the specimen

years later, but with tragic results. In Tanzania, in the summer

collapsed on him during undercutting, resulting in serious

1925, Cutler, a workaholic who ignored sensible health pre-

upper body injuries (Tanke in prep.).

cautions and co-workers’ concerns for his health, died of ma-

During WWI, Cutler volunteered for military service in June

laria (Maier 2003). From this point, the whereabouts of his

1915 (Anonymous 1915) and went overseas. After his rather

Eoceratops was lost to history. Cutler’s will mentioned fossil

undistinguished military service (Maier 2003), Cutler returned

specimens located in London and Calgary:

to Canada on the Canadian Pacific passenger liner Empress of Britain, landing in St. John on April 2, 1919, and expected to arrive back in Calgary with a large number of other former soldiers on April 5th (Anonymous 1919a). He soon resumed his paleontological ways, returning to DPP early in the summer of 1919 (Anonymous 1919b). Over the fall and winter of 1919–1920, Cutler camped alone in the badlands in DPP (Tanke 2004), near the site of a disarticulated ceratopsian skeleton he had discovered. Despite serious illness (L. S. Russell

Another [British] law firm, Burnie and Coleman, announced that they possessed Cutler’s will and were acting on behalf of interested parties. The will referred to ‘‘scientific collections’’ in Canada and England. Cutler’s cabin trunks in storage at the museum [BMNH] had been opened and contained notebooks listing specimens stored in Calgary, presumably the scientific collection. Here the trail of letters ends; it is not clear how the situation was resolved. (Maier 2003: 182)

1966: 32) and the hardships of a cold Alberta winter, he ex-

The skeleton was first stored at Johnson Storage and Cartage

posed the skeleton, collected it in the summer of 1920 (Anony-

Co. Ltd. (Maier pers. com. 2003) who contacted Cutler’s ex-

mous 1920; Tanke 2007) and shipped the plaster field jackets

ecutors seeking payment for the storage of his specimen [and

to Calgary for storage while a buyer was sought. Cutler (1922:

possibly other specimens?]. Trafford (2005: 21) states that 90

24) and Maier (2003: 121) indicate 14 crates of plaster blocks

unclaimed crates of Cutler’s fossil finds were in Calgary after

were secured from this site. Cutler called his dinosaur find ‘‘Eo-

Cutler passed away but provides no reference. It was possible

Ceratops’’ (Anonymous 1920), a reference to the ceratopsid

that Cutler’s ceratopsian was later stored at the Calgary Public

genus Eoceratops (Lambe 1915), a form now considered a ju-

Museum,2 which first opened as a private venture in 1911

nior synonym of Chasmosaurus (Lehman 1989; Godfrey and

and occupied (starting in 1912) one floor of the new Memo-

Holmes 1995).

rial Park Library Building at 1221 2nd St. SW (Anonymous 2003a). The staff and board of directors hoped to develop

POST-COLLECTION HISTORY

this museum and/or its collection into a larger Provincial Museum in Calgary. Canadian vertebrate paleontologist,

The subsequent history of Cutler’s Eoceratops after its arrival

the late Dr. Loris Russell, got his start in museums in this

in Calgary is shrouded in mystery. Over the next 80+ years,

facility, working mainly with modern zoological specimens

the mystery continued to gain notoriety in the western Cana-

(Anonymous 2003b), and he may have been influenced by

dian paleontological community, and the author first became

Cutler to pursue a career in paleontology (Currie 1985). An-

aware of ‘‘Cutler’s lost dinosaur’’ when he joined a Philip

other possible influence was through the museum’s curator,

542 tanke

FIGURE 35.1. William Edmund Cutler (1878–1925). (A) Circa 1910 when he lived along Threehills Creek, 12.5 km NNW of Carbon, Alberta; (B) autographed portrait. Cutler’s appearance and clothing in (B) suggest that the picture was taken around 1920, during the time of the discovery and collection of his Eoceratops specimen. Images courtesy of the Royal Tyrrell Museum.

Robert Thurston,3 whose son James befriended Russell; both

was Dr. Euston Sisley (Anonymous 1940a), another colleague

shared a keen interest in vertebrate paleontology (Rowland

of Cutler and head of the Calgary Syndicate for Prehistoric

and Tanke 2007). On one floor of the Memorial Park Library

Research, an organization that, for a short time in 1913, sup-

was a public library, run by Calgary’s first librarian and a

ported Cutler’s work in DPP:

friend of Cutler, Alexander Calhoun, a man also interested in paleontology. On another floor was the Calgary Museum, dedicated mostly to modern natural history, but also containing a few fossils. Also affiliated with this museum venture

2. In the course of research for this paper, it became evident that the proper and official name for the first natural history museum in Calgary is uncertain. The author was able to find the following names for the same institution: Calgary Natural History Museum; Calgary Natural History Society; Calgary Museum; Calgary Public Museum; City of Calgary Museum and Art Gallery; and Civic Museum. It is not known which of these is correct. For the purposes of this paper, it is called the Calgary Public Museum.

In 1913 Cutler persuaded a group of business and professional men in Calgary to organize the Calgary Syndicate of Prehistoric Research for the purpose of supporting his fossil collecting in the Red Deer badlands. One of his finds

3. James (Robert) Thurston’s son James E. Thurston (1905–1932) was a good friend of vertebrate paleontologist Loris S. Russell. The pair worked together on Russell’s first dinosaur dig, a GSC excavation that yielded a good Edmontosaurus skeleton near Drumheller in 1923; see Russell (1986). In Alberta, J. E. Thurston worked with ROM (1922) and GSC crews (1923–1926; Tanke 2008: fig. 20).

Lost in Plain Sight 543

that year was the skeleton of a small duck-billed dinosaur, which was partly prepared and exhibited for some years in the original Calgary Museum. Not properly protected, it suffered much damage from the public, but was eventually obtained for the National Museum of Canada . . . (Loris Russell 1966: 31)4

making comments regarding the loss of artifacts from the Calgary Museum. Subsequent council records show no action was taken. However, as a follow-up, Wood (1957) records that in 1946 the director of the Allied Art Centre, Archie Key,6 and persons formerly associated with the museum deemed the collection not worth saving and nearly the entire collec-

The museum operated at the Memorial Park Library building

tion was burned. Among the destroyed items were dinosaur

for a time, then was moved to the basement of the courthouse

fossils, Wood notes:

building. In 1928, it moved to the North-West Travellers Building at 515 1st St. SE (Anonymous 2003c), but the Great Depression caused it to fall into financial difficulties and the doors closed in 1935. Hints of the museum’s early financial troubles and unfulfilled efforts to build a larger museum are

A stuffed buffalo, a stuffed wolf, some traditional costumes, the remnants of a butterfly collection and a few petrified dinosaur bones went up in flames in the incinerator at Coste House in 1946. (Wood 1957)

revealed in a letter dated April 24, 1921. The author of this let-

Someone involved in this sad state of affairs further com-

ter is unclear from the signature [ J. Dumas?], but it appears on

mented in Wood (1957) that he thought the museum’s collec-

Canadian Pacific Railway Company letterhead, and is from

tion was stored for a time at ‘‘St. George’s Island,’’ the location

the office of the Chief Commissioner in the Department of

of the Calgary Zoo, though Jameson (1965: 19), commenting

Colonization and Development. The letter states:

on the decline and closing of the museum, stated that ‘‘the artifacts, after a number of unsatisfactory storage places, even-

Dear Mr. Black,5

tually disappeared.’’ Many items were apparently stolen or Replying to your letter of the 8th.

lost. The following fantastic claim regarding the fossils was

I had previously heard from Mr. Cutler relative to the sale of our collection and had advised him that I was personally prepared to approve of the sale. I note your remarks with regard to the possibility of a Museum being provided at Calgary, but, as you know, we have lived on this hope for some considerable time and offered the Provincial Government a present of most of our collection without reaching any result. However, I will be quite prepared to abide by whatever you and the other members of the Syndicate think best to be done with the collection. Yours very truly, [Signature]

The defunct museum’s collection, now numbering some 7,500 natural and human history artifacts, was moved and stored in the basement of the Coste House at 2208 Amherst St. SW; for a brief time in 1936, some of these items were put on display there (Anonymous 2003d). The house was then taken over by Calgary Allied Arts Council for the Allied Art Centre. Subsequently there is some mystery as to where the

made in one magazine: It was at Drumheller that he [Dr. Omer Patrick, Drumheller coal mine owner and founder of the Calgary Zoo] became interested in the fossil and dinosaur deposits of the ‘‘Badlands.’’ There he had watched an American, Dr. Barnum Brown, dig up and ship, bone by bone, a brontosaurus, to the Smithsonian Institution in Washington, D.C., where the assembled skeleton received world acclaim as the largest known land-dwelling animal in the world. He realized that the people on whose earth the Brontosaurus had originally walked, had lost part of their history. Dr. Patrick had become good friends with Dr. C. M. Sternberg, curator of Canada’s National Museum who frequently came to Drumheller to ‘‘dig.’’ Unfortunately the impoverished National Museum could not even pay for transportation of the heavy unpolished bones to Ottawa and as a result some were just left in a basement room in Calgary’s old courthouse to collect dust. Dr. Patrick took them home, sorted and polished them and they became the start of the St. George’s [Calgary Zoo] fossil collection. (Anonymous 1967)

artifacts were stored, but most met an unfortunate end. Years

The suggestion by Wood (1957) is supported by comments in

later, Calgary alderman Grant MacEwan raised new business

Osakiwsky (1979) that a ‘‘massive pile of geological specimens

at a council meeting (City of Calgary, September 4, 1957),

and fossils . . . were presented to the Calgary Zoo,’’ and Nutt

4. This specimen was apparently partly prepared by Cutler. A letter penned in the field in DPP by Barnum Brown to William D. Matthew (both AMNH) and dated August 19, 1914, speaks of Cutler doing partial preparation on a small hadrosaur skeleton ‘‘last winter,’’ suggesting he resided in Calgary and was working on dinosaurs for Dr. Sisley at that time. In 2007, the author located the lost quarry that yielded this specimen; it is now numbered as quarry 252. The specimen is a juvenile Gryposaurus (CMN 8784).

5. Mr. Black is David E. Black, founder of what was then Calgary’s largest jewelry company, D. E. Black Jewelers. This later became Birks Jewelers. 6. Archie Key (1894–1989) was the editor at the Drumheller Mail newspaper from 1927 to sometime early in WWII. Beginning in the late 1920s, he was the town’s main advocate to build a fossil museum or park in Drumheller (Tanke 2008), but his efforts failed, largely because of the Great Depression and WWII. He was awarded an honorary doctorate from the University of Calgary in 1968.

544 tanke

(1974: 770), who wrote briefly about the Calgary Zoo Natural

His field notes (Sternberg 1937) again discuss the Cutler col-

History Park, stated, ‘‘The W.E. Cutler collection is exhibited

lection. While work that summer centered on the Lost River

at the park.’’ Finally, Spalding (1999: 105) adds that ‘‘When

badlands near Manyberries, AB (Tanke 2008), Sternberg did

the Calgary Museum closed, its dinosaurs ended up in the

make a trip to Calgary to see the Cutler fossils. He writes on

zoo.’’

June 17:

Were the ceratopsian and other dinosaur bones collected by Cutler among the incinerated group? Probably not; it would be impossible to incinerate something as large as a dinosaur skeleton embedded in non-flammable matrix and wrapped in its large plaster field jackets. In 1936, Charles M. Sternberg came to Alberta to engage in fieldwork at DPP. He made a quick trip to Calgary and visited

I decided to go to Calgary today. I want to see the Johnson Storage Co., with reference to the Cutler collection of fossils which I think they will present to the Survey. I took F. [Fred] Shindler [the camp cook and field assistant] along as I expect to need help in Calgary if I examine the Cutler collection. (Sternberg 1937)

the Calgary Zoo, where construction of a life-sized concrete Chasmosaurus was underway for the zoo’s Prehistoric Park.

The next day Sternberg writes:

There he made a disturbing discovery, recorded in his field notes: I called on Dr. [Omer] Patrick7 and he took me out to St. George’s Island. I looked at the dinosaur skeleton which the Zoological Society has been given from the old museum [Calgary Public Museum]. He wants us to ship the skeleton to Ottawa and mount it for them. It is wrapped in sections so I could not tell what condition it was in. They have all of the collection, which was in the Calgary Museum, thrown about in an old building. I fear what they had in the museum will soon be destroyed from lack of care. Dr. Patrick does not seem to care about the scientific value of the collection. (Sternberg 1936)

Clearly, the Calgary Zoo was too inexperienced or unwilling to provide the proper care for the dinosaur fossils. That summer, C. M. Sternberg and his son Ray carefully packed and boxed up the dinosaur skeleton, a juvenile Gryposaurus, and shipped it to Ottawa (Anonymous 1936). There it was repaired, further prepared and later described by Waldman (1969). Sternberg does not record anything about Cutler’s ceratopsian. In his historical overview of dinosaur collecting in western Canada, Loris Russell confirms that fossils from the Calgary Museum were later stored at the Calgary Zoo, and provides a clue as to the possible fate of Cutler’s ceratopsian. He writes:

I called on the Johnson Storage Company but was informed by the lady in the office that there would be no decision re: Cutler collection until Mr. H. Johnson returned in July. She promised to write me as soon as a decision is made. (Sternberg 1937)

The subsequent field notes provide no clues. However, C. M. Sternberg wrote a letter to Barnum Brown on December 21, 1937 stating he had examined the Cutler collection at the Calgary Zoo and found loose bones and teeth, but no ceratopsian skull (D. A. Russell 1966). With the onset of WWII, Sternberg did not return to Alberta until 1946 and, having done his part to try and resolve this mess, no doubt considered the matter closed. The comments of Wood (1957), L. S. Russell (1966), Osakiwsky (1979), and Spalding (1999) indicate that some of the fossil collection was at the Calgary Zoo. Russell’s comment about ’’residue’’ of the Cutler collection indicates it may have been divided up. The plaster blocks were probably too large and heavy to be incinerated at Coste House. But where were they? Baptie (1972: 7) provides some information on the Calgary Public Museum and its missing collection. She notes that the material was stored away and ‘‘became more and more neglected, with the specimens literally falling victims to dust,

It [ceratopsian skeleton] was stored for years in Calgary, and as far as I know was part of the residue of the Cutler collection acquired by the Calgary Zoological Society. (L. S. Russell 1966: 32)

moths, water damage, vandals, and theft.’’ The Calgary Zoo

Another tantalizing clue involves C. M. Sternberg in 1937.

plans were made to develop a Prehistoric Park,8 and that year

7. Omer H. Patrick (d. 1947, age 78) was founder and president of the Calgary Zoological Society from 1928 to 1944, and a prime supporter of the Calgary Zoo Prehistoric Park. He has ties to the early coal mining industry in the Drumheller, Alberta, district, having co-founded the Atlas #1 coal mine in Drumheller in 1917 and subsequently others in the area, including the Atlas Coal Mine in East Coulee, AB.

8. This park and most of its dinosaurs were torn down in the late 1970s/early 1980s. The only known survivor of this unfortunate purge was ‘‘Dinny,’’ the Apatosaurus, which was designated in 1987 as a historical landmark. The Calgary Zoo Prehistoric Park seen today was opened in 1983 (east section) and 1984 (west section). For a recent history of the original Prehistoric Park, see Debus (2006).

also played a role in this history, so it is appropriate to provide some historical information about the zoo at this point. The Calgary Zoological Society was incorporated in 1929. In 1932,

Lost in Plain Sight 545

construction began on the first life-sized concrete dinosaur

but were, ultimately, fruitless. After mulling over this prob-

model. The Prehistoric Park officially opened on August 28,

lem, the author decided to try a different approach.

1937 (Anonymous 2003e). This consisted of an extensive

Cutler’s fossil specimens are known to have been donated to,

treed park and pathways containing numerous life-sized con-

or purchased by only four institutions: the aforementioned

crete dinosaurs and other prehistoric life. This popular park

Calgary Public Museum; the Calgary Zoo; the University of

also included two long, walk-through display buildings con-

Manitoba in Winnipeg; and the Natural History Museum

taining a Corythosaurus skeleton9 and individual dinosaur

(London). The Calgary Public Museum and Calgary Zoo leads,

bones, some of the latter of which might have been part of the

although seemingly the most promising, have proved unpro-

Cutler Collection from the old City Museum. As further evi-

ductive, if not distracting. Fossils collected by Cutler and cura-

dence to this effect, the Royal Tyrrell Museum acquired 52

ted in Winnipeg consist only of Paleozoic invertebrates col-

dinosaur bones and other fossils from the Calgary Zoo in 1984

lected from southern Manitoba (Leith 1952; Elias 1983; Maier

(see TMP 84.121 series). A coronoid process (TMP 84.121.20)

2003). Could Cutler’s lost dinosaur have been sent to England?

was the only identifiable ceratopsian specimen among them,

This line of reasoning suggested that it might be profitable to

with the remainder consisting of various Late Cretaceous ver-

examine the NHMs collection to see if there are any cera-

tebrate and plant remains and Paleozoic invertebrates. Addi-

topsian dinosaur skeletons from Alberta. This quickly proved

tionally, in the early 1990s, the RTMP acquired from the Cal-

to be the line of inquiry that would resolve the mystery.

gary Zoo some of Cutler’s field crates bearing 1919 and 1920

In London, there is an incomplete skeleton referable to

and ‘‘Steveville, Alta’’ markings. The latter shows some of Cut-

Chasmosaurus from DPP, bearing catalogue number BMNH

ler’s collection did make it to the zoo and were stored there for

R4948. This specimen consists of articulated premaxillae and

a period of time.

rostrum, both orbital regions with articulated jugals, brain-

Newspaper clippings archived in the Local History Collec-

case, dentaries, predentary, edentulous maxillae, and both

tion at the Calgary Public Library also provide some clues.

squamosals. Godfrey and Holmes (1995) indicated that post-

Anonymous (1940b) relates how zoo director Dr. Omer H. Pat-

cranial remains were also present. In late May 2004, the au-

rick advocated building a second fossil display house at the

thor asked a colleague, William T. Blows, to have a closer look

Calgary Zoo at an estimated cost of $1,500, and that such a

at the acquisition history of this specimen. Blows conducted

building, according to Patrick, ‘‘could be immediately filled

searches of archival material in the Natural History Museum

with specimens from the ‘Cutler Collection’ which the Society

and found letters and documents from 1919 to 1921 and 1923

acquired some years ago.’’ This fossil house was indeed built,

(see Appendix 35.1) pertaining to the eventual sale of Cutler’s

and the author recalls seeing the dinosaur fossils numerous

Eoceratops to the NHM.

times as a boy growing up in Calgary. The Calgary Zoological

In March 2005, Blows and the author met at the Natural His-

Society also appears to have intended to resurrect the defunct

tory Museum in London to compare original field photo-

Calgary Public Museum and its holdings on the zoo grounds.

graphs of Cutler’s Eoceratops specimen with BMNH R4948.

O. H. Patrick even had $6,500 donated toward seeing this hap-

Whereas the photographs are not of the highest quality, and

pen. Plans fell through, however, and in a surprising move,

preparation of the specimen was crude by today’s standards—

Patrick and most of the zoo’s directorate resigned and the

plaster fills in or otherwise obscures several anatomical

money was returned (Anonymous 1944).

features—the cracks on a BMNH R4948 humerus roughly match those on an in situ humerus that had been photo-

RESOLUTION

graphed by Cutler in a site that was subsequently mapped as original quarry number 78 in Sternberg (1950) and the RTMP

Historically, all attempts to locate Cutler’s lost dinosaur had

collections data base (Currie and Russell 2005). Whereas the

involved (1) following up on Cutler’s personal history, includ-

matching cracks are not entirely conclusive, the paperwork

ing his associations with a variety of Calgary institutions,

dealing with the sale of the specimen is persuasive evidence: to

which may have housed the specimen for a time, and (2) fol-

our knowledge Cutler collected only one ceratopsian skeleton.

lowing the historical trails of those institutions and their collections. Both approaches yielded interesting historical data

Implications This research has revealed some interesting paleontological

9. This Corythosaurus skeleton was collected by C. H. Sternberg and crew in 1915 (field no. 1915-9). The skeleton was acquired by the Royal Tyrrell Museum from the Calgary Zoo on October 2, 1990, and now bears the catalogue number TMP 84.121.1. The specimen in the field is figured in Sternberg (1917: fig. 16). The quarry site is now numbered as quarry 243.

546 tanke

and historical facts. More of BMNH R 4948 is present than is suggested by the literature (historical and scientific), and much of the skeleton actually consists of measurable elements (Appendix 35.2). Thus, although the incomplete skeleton alone probably doesn’t warrant a detailed anatomical descrip-

tion, after repair it could serve to complement future studies

last remaining pieces of the Cutler fossil collection from the

on variation and biostratigraphy in the genus Chasmosaurus.

Calgary Zoo ended up at the RTMP in 1984 (see TMP 84.121

During preparation of his classic tome, ‘‘Revision of the Cer-

series). In 2008, additional material, some of which may have

atopsia’’ (Lull 1933), R. S. Lull wrote to numerous museums,

been collected by Cutler, was received by RTMP’s education

including the Calgary Public Museum [ July 2, 1932], to in-

department. Some specimens are still at the Calgary Zoo serv-

quire about major ceratopsian specimens in their care. The

ing an important educational role.

Calgary Public Museum was a tiny facility even by the standards of the day, so, although it was surprising to find Lull’s letter among archived materials,10 its presence clearly demon-

Conclusions

strates the thoroughness of his research. [There is no known

Cutler’s ‘‘lost’’ Eoceratops has never actually been lost at all,

reply from the Calgary Public Museum.] Unfortunately, it is

and one of Alberta’s more enduring paleontological mysteries

unknown whether or not Lull wrote to the NHM. However,

centers on a dinosaur fossil that has been lost in ‘‘plain sight’’

this seems likely given that it was widely known in the paleon-

for almost 90 years. Cutler’s Eoceratops was safely in England

tological circles of the day that Cutler and C. H. Sternberg had

2 years before its collector’s death.

sold Albertan dinosaur material to the NHM. However, Lull’s

Once the sale was completed, Cutler left Alberta for the last

1933 compilation includes no mention of Cutler’s Eoceratops.

time, passing through Saskatchewan and spending some time

How could he have overlooked the specimen in London? We

at the University of Manitoba in Winnipeg (Maier 2003). His

can only speculate that it was not prepared or that it was

departure from Alberta likely meant his contact with paleon-

otherwise unavailable for viewing. Had Lull acquired some

tological crews in the province was over, and he evidently

basic information about this ceratopsian from the NHM, it is

never told them that he had sold his Eoceratops skeleton to

likely that the mystery never would have existed.

the NHM.

Several scientific and historical changes are now required to

With his death, the myth of Cutler’s lost dinosaur was born,

the quarry lists given in Sternberg (1950) and Currie and Rus-

and for the next eight decades it became an intriguing mys-

sell (2005) as a result of this research. Both sources identify

tery that would, from time to time, preoccupy paleontologists

quarry 78 at Dinosaur Provincial Park as a Cutler site yielding

and historical researchers alike. In this context, the author

a Centrosaurus skull. It is now known that quarry 78 is indeed

has gained much satisfaction from solving Alberta’s longest-

a W. E. Cutler site, but that it yielded BMNH R4948, an in-

running and intriguing paleontological mystery.

complete Chasmosaurus skull and associated skeleton, rather than a Centrosaurus skull. However, the exact locality in DPP from which Cutler’s Centrosaurus skull (BMNH R4859) was collected is presently unknown, and the site now has ‘‘lost quarry’’ status. It appears that the rest of the Cutler dinosaur bone collection, in quantities and qualities now unknown, stayed with the Calgary Public Museum (material he collected on their behalf around 1913–1914). As previous research into this topic has shown (see above), after the museum closed, the fossil collection was improperly stored, and some of it was lost through damage, theft, or by design. Some of the stolen material could potentially still exist in private collections. The remaining fossils from the Calgary Public Museum collection went to the Calgary Zoo sometime prior to the summer of 1936, when C. M. Sternberg observed the material there. Sometime after the summer of 1937, the Cutler fossil crates held at Johnson Storage and Cartage for some 17 years must have also ended up at the Calgary Zoo. There, over the next

Acknowledgments

The author has benefited from discussions and collaborations with Gerhard Maier (Calgary, AB); Carol Stokes, archivist (City of Calgary); staff of the Natural History Museum (London, England); and the Glenbow Museum (Calgary, AB). The author is especially grateful to Rob Rondeau (Hardisty, AB) for making the 2005 trip to London, England, possible; William T. Blows for finding and copying the documents pertaining to the sale of Cutler’s Eoceratops and research assistance; and Patty Ralrick for research assistance, discussions, and reviewing the manuscript. Special thanks to David Eberth and Craig Scott (Royal Tyrrell Museum, Drumheller) for reviewing and editing the manuscript. Appendix 35.1. Report on the Cutler Documents in the Natural History Museum, London

Author. Dr. W. T. Blows.

four and a half decades, the fossils were likely consumed, in

Date of Museum Visit. May 25, 2004

part, by public exhibits and as educational hand samples. The

Report Written. May 29, 2004. Notes

10. Letter at the Glenbow Museum, Calgary. See 1932 history entry #6 on the author’s Alberta ceratopsian history CD-ROM included in this volume.

The documents viewed were letters sent to Dr. A. Smith Woodward, Keeper of Geology, British Museum (Natural History) in

Lost in Plain Sight 547

London, and documents related to shipment of boxed fossils,

being almost as broad as high whilst the supra orbital horn

all were dated 1919, 1920, or 1923. These documents are held

is crooked up and of considerable size, almost equaling in

in the General Library of the museum. The access codes for

length the prenasal instead of a rudiment or boss as is usual

this reference material follow: DF 100/65: letters/documents

amongst the Canadian Ceratopsia, almost in fact an approach

dated 1919; DF 100/66: letters/documents dated 1920; DF

to Triceratops.’’

100/105/6: letters/documents dated 1923. 3. Letter to Woodward, April 3rd 1920 from Cutler, 1. Letter to Woodward, Friday Nov. 14th 1919, from Cutler, PO Box 223, Calgary

‘‘I have during the last 14 days removed a hill 12 feet high

PO Box 223, Calgary

‘‘I am still working on the Ceratopsian of Eoceratops affinities and shall finish this month if I experience any good weather’’

by 16 feet long by 9 feet wide to examine some ceratopsian remains with temnospondylus centra which appear to

4. Letter to Woodward, April 5th 1921 from Cutler,

me unusual. Today I have made my level platform and have

PO Box 223, Calgary

outlined a beautiful sacrum of 5 centrae [sic], perfect with spines in beautiful preservation, also several vertebrae, fragment of femur with great trochanter and one metacarpal so far, the right dentary, a splendid bone was unfortunately almost destroyed by a heavy rock fall when the side of the cliff slid out, this was impossible to foresee and I hope that I shall yet find the skull as I have some 100 square feet to uncover. In the butte where I am now working (Ceratops ? butte) are quantities of leaf impressions in the friable sandstone of the many angiosperms and by taking neatly trimmed pedestals of

‘‘I can sell the collections of 1919–1920 of Dinosaurian remains, should however require $2,000—two thousand dollars to repay the financing etc. undertaken when prices were at their highest. This collection includes a partial Eo-Ceratops skeleton (so far as I know the only skeleton of this extant) with a finely preserved but disarticulated skull and a good skull of Euphocephalus [sic] tutus Lambe etc.’’ 5. Letter to Woodward, February 26th 1923 from Cutler, 4 St. Mary Place, Winnipeg

the rock they are to be collected and are quite diagnostic.’’

‘‘Would you please inform me whether you could interest the 2. Letter to Woodward, Dec. 31st 1919 from Cutler,

Trustees of the Museum to purchase my Eo-Ceratops skeleton

PO Box 223, Calgary

in some 12 or so medium sized boxes and amounting to per-

‘‘The winter commenced on October 20th four weeks before the usual date and whilst taking a last survey in the snow I discovered the Butte shown in picture 3 (untouched) seeing that a dentary showed on one side and at similar level across diameter were partial femur, associated vertebrae and spines, and other large bone which proved to be a complete sacrum. I

haps two tons (or less) in weight.’’ 6. Bill of Lading no. 264, April 19th 1923 to Dr. Smith Woodward, British Museum (Natural History), London, England, from Canadian Pacific Railway Co., 15 boxes of Fossil remains, from Montreal to London by Canadian Pacific SS Ltd.

took the initiative of removing some 75 cubic yards of rock alone, which is now complete and in pictures 9–13 inclusive are shown parts of the ceratops skeleton. I believe I am the first to take up a specimen in winter in these fields but as I was expecting to be at your service and in the employ of your museum, very soon, I undertook this work now. There are

7. Marine Certificate, dated April 30th 1923 insured by American Express Co., 1 box skull of Denasaur [sic] for 111 dollars shipped on board SS AQUITANIA from New York to London, UK. Signed by Despard and Co., Insurance Brokers, 6 Hanover Street, New York.

some various complete vertebrae, a distal end of femur, an perfect ditto [ femur] and tibia, also fibula ? [question mark is Cutler’s] splendid sacrum some as yet undiagnosed pelvic

8. Letter to Woodward, May 9th 1923 from Cutler, Winnipeg, Manitoba

bones, humerus, radius, ulna, coracoid, scapula ? [question mark is Cutler’s] sternum ? [question mark is Cutler’s] atlas, den-

‘‘To refer now to the most important matter in hand, the Eo-

tary right, two large spiked processes of the parietal frill (post

Ceratops skeleton comprised in 15 boxes, was dispatched on

occipital ?), [question mark is Cutler’s] supra-orbital and pre-

April 27th ultimo as denoted by the two bills of lading which

nasal and an area of circa four feet square yet to trace out.

I enclose. The date of receipt noted as April 19th owes its

Undoubtedly the specimen belongs to the genus Styraco-

discrepancy to the fact that as first they would not accept

saurus but both in the vertebral and skull elements I see great

shipment unless prepaid. I trust that these remains will reach

differences. For instance the vertebrae are temnospondylous,

you safely in due course and I hope that the material will

and the prenasal is very massive and barely 6 inches in length

please you.’’

548 tanke

9. Letter to Woodward, July 7th 1923 from Cutler, PO Box 241, Winnipeg, Manitoba

‘‘If the funds from my sale of Eo-Ceratops are sent by you so as to arrive here prior to 28th July, all will be well, otherwise I think that as my postbox will be closed during my absence in the west and forwarding registered mail is dubious it might be wiser to address said letter to the post office at Swift Current [Saskatchewan] there to be held pending my arrival. I hope by now you have received the 15 Eo-Ceratopsian boxes.’’ 10. Certificate of Insurance, May 31st 1923 for shipment in the name American Express Co. Lot number 39511 [no other useful information was given] End of Report Appendix 35.2

NHM R 4948 Chasmosaurus (from quarry 78, DPP) consists of the following skeletal elements. Cranial. Squamosals; both orbital regions with articulated jugals and orbital horncores; orbital horncores; jugal; articulated premaxillae and rostrum; mandibles (mostly restored in plaster); unidentified skull bone (?part of parietal); braincase; pterygoids; predentary; and edentulous maxillae. Postcranial. Fourteen dorsal vertebrae and extra processes; complete sacrum (two sections); proximal scapula fragment; coracoids; ?pubes; humeri; ulnae; radius; ischium; femora; tibia; numerous ribs and rib fragments; and hundreds of broken bone pieces in boxes. References Cited Anonymous. 1915. [W. E. Cutler attestation papers.] www.collec tionscanada.ca/02/020106—e.html ———. 1919a. More large parties of Alberta men returning. Calgary Daily Herald, April 2, p. 23. ———. 1919b. Cutler back. Brooks Bulletin, June 8, p. 1. ———. 1920. Horned dinosaur being dug out by W. E. Cutler— Calgary scientist is working on rich find he made in the Steveville District. Calgary Daily Herald, August 31, p. 15. ———. 1936. Fossil garden work at St. George’s Island is resumed. Calgary Herald, August 28, p. 14. ———. 1940a. Dr. Sisley dies, was authority on Alberta game. Calgary Herald, March 20, p. 11. ———. 1940b. Zoological groups re-elects slate—Calgary fossil garden unique Dr. Patrick declares in report. Calgary Herald, April 16, p. 18. ———. 1944. Dr. O. H. Patrick, Zoo directorate, resign in body—no reason given; $6,500 donation is returned. Calgary Herald, April 24. ———. 1967. How St. George’s got its dragons. My Golden West 2 (March–April): 16–17. ———. 1986. PRESANT, Frederick S. In Carbon—Our History Our Heritage, p. 722. Carbon: Carbon Historical Committee.

———. 2003a. Cornerstones: Memorial Park Library. http:// calgarypubliclibrary.com/calgary/historic— tours/corner/mpl.htm. ———. 2003b. Great ROM dinosaur hunters—Loris S. Russell. www.rom.on.ca/dinohunters/russell.html. ———. 2003c. Cornerstones: North-West Travelers Building. http:// calgarypubliclibrary.com/calgary/historic— tours/corner/nwtravel.htm. ———. 2003d. Cornerstones: Coste House. http:// calgarypubliclibrary.com/calgary/historic— tours/corner/cos.htm. ———. 2003e. Calgary Zoo History. www.calgaryzoo.ab.ca/general—zoo—info/zoo—history.shtml. Baptie, S. 1972. The case of the missing museum. Glenbow 5: 6–7. Currie, P. J. 1985. Letter to Mr. Henry F. Irwin (Edmonton), August 27, 1985. On file in library, Royal Tyrrell Museum of Palaeontology, Drumheller, Alberta. Currie, P. J., and D. A. Russell. 2005. The geographic and stratigraphic distribution of articulated and associated dinosaur remains. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed, pp. 537–569. Bloomington: Indiana University Press. Cutler, W. E. 1922. The badlands of Alberta. Canadian Illustrated Monthly ( January): 19–25, 50–51. Debus, A. A. 2006. Calgary’s prehistoric zoo. Prehistoric Times 80: 52–53, 56. Elias, R. J. 1983. Late Ordovician rugose corals of the Stony Mountain formation, southern Manitoba and its equivalents. Journal of Paleontology 57: 924–956. Godfrey, S. J., and R. Holmes. 1995. Cranial morphology and systematics of Chasmosaurus (Dinosauria: Ceratopsidae) from the Upper Cretaceous of western Canada. Journal of Vertebrate Paleontology 15: 726–742. Graf, G. 1986. CUTLER, William E. In Carbon—Our History Our Heritage, pp. 394–395. Carbon: Carbon Historical Committee. Jameson, S. S. 1965. A visit to Calgary’s new museum. Alberta Historical Review 13: 19–22. Lambe, L. M. 1915. On Eoceratops canadensis, gen. nov., with remarks on other genera of Cretaceous horned dinosaurs. Canada Geological Survey Museum Bulletin 12, Geological Series 24: 1– 49. Lehman, T. M. 1989. Chasmosaurus mariscalensis, sp. nov., a new ceratopsian dinosaur from Texas. Journal of Vertebrate Paleontology 9: 137–162. Leith, E. I. 1952. Schizocoralla from the Ordovician of Manitoba. Journal of Paleontology 26: 789–796. Lull, R. S. 1933. A revision of the Ceratopsia or horned dinosaurs. Peabody Museum of Natural History Bulletin 3: 1–175. Maier, G. 2003. African Dinosaurs Unearthed—The Tendaguru Expeditions. Bloomington: Indiana University Press. Nutt, R.G. 1974. Alberta badland trails. Lapidary Journal 28: 770, 772–774, 776, 792–796. Osakiwsky, I. 1979. City’s early art and museums have varied, troubled past. Calgary Herald, February 24, p. 16. Ritchie, C. 1934. Museum holds world of wonders. Calgary Daily Herald (magazine section), December 15, p. 23.

Lost in Plain Sight 549

Rowland, S. and D. H. Tanke. 2007. The career of James E. Thurston and the extinction of the professional field collector in North American vertebrate paleontology. Geological Society of America Abstracts with Program 39: 382. Russell, D. A. 1966. Notes on Oldman Formation, (Alberta) specimens. Unpublished National Museum of Canada notebook. Russell, L. S. 1966. Dinosaur hunting in western Canada. Royal Ontario Museum, Life Sciences Contribution 70: 1–37. ———. 1986. Exploring a great dinosaur graveyard. Rotunda 19: 20– 29. Spalding, D. 1999. Into the Dinosaurs’ Graveyard—Canadian Digs and Discoveries. Toronto: Doubleday Canada. Sternberg, C. H. 1917. Hunting Dinosaurs in the Badlands of the Red Deer River, Alberta, Canada. San Diego: Privately published. Sternberg, C. M. 1936. Field Notes. On file at the Royal Tyrrell Museum of Palaeontology, Drumheller, Alberta. ———. 1937. Field Notes. On file at Royal Tyrrell Museum of Palaeontology library, Drumheller, Alberta. ———. 1950. Steveville west of the 4th Meridian, with notes on fossil localities. Geological Survey of Canada, Map 969A. Scale 1:31,680. Tanke, D. H. 2004. Discovery of William E. Cutler’s winter 1919– 1920 fieldcamp, Dinosaur Provincial Park, Alberta. In H. Allen,

550 tanke

ed., Alberta Palaeontological Society, Eighth Annual Symposium, Abstracts Volume, pp. 58–60. Calgary: Mount Royal College. ———. 2007. Lost in plain sight: Discovery of William E. Cutler’s lost ‘‘Eoceratops.’’ In D.R. Braman, comp., Ceratopsian Symposium: Short Papers, Abstracts, and Programs, pp. 144–147. Drumheller: Royal Tyrrell Museum of Palaeontology. ———. 2008. Remember me: Harold D. R. Lowe (1886–1952)—a forgotten name in early Albertan vertebrate palaeontology history. Alberta Paleontological Society Bulletin 23: 4–34. Tarbuck, J. 1972. The Tarbuck story. In Memories, Yours and Mine: A History of Beveridge Lake, East View, Garrett, Hesketh, Humbolt, Kirby, Lenox, Marne, Webb School Districts, pp. 119–121. Hesketh: Hesketh Pope Lease Historical Society. Trafford, T. 2005. The Evolution of the Calgary Zoo. Calgary: Calgary Zoo. Waldman, M. 1969. On a new specimen of Kritosaurus notabilis (Lambe) (Ornithischia: Hadrosauridae) from the Upper Cretaceous of Alberta, Canada. Canadian Journal of Earth Sciences 6: 569–576. Wood, R. 1957. Exhibits incinerated—Investigation continues— Case of mystery museum solved: Its contents were burned in 1946. Calgary Herald, October 15, p. 15.

36 Historical Collecting Bias and the Fossil Record of Triceratops in Montana MARK B. GOODWIN AND JOHN R. HORNER

Triceratops is one of the most familiar and recognizable

Introduction

Late Cretaceous dinosaurs from North America due to its prominent postorbital and nasal horns and solid frill com-

By anyone’s measure, Triceratops, or ‘‘three-horned face,’’ is

posed of the parietal and paired squamosals. Complete,

one of the most familiar genera of Late Cretaceous dinosaurs

undistorted adult Triceratops skulls are found primarily in

since it was named by O. C. Marsh in 1889. Early expeditions

sandstones and siltstones, which contribute generously to

near the turn of the twentieth century by Barnum Brown,

the sedimentary composition of the Hell Creek Formation

John Bell Hatcher, the team of Charles H. Sternberg, and oth-

in Montana, North and South Dakota, the equivalent

ers resulted in the discovery and collection of dozens of adult

Lance and Evanston formations, Wyoming, the Laramie

skulls from Upper Cretaceous sediments of the Western Inte-

Formation of Colorado, and the Scollard and Frenchman

rior of North America, particularly from Montana and adja-

formations of Alberta and Saskatchewan, Canada. Disar-

cent states. Only a limited number of what we now recognize

ticulated and occasionally highly concreted adult skulls

as subadult Triceratops were collected (Schlaikjer 1935; Dodson

may occur in mudstones; however, nonadult Triceratops

et al. 2004). Hatcher collected for O. C. Marsh and Yale Uni-

are seldom found in sandstones and siltstones, except as

versity in the Upper Cretaceous of east central Wyoming in

isolated cranial elements in channel lag deposits. In Mon-

the Laramie ‘‘Ceratops’’ beds, now formally referred to as the

tana, relatively complete baby (post-neonate), juvenile,

Lance Formation of Maastrichtian age. Hatcher discovered

and subadult Triceratops skulls are found almost exclu-

over 30 partial to complete skulls, most of which were as-

sively in mudstone facies of the Hell Creek Formation. We

signed to the genus Triceratops, from a relatively restricted

hypothesize that a historical collecting bias and facies and

area near Lusk, in Niobrara County (previously a part of Con-

taphonomic factors are responsible for the limited num-

verse County), Wyoming. The result of Hatcher’s early work

ber of non-adult Triceratops skulls and skull elements

in the Lance Formation, supplemented by collections from

known prior to the publication of the first Triceratops cra-

the Hell Creek Formation of Montana and equivalent beds in

nial growth series by Horner and Goodwin in 2006. Baby,

North and South Dakota, Colorado, and Alberta and Saskatch-

juvenile, and subadult Triceratops are not as rare in the Hell

ewan, Canada, was the description of 16 species of Triceratops

Creek Formation as previously reported, and permit a

(Dodson et al. 2004). Most of these taxa are not diagnosable

more complete assessment of Triceratops systematics,

and instead reflect individual cranial variation and ontog-

ontogeny, morphology, and variation.

eny (Ostrom and Wellenhofer 1986; Forster 1996a, b; Dod-

551

FIGURE 36.1.

Index map of Montana showing the general location of the Hell Creek Project and area where the Triceratops described in this study occur near Jordan, Montana, at the black arrow. The CretaceousPaleocene boundary is delineated by the lowest laterally extensive coal bed (‘‘Z’’ coal).

son 1996). Today, most students of Triceratops recognize one

Natural History, New York; MOR: Museum of the Rockies,

taxon, Triceratops horridus (Ostrom and Wellenhofer 1986;

Bozeman; UCMP: University of California Museum of Paleon-

Dodson 1996; Lehman 1998), or alternatively two species,

tology, Berkeley.

T. horridus and T. prorsus (Forster 1996b). Regardless of your opinion on these matters, adult Triceratops skulls are well represented in museum collections across North America and

Geology

Europe (Ostrom and Wellnhoffer 1986; Dodson et al. 2004;

The Upper Cretaceous Hell Creek Formation of eastern Mon-

Goussard 2006).

tana (Fig. 36.1) was deposited as the continental facies of the

We hypothesize that a historical collecting bias, controlled

last major regression of the Western Interior Eperic Sea (Lof-

in part by facies and taphonomic factors, influenced the

gren 1995). The Hell Creek Formation in Garfield and Mc-

known fossil record of Triceratops for many decades. We pro-

Cone Counties is a predominantly fluvial deposit dominated

pose that these factors are likely responsible for the limited

by sandstones and siltstones that encroached eastward as a

number of non-adult Triceratops skulls and skull elements

prograding wedge of clastic sediment across a low, broad

known prior to the publication of the first cranial growth se-

coastal plain as the seaway regressed (Archibald 1982). The

ries by Horner and Goodwin (2006). Historically, this collect-

base of the Z coal complex that defines the upper formational

ing bias also played a considerable role in the evolution of

boundary near Hell Creek has a weighted mean 40Ar/ 39Ar age

Triceratops systematics, based primarily on horn and frill mor-

determination of 65.0 Ma (Swisher et al. 1993).

phology, and until recently, limited the evaluation of behavioral hypotheses and sexual dimorphism.

Beginning in the summer field season of 2000, J. Horner initiated the Hell Creek Project, at the time a five-year paleon-

This study is based largely on an assemblage of Tricera-

tological and geological investigation of the entire Hell Creek

tops from the Hell Creek Formation of Garfield and McCone

Formation exposed in Garfield and McCone Counties, eastern

Counties, Montana, in the collections of the Museum of the

Montana. This multi-institutional project continues to inves-

Rockies, Montana State University and the University of Cali-

tigate aspects of the sedimentology and stratigraphy, verte-

fornia Museum of Paleontology. Future work will test this

brate and invertebrate paleontology, paleobotany, taphon-

hypothesis of a facies-related collecting bias in contempo-

omy, and paleoecology of the Hell Creek Formation beyond

raneous beds that yield Triceratops in neighboring states and

the initial five-year plan. More than one colleague com-

Canada.

mented at the start, ‘‘Why are you exploring the Hell Creek

Institutional Abbreviations. AMNH: American Museum of

552 goodwin & horner

Formation? We already know everything about it.’’ To the

FIGURE 36.2.

An aerial view of the MOR ‘‘B. rex’’ site shows the significant amount of overburden (over 50 feet) removed to collect this partial skeleton, and the steep slope where the first bone was found by Bob Harmon. The grasscovered, rolling topography of the underlying Fox Hills Formation is exposed at the base of the cliff. This is the lowest stratigraphically occurring, and thus geologically oldest, Tyrannosaurus from the Hell Creek Formation of Montana.

contrary, the Hell Creek Formation is about 150 m (500 ft.)

Prospecting

thick and historically only the upper 45 m (148 ft.) was seriously prospected and collected in—for two very good rea-

Prospecting is not a ‘‘random walk.’’ Prior to actually walking

sons: (1) abundant vertebrate fossils weather out of the upper-

up and down outcrops of sedimentary rock, a number of steps

most part of the section; and (2) these exposures are relatively

occur. For instance, geological maps are consulted in detail;

easier to get to, compared with the lower portion of the forma-

archives of field notes and photographs are reviewed; anec-

tion. A comprehensive survey of the entire Hell Creek Forma-

dotal evidence is examined; badlands are scouted from the air;

tion in Montana, from the basal contact with the Fox Hills

and finally, boots on the ground confirm whether there are

Formation to the overlying Cretaceous-Tertiary boundary at

fossils to be found. Nonetheless, prospecting for fossils re-

the contact with the earliest Paleocene Tullock Formation was

mains both an art and a science. In 2000, Bob Harmon dis-

undertaken.

covered an exceptionally well-preserved, disarticulated skele-

Historical Collecting Bias and the Fossil Record of Triceratops in Montana 553

FIGURE 36.3.

An isolated right squamosal (MOR 2572) of juvenile Triceratops, discovered by M. B. Goodwin in 2006 (inset). Note the wavy posterior margin. Brush is 3.8 cm wide. This isolated specimen was found in situ (white arrow) in an upper grey mudstone. The Hell Creek Formation exposed in this area has very little sand and is composed primarily of siltstones and claystones. The weathered fractured surface of exposed bone was bleached white and looked like a fragment of a larger bone until it was exposed.

ton of Tyrannosaurus rex (MOR 1125; also known as ‘‘B. rex’’) at

The goal of early collectors was to find and collect new and

the base of the Hell Creek Formation, 8 m (26 ft.) above the

complete dinosaurs—preferably skulls or skeletons, and Tri-

contact with the underlying Fox Hills Formation (Fig. 36.2).

ceratops was one of the most common Late Cretaceous dino-

Bob stopped for lunch along a steep slope above the grassy

saurs in the Western Interior of North America. This fact has

knolls of the Fox Hills Formation to take advantage of the

not changed, but as the early collectors discovered, and we

shade. Looking up, he noticed the distal end of a metatarsal

have confirmed during the Hell Creek Project in the field and

poking out of a well consolidated, crossbedded sandstone. Af-

in the archives and collections of the MOR and UCMP, if you

ter careful consideration, and a walk back to camp to retrieve a

desire a relatively complete and undistorted dinosaur skull or

chair to stand on in order to reach the fossil, Bob proceeded

skeleton, success is more likely by prospecting in the sand-

to expose the metatarsal, not knowing at the time that it was

stones and coarser grain sediments (Fig. 36.4A–D).

but one bone of a beautifully preserved partial T. rex skeleton

Would you look in the relatively somber mudstone beds of

under 16 m (]50 ft.) of overburden. Bob’s efforts paid off

the Hell Creek Formation for a dinosaur here in Fig. 36.5?

handsomely when future studies documented the unexpected

Much of the Hell Creek Formation is comprised of these over-

preservation of soft tissue vessels and medullary bone in MOR

bank mudstones. They are laterally extensive and may also in-

1125 (Schweitzer et al. 2005a, b, 2007).

clude local sandy lag deposits. These beds often have well pre-

We all have our favorite facies and lithology when it comes

served fossil leaves ( Johnson 2002) and fresh water mollusks

to prospecting for dinosaurs, and dinosaur paleontologists are

(Hartman and Kirkland 2002)—and even some dinosaurs—but

known to be a competitive bunch. Some like to search in sand-

frequently the dinosaur bones are ‘‘exploded’’ on the surface

stones, or at the interface between sandstone and mudstone

as float, or are weathering out in situ, like this Triceratops skull

beds where bones might accumulate, or at the base of sand-

with its partial frill, horn and frill fragments littering the sur-

stone channels, or even in shell beds and lag deposits. Others

face (Fig. 36.5). This is caused, in part, by the expansion and

prefer to prospect in mudstones (Fig. 36.3), as these beds are

contraction of the mudstones due to fluctuations in the local

often easier to walk over, except when the popcorn-like ben-

water table and precipitation. Many of the dinosaurs from the

tonitic clays become unstable and create a surface as difficult

Hell Creek Formation mudstones are covered in a fine-grained

to negotiate as a hill covered in marbles. Mudstones are softer

concretion (Fig. 36.6), and the bone is often black and heavily

than sandstones, which may make it easier to remove over-

mineralized by iron and manganese (Goodwin et al. 2007).

burden with hand tools instead of power equipment, to fur-

Often, these dinosaur bones exhibit perpendicular fractures

ther expose fossil bones.

indicating subaerial exposure. Some Triceratops specimens

554 goodwin & horner

FIGURE 36.4.

Relatively complete and undistorted dinosaur skulls and skeletons are found primarily in the consolidated sandstones of the Hell Creek Formation in eastern Montana. (A) The articulated ‘‘Wankel T. rex’’ (MOR 555) in overhead view at the locality in the field. The skull is visible in dorsal view at the white arrow in the center of the photograph; (B) a close-up of MOR 555 in left lateral view from the cast of the skeleton mounted in the atrium of the UCMP; (C) adult Triceratops skull (MOR 004), aka MORT, in right lateral view (reversed); and (D) adult skull of Edmontosaurus, UCMP 128372, in right lateral view (reversed).

from the Hell Creek Formation mudstones show signs of

John Bell Hatcher observed Triceratops to be preserved in ex-

transportation, followed by burial and reburial by rework-

traordinary numbers in relatively coarser sedimentary facies,

ing. Compressed over millennia, followed by modern erosion

weathering out of the Lance Formation, Wyoming, literally by

and exposure, these fossils appear very fragmented as they

the dozens. As Hatcher (1896) notes, ‘‘In the nearly four years

weather out in the badlands. Skulls and skeletons are regularly

spent by the writer in working these beds, 31 skulls and several

disarticulated. This is in contrast to articulated, very well pre-

fairly complete skeletons of horned dinosaurs were secured.’’

served Triceratops and contemporaneous dinosaurs found in

Barnum Brown (1917: 281) ‘‘identified no less than five hun-

harder, well-cemented, coarser sandstones (Fig. 36.4).

dred fragmentary skulls and innumerable bones’’ of Triceratops during seven field seasons between 1902 and 1909. This is

The Fossil Record of Triceratops

likely not an exaggeration based on the field observations of the authors throughout seven field seasons of the Hell Creek

The discovery of Triceratops, as well as giant sauropods from the

Project. Early dinosaur collectors were days, if not weeks from

Jurassic Morrison Formation at the end of the nineteenth cen-

the steamboat landing or railroad station that served as their

tury, generated considerable interest from both the scientific

transportation hub. Resupplying was difficult and plaster and

community and the public. Major museums, many just founded

burlap would not be wasted on incomplete material. Sand-

in North America, desired complete skulls and skeletons for their

stones and siltstones were explored and less promising mud-

exhibits and collections. In the decades that followed and up

stone deposits with eroded and often highly weathered and

to the present day, natural history museums across the United

exploded Triceratops skulls (Fig. 36.2) were passed over in the

States and Canada invested heavily in collecting and mount-

search for more impressive, complete skulls from Hell Creek

ing complete dinosaurs for exhibit. Triceratops became an icon

Formation sandstones (Lull 1903). This was the strategy em-

for the end of the Age of Dinosaurs and this period is often

ployed by Mick Hager, director of the MOR in 1981, to col-

illustrated in exhibits and books showing agonistic encounters

lect a Triceratops skull in the Hell Creek Formation—the result

between Triceratops and its presumed rival, Tyrannosaurus.

was the complete adult skull referred to as ‘‘MORT’’ (MOR

Historical Collecting Bias and the Fossil Record of Triceratops in Montana 555

FIGURE 36.5.

A weathered Triceratops frill is visible at the white arrow above the Marsh pick in the center of the photograph. Fragments of horn, frill, and unidentified cranial elements litter the slope of this grey, bentonitic mudstone (Hell Creek Formation, Montana).

004), found in an upper Hell Creek Formation sandstone

AMNH 5006, a pair of diminutive, postorbital horns de-

(Fig. 36.4C).

scribed by Tokaryk (1997) from the Frenchman Formation of

Previously, only adult and a limited number of what we

southern Saskatchewan were referred to the Chasmosaurinae

now recognize as subadult Triceratops skulls were collected

since it could not be ruled out they may belong to a very

in Upper Cretaceous sediments from the Western Interior

young Torosaurus. In 1987, Harley Garbani, a UCMP field

of North America. Barnum Brown collected an isolated juve-

associate, discovered the smallest ceratopsid skull yet known

nile postorbital horn (AMNH 5006) in 1906 from the Hell

when he uncovered a baby (post-neonate) Triceratops (UCMP

Creek Formation, eastern Montana. Brown and Schlaikjer

154452) in the Hell Creek Formation, Garfield County, Mon-

(1940) later confirmed that the orbital horns of Triceratops

tana (Goodwin et al. 2006). In the summer of 2006, Sonya

were an outgrowth of the postorbital bones and not derived

Scarff, a member of an MOR field crew, discovered a second

from a separate dermal or epidermal ossification based on

baby Triceratops skull in the Hell Creek Formation, eastern

this specimen. Nearly a decade after Brown’s discovery of

Montana.

556 goodwin & horner

FIGURE 36.6.

Some dinosaurs found in Hell Creek Formation mudstones are covered with a white to grey silty concretionary layer. Bones are often black under these conditions. (A) Mary Schweitzer holds a maxillary fragment from this very weathered, broken subadult Triceratops skull in the field; (B) as badly preserved as this partial skull appears on the surface, the specimen confirms morphological details of the premaxillae, epinasal, and epijugal.

FIGURE 36.7.

Mark’s Trike II locality in the uppermost Hell Creek Formation. (A) Under excavation; (B) overburden being removed by graduate students and volunteers, and the popcorn-like surface of bentonitic mudstones visible in the background; (C) a very small and badly fractured frill fragment weathering out on the surface was the only bone visible before excavation. Brush is 3.8 cm wide.

Historical Collecting Bias and the Fossil Record of Triceratops in Montana 557

Results LOCALITIES IN MUDSTONES

Sierra Skull. Only 870 mm long, this skull (MOR 1199) is the best example yet known of the ‘‘Small Juvenile’’ ontogenetic stage of Horner and Goodwin (2006). This skull was found

Mark’s Trike II Site. This locality (Fig. 36.7) is about 15 m (50 ft.)

near Nelson Creek by a Sierra College field crew, under the

below the Z coal in the uppermost Hell Creek Formation. Only

direction of Dick Hilton, in a dark grey mudstone lens above a

a badly weathered piece of the frill was initially exposed (Fig.

carbonaceous shale (Fig. 36.13). Only the epinasal was miss-

36.7). Stop, or go on? With a crew of graduate students and

ing. This find was remarkable, as the skull was completely

volunteers available for the necessary labor and time required

disarticulated in the field. After careful preparation by Carrie

to resolve this question, we could determine what lies in the

Ancell, the elements articulate along sutural contacts and the

hill behind this very weathered, ‘‘trashy’’ dinosaur bone. Our

skull shows only modest distortion.

experience has confirmed that investing this time and energy,

Afternoon Delight. In the summer of 2006, (Fig. 36.14), an

that is, removing large amounts of overburden, followed by

MOR field crew came across a postorbital horn and ceratop-

careful hand quarrying in the search for more bones in situ,

sid skull fragments eroding out of the slope of a dark grey

pays off, particularly in the Hell Creek Formation mudstones

mudstone in the Hell Creek Formation. Careful hand quarry-

(see Horner and Goodwin 2006). Here, at Mark’s Trike II, along

ing and screening recovered the second partial baby (post-

strike next to the frill fragment, the proximal end of what

neonate) Triceratops skull (MOR 2569) yet known. Elements

turned out to be a complete postorbital horn was uncovered

preserved in situ included the parietal, left and right squamo-

and collected. Additional hand quarrying exposed a pair of

sals, left and right quadrates, right maxillary, and right jugal. A

dentaries, all belonging to one subadult Triceratops, MOR 2597

more detailed description of the skull is in preparation.

(Fig. 36.8). Skilled fossil preparation supported all the effort that went into collecting these well preserved fossils (Fig. 36.9), which at first glance, was uninspiring (Fig. 36.7C). This

Conclusions

specimen (MOR 2597) is an important datum in a larger study

Complete, undistorted adult Triceratops skulls are found pri-

of Triceratops cranial ontogeny and morphology by the au-

marily in sandstones and siltstones, which contribute gen-

thors. Additional Triceratops specimens, all nonadult skulls

erously to the sedimentary composition of the Hell Creek For-

and cranial elements found in mudstone beds of the Hell

mation, Montana, and the equivalent Lance Formation of

Creek Formation, are listed from the localities below.

Niobrara County, Wyoming (Hatcher et al. 1907; Dodson

High Triceratops. This locality, UCMP V88001, is in a local

1996). While disarticulated and sometimes highly concreted

basin northwest of McGuire Creek, McCone County, Mon-

adult Triceratops skulls may occur in Hell Creek Formation

tana (Fig. 36.10). Along the southern margin of the basin, in

mudstones, nonadult Triceratops are seldom found in sand-

the uppermost Hell Creek Formation, a disarticulated sub-

stones and siltstones, except as isolated cranial elements

adult Triceratops skull (UCMP 137263) was collected by a

within channel lag deposits or as ‘‘float’’ bone on the surface.

UCMP field crew in a grey fissile mudstone. A pair of dentary

These small, isolated bones were likely overlooked or simply

bones, most of the frill, both postorbital horns, and a quadrate

not recognized by early collectors in the field.

were preserved. Plant fossils were also found in these beds.

Ontogenetically younger cranial elements eroding out on

Russell Basin. In the early 1980’s, a UCMP field crew col-

the surface look a lot like fragmentary pieces of larger and pre-

lected a subadult Triceratops skull (UCMP 136092; est. 160 cm

sumably incomplete adult skulls and bones. Consequently, it

long) from locality V88081 (Fig. 36.11). At least three addi-

is possible that partial and disarticulated nonadult skulls and

tional, but badly weathered partial Triceratops skulls and/or

isolated Triceratops bones have been misidentified as the frag-

cranial elements were observed weathering out of the abun-

mentary remains of adults and were subsequently left uncol-

dant mudstones within 1 km (0.6 mi) from this locality in

lected. As it turns out, the mudstone facies in the Hell Creek

Russell Basin during the 2007 summer field season by mem-

Formation are where numerous isolated elements, as well as

bers of the Hell Creek project. The fossils were photographed

closely associated and often disarticulated, but relatively com-

and GPS coordinates were recorded for an ongoing census.

plete baby, juvenile, and subadult Triceratops skulls are found

Getaway Trike. Excavation of the Getaway Trike site (Fig.

(Horner and Goodwin 2006, 2008). In the past, targeting only

36.12) revealed a nearly complete, exceptionally well pre-

sandstones and siltstones, whilst overlooking mudstone beds

served disarticulated subadult Triceratops skull (MOR 1120) in

effectively guaranteed that nonadult Triceratops would remain

a mudstone bed. This skull is an excellent example of a speci-

under-collected in the field and under-represented in museum

men in the field that required quarrying by hand. The cranial

collections until recently.

elements were all separate and in different orientations. After

This record played a considerable role in the evolution of

careful preparation, the 165 cm long skull articulates along

Triceratops systematics, resulting in the naming of as many as

sutural contacts with very little distortion.

16 species on the basis of horn and frill morphology. Histori-

558 goodwin & horner

FIGURE 36.8.

Mark’s Trike II locality. (A) Graduate student Gina Sorrentino points to a pair of dentaries exposed behind the weathered frill and complete postorbital horn; (B) close-up of the paired dentaries. Brush is 3.8 cm wide.

FIGURE 36.9.

After preparation by Carrie Ancell, the fine preservation of the dentaries and right postorbital horn of MOR 2597 is revealed in this photograph taken in the MOR collections. Note the tip of the right postorbital horn is oriented posteriorly, whereas the remaining section of the horn is now anterior, a subadult ontogenetic character.

Historical Collecting Bias and the Fossil Record of Triceratops in Montana 559

FIGURE 36.10.

UCMP locality V88001 (High Ceratopsian) in the Charles M. Russell Wildlife Refuge, McCone County, Montana. (A) Current photograph; (B) slight depression still visible in the bentonitic mudstone (Hell Creek Formation) at the quarry site (excavated in 1990). This locality yielded a disarticulated subadult Triceratops skull (UCMP137263). The left (C) and right (D) postorbital horns are directed posteriorly. The right parietal is shown exposed in the field in ventral view (E).

FIGURE 36.11.

Subadult Triceratops skull (UCMP 136092; Hell Creek Formation) from UCMP locality V88081 in Russell Basin near Bug Creek, McCone County, Montana. The anterior portion of the skull was found weathering out of a grey mudstone along the base of the badlands. Graduate student Denver Fowler stands at the locality for scale at the white arrow, just below the CretaceousTertiary boundary (coaly interval).

560 goodwin & horner

FIGURE 36.12.

The Getaway Trike site yielded a subadult Triceratops skull (MOR 1120; 165 cm long; disarticulated) in mudstone. When reconstructed the skull was three-dimensional and beautifully preserved, showing sutural surfaces and morphological detail (inset). The basioccipital is visible in the quarry at the black arrow where Bob Harmon (middle) exposes additional elements with two MOR crew members.

FIGURE 36.13.

The Sierra Skull (MOR 1199; 87 cm long) is a small juvenile Triceratops (A) found by Dick Hilton and a field crew from Sierra College; (B) host mudstone (Hell Creek Formation) in McCone County, Montana; (C) crew puts a plaster jacket on the frill. The skull was entirely disarticulated when found and is undistorted and nearly complete except for the epinasal.

Historical Collecting Bias and the Fossil Record of Triceratops in Montana 561

FIGURE 36.14.

Aerial view of the Afternoon Delight locality shows field crew (white arrow) collecting second baby Triceratops skull known (MOR 2569). Postorbital horn was found on the surface by student Sonya Scarff. Excavation in a grey mudstone uncovered a well-preserved, disarticulated skull including frill (inset; scale bar is 10 cm), left and right quadrates, right maxillary, and right jugal.

cally, the scarcity of non-adult Triceratops fossils collected from Upper Cretaceous sediments of North America limited our understanding of Triceratops systematics, cranial ontogeny and morphology. A more complete assessment of sexual dimorphism and variation is now possible. Acknowledgments

We thank Carrie Ancell for her exceptional fossil preparation, and Bob Harmon, Nels Peterson, Patrick Leiggi, and the many MOR and UCMP field crew members who participated in the Hell Creek Project. Harley Garbani discovered and patiently reassembled the partial juvenile Triceratops skull, UCMP 128562. Harley has greatly advanced the field of dinosaur paleontology by his numerous discoveries and unending enthusiasm for taking the time to ‘‘look over that next hill’’ in the badlands of eastern Montana. We thank Bill Clemens for his thoughtful comments and insightful discussion on this research topic. The generous financial support of Nathan Myhrvold for the Hell Creek Project is gratefully acknowledged. The University of California Museum of Paleontology (UCMP) provided funding to MG. The support and assistance of the Museum of the Rockies, UCMP, the Bureau of Land Management, the United States Fish and Wildlife Service, and the Charles M. Russell Wildlife Refuge are sincerely appreciated. References Cited Archibald, J. D. 1982. A study of mammalian and geology across the Cretaceous-Tertiary boundary in Garfield County, Mon-

562 goodwin & horner

tana. University of California Publications in Geological Sciences 122. Berkeley: University of California Press. Brown, B. 1917. A complete skeleton of the horned dinosaur Monoclonius, and a description of a second skeleton showing skin impressions. Bulletin of the American Museum of Natural History 37: 281–306. Brown, B., and E. M. Schlaikjer. 1940. The origin of ceratopsian horncores. American Museum Novitates 1065: 1–7. Dodson, P. 1996. The Horned Dinosaurs. A Natural History. Princeton: Princeton University Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, 2nd ed., pp. 494–513. Berkeley: University of California Press. Forster, C. A. 1996a. New information on the skull of Triceratops. Journal of Vertebrate Paleontology 16: 246–258. ———. 1996b. Species resolution in Triceratops: Cladistic and morphometric approaches. Journal of Vertebrate Paleontology 16: 259–270. Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian. 2006. The smallest known Triceratops skull: New observations on ceratopsid cranial anatomy and ontogeny. Journal of Vertebrate Paleontology 26: 103–112. Goodwin, M. B., P. G. Grant, G. Bench, and P. A. Holroyd. 2007. Elemental composition and diagenetic alteration of dinosaur bone: Distinguishing micron-scale spatial and compositional heterogeneity using PIXE. Palaeogeography, Palaeoclimatology, Palaeoecology 253: 458–476. Goussard, F. 2006. The skull of Triceratops in the palaeontology gallery, Muséum National d’Histoire Naturelle, Paris. Geodiversitas 28: 467–476.

Hartman, J. H., and J. I. Kirkland. 2002. Brackish and marine mollusks of the Hell Creek Formation of North Dakota: Evidence for a persisting Cretaceous seaway. In J. H. Hartman, K. R. Johnson, and D. J. Nichols, eds., The Hell Creek Formation and the Cretaceous-Tertiary Boundary in the Northern Great Plains: An Integrated Continental Record of the End of the Cretaceous, pp. 271–296. Geological Society of America Special Paper 361. Hatcher, J. B., 1896. Some localities for Laramie mammals and horned dinosaurs. American Naturalist 30: 112–120. Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49: 1–300. Horner, J. R., and M. B. Goodwin. 2006. Major cranial changes during Triceratops ontogeny. Proceedings of the Royal Society B 273: 2757–2761. ———. 2008. Ontogeny of cranial epi-ossifications in Triceratops. Journal of Vertebrate Paleontology 28: 134–144. Johnson, K. R. 2002. Megaflora of the Hell Creek and lower Fort Union Formations in the western Dakotas: Vegetational response to climate change, the Cretaceous-Tertiary boundary event, and rapid marine transgression. In J. H. Hartman, K. R. Johnson, and D. J. Nichols, eds., The Hell Creek Formation and the Cretaceous-Tertiary Boundary in the Northern Great Plains: An Integrated Continental Record of the End of the Cretaceous, pp. 329–391. Geological Society of America Special Paper 361. Lehman, T. M. 1998. A gigantic skull and skeleton of the horned dinosaur Pentaceratops sternbergi from New Mexico. Journal of Vertebrate Paleontology 72: 894–906. Lofgren, D. L. 1995. The Bug Creek problem and the CretaceousTertiary transition at McGuire Creek, Montana. University of

California Publications in Geological Sciences 140. Berkeley: University of California Press. Lull, R. S. 1903. Skull of Triceratops serratus. Bulletin of the American Museum of Natural History Article 19: 685–695. Marsh, O. C. 1889. Notice of gigantic horned Dinosauria from the Cretaceous. American Journal of Science Series 3, 38: 501–506. Ostrom, J. H., and P. Wellnhofer. 1986. The Munich specimen of Triceratops with a revision of the genus. Zitteliana 14: 111–158. Schlaikjer, E. M. 1935. The Torrington Member of the Lance Formation and a study of a new Triceratops. Bulletin of the Museum of Comparative Zoology 76: 29–68. Schweitzer, M. H., Z. Suo, R. Avci, J. M. Asara, M. A. Allen, F. T. Arce, and J. R. Horner. 2007. Analyses of soft tissue from Tyrannosaurus rex suggest the presence of protein. Science 316: 277– 280. Schweitzer, M. H., J. L. Witteyer, and J. R. Horner. 2005b. Genderspecific reproductive tissue in ratites and Tyrannosaurus rex. Science 308: 1456–1460. Schweitzer, M. H., J. L. Witteyer, J. R. Horner, and J. K. Toporski. 2005a. Soft-tissue vessels and cellular preservation in Tyrannosaurus rex. Science 307: 1952–1955. Swisher, C. C., III, L. Dingus, and R. F. Butler. 1993. 40Ar/39Ar dating and magnetostratigraphiccorrelation of the terrestrial Cretaceous-Paleogene boundary and Puercan mammal age. Canadian Journal of Earth Sciences 30: 1981–1996. Tokaryk, T. T. 1997. First evidence of juvenile ceratopsians (Reptilia: Ornithischia) from the Frenchman Formation (late Maastrichtian) of Saskatchewan. Canadian Journal of Earth Sciences 34: 1401–1404.

Historical Collecting Bias and the Fossil Record of Triceratops in Montana 563

AFTERWORD PHILIP J. CURRIE

And there you have it! A remarkable roundup of 36 chapters,

Sternberg and many of the other, earlier ceratopsian work-

plus two extensive articles on the CD-ROM, contributed by 66

ers were unable to attend of course, but were there in spirit.

ceratopsian researchers and students from around the world.

John B. Hatcher was even represented ‘‘in the flesh’’ by one of

Only a few short years ago, the number of people working on

his descendants, Joseph Hatcher, who gave a talk about Tri-

dinosaurs in the entire world was considerably less than the

ceratops on the final day of the conference (Hatcher 2007). At

number of authors in this book. In fact, it was very rare during

least three successional generations of ceratopsian researchers

the first century of ceratopsian studies (starting with Cope’s

—from Wann Langston Jr. who started publishing on horned

description of Agathaumas in 1872) for more than two cera-

dinosaurs in the 1960s (just think of it, five decades of work-

topsian researchers to be active at the same time. A good in-

ing on ceratopsians!), to Peter Dodson and Jack Horner who

dication of how little work was being done was mentioned in

began to publish in the 1970s, all the way to the youngest

the first chapter (Dodson, this volume)—no new genera of

ceratopsian workers (in this case, probably Tetsuto Miyashita,

ceratopsians were reported between 1950 and 1986!

who is still an undergraduate at the time this was written)—are

Review chapters on basal ceratopsians (You and Dodson

all represented.

2004) and the Ceratopsidae (Dodson et al. 2004) recently rec-

With the increased number of people doing research on cera-

ognized about 40 valid species of ceratopsians. There are an

topsian dinosaurs, it is no surprise that this has led to an incr-

additional 50 specific names (and combinations of names)

eased rate of discovery of new ceratopsian genera and species.

that are no longer considered valid. Cope and Marsh in fact

Since the Ceratopsian Symposium was held in Drumheller in

assigned half of these invalid names during the last three de-

2007, 14 new names have appeared (Albertaceratops nesmoi,

cades of the nineteenth century. Within five years of publica-

Ryan 2007; Archaeoceratops yujingziensis, You et al. 2010; Cer-

tion of the two review papers in 2004, however, the number

asinops hodgskissi, Chinnery and Horner 2007; Coahuilacera-

of ceratopsid species that were considered valid had doubled

tops magnacuerna, Loewen et al. 2010; Diabloceratops eatoni,

(Sampson 2008).

Kirkland and DeBlieux 2010; Eotriceratops xerinsularis, Wu et al.

To put this publication in context, this is undoubtedly

2007; Gobiceratops minutus, Alifanov 2008; Medusaceratops

the most significant ceratopsian volume since the Hatcher,

lokii, Ryan et al. 2010; Microceratus sulcidens, Mateus 2008;

Marsh, and Lull monograph that was published in 1907, and

Ojoceratops fowleri, Sullivan and Lucas 2010; Pachyrhinosaurus

Lull’s ‘‘Review of the Ceratopsia’’ in 1933. Several other cera-

lakustai, Currie et al. 2008; Psittacosaurus major, Sereno et al.

topsian publications of note include Dodson’s The Horned

2007; Rubeosaurus ovatus, [Gilmore 1930] gen. nov. and sp.

Dinosaurs (1996), written for a broad audience, and a collec-

com. nov., McDonald and Horner 2010; Tatankaceratops sacri-

tion of ceratopsian and ornithopod papers edited by Car-

sonorum, Ott and Larson 2010), seven of them in this volume!

penter (2007). Another, more recent volume (Currie et al.

Although ceratopsians were relative latecomers in the dino-

2008) contains multiple papers focused on just one species of

saurian world, they were clearly undergoing an explosive radi-

Pachyrhinosaurus.

ation in the northern hemisphere during Cretaceous times.

Whatever expectations we may have had when the cera-

Even more astonishing than the sudden increase in cera-

topsian conference was originally discussed, the number of

topsian researchers and taxa is the diversity of the research

papers presented at the conference (and now the number of

being done on these incredible animals. This is evident in this

papers written for this volume) surpasses anything that has

volume. The historical data and perspectives presented by

gone before. Barnum Brown, Edward Drinker Cope, John B.

Dodson in chapter 1, by Goodwin, Horner, and Tanke in the

Hatcher, Richard Swann Lull, Othniel C. Marsh, Henry Fair-

concluding chapters of the book, and by Tanke on the supple-

field Osborn, Halszka Osmólska, John Ostrom, Charles M.

mentary CD-ROM help consolidate the foundations of cera-

565

topsian research by giving us a better understanding of who,

ertus bonebeds in Dinosaur Provincial Park alone, plus one

when, and why specimens were collected. Collecting biases

Centrosaurus brinkmani bonebed and two Styracosaurus alber-

have always existed, and will probably continue to exist in

tensis bonebeds, and the minimum number of individuals add

different ways. These can only be identified and rectified by

up into the thousands (Eberth and Getty 2005). Other cera-

studying the historical context within which specimens were

topsian bonebeds include Psittacosaurus (Gardner 2007

(and are) recovered. Basic data on specimens is available in

reported the discovery of 67 individuals at one site alone) and

the Ford and Tanke contributions on the supplementary CD-

Protoceratops in Asia, and Agujaceratops, Albertaceratops, Einio-

ROM, and throughout the volume in individual papers. With-

saurus, Torosaurus, and Triceratops in North America (Hunt and

out knowing where specimens were collected, the potential

Farke, this volume). Regardless of whether one accepts that

for incorporating existing specimens into more modern stud-

ceratopsians were herding animals or that other processes are

ies is seriously impaired. Tanke has performed an amazing

responsible for these monodominant bonebeds (Brinkman et

service to future research by tirelessly ferreting out ‘‘lost’’ in-

al. 2007), these now common and spectacularly rich accumu-

formation that now provides new raw data for analysis.

lations of fossils provide an amazing resource for analyzing

In addition to describing the anatomy of new taxa (chapters

virtually all aspects of ceratopsian paleobiology, including

3, 8, 9, 11, 12, and 14), new information is provided by newly

(but not limited to) age, behavior, evolutionary rates, diversifi-

discovered fossils or by the re-examination of existing speci-

cation, growth rates, and variation (individual, ontogenetic,

mens of known taxa (chapters 4, 6, 10, and 13). The anatomi-

and sexual).

cal work, often with information on different ontogenetic lev-

The number of specimens available as newly collected skele-

els, helps to establish and clarify the relationships among

tons or specimens from bonebeds explains (in part) the explo-

various ceratopsian taxa. Descriptive work, sometimes at a

sion of research activity that is currently centered on cera-

histological level (chapter 17), is also essential before delving

topsian dinosaurs. In spite of the large number of papers in

into the structure and function of horns (chapters 19 and 20)

this volume, the work is far from done, and most of the

and frills (chapters 17, 18, and 20).

authors (ceratophiles in Dodson terminology!) represented

Whereas the majority of taxa come from regions like the

here will continue to study ceratopsians. Many are already

Gobi Desert of China and Mongolia, or the Western Interior

collecting additional data and specimens, restudying signifi-

Basin of Canada and the United States, chapter 7 describes

cant specimens, writing new papers, and training new stu-

ceratopsian material from Coahuila, Mexico, a region that few

dents. To date, the only other dinosaurs that have attracted

know about. Larger scale biogeographic information is dis-

similar levels of activity are the Theropoda (and of course their

cussed in chapters 26 and 27. Refined biostratigraphic and

descendants—birds).

taphonomic research has helped us to better understand the environments in which ceratopsians lived.

In assembling this magnificent volume with all its diversity in near record time following the symposium that spawned

The behavior and habits of ceratopsians are assessed using

and/or focused the research, the editors merit our undying

anatomy (chapters 22 and 23), pathology (chapters 24 and

praise. In spite of the relative speed with which the papers

25), and taphonomy (chapters 29, 30, and 32), and come up

have been assembled and published, each one has been peer-

with surprising interpretations suggesting that some cera-

reviewed and edited for consistency to produce this quality

topsians were nocturnal, others were semi-aquatic, and yet

publication that provides a new foundation for the current

others moved in large herds. The evidence for gregarious or

state of ceratopsian affairs.

aggregate behavior (Brinkman et al. 2007) in ceratopsians has become overwhelming since the Provincial Museum of Alberta started to excavate their first Centrosaurus bonebed (Quarry 143) in Dinosaur Provincial Park in 1979. We know that collections had been made in this bonebed previously by Barnum Brown in 1915 because of specimens and photographs in the American Museum of Natural History. The Sternbergs had also collected from ceratopsian bonebeds (see chapter S-2 on the supplementary CD-ROM) in the same region and around the same time. The concentrations led C. M. Sternberg (1970) to speculate that Anchiceratops, Centrosaurus, Pachyrhinosaurus, and Styracosaurus congregated and died in swampy areas at certain times. Over the past three decades in particular, the number of monodominant bonebeds has increased dramatically. There are more than 17 Centrosaurus ap-

566 afterword

References Cited Brinkman, D. B., D. A. Eberth, and P. J. Currie. 2007. From bonebeds to paleobiology: Applications of bonebed data. In R. R. Rogers, D. A. Eberth, and A. R. Fiorillo, eds., Bonebeds: Genesis, Analysis and Paleobiological Significance, pp. 221–265. Chicago: University of Chicago Press. Carpenter, K., ed. 2007. Horns and Beaks: Ceratopsian and Ornithopod Dinosaurs. Bloomington: Indiana University Press. Cope, E. D. 1872. On the existence of the Dinosauria in the transition beds of Wyoming. American Philosophical Society, Proceedings 12: 481–483. Currie, P. J., W. Langston Jr., and D. H. Tanke. 2008. A new species of Pachyrhinosaurus (Dinosauria, Ceratopsidae from the Upper Cretacious of Alberta, Canada. In P. J. Currie, W. Langs-

ton, Jr., and D. H. Tanke, A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta. Ottawa: NRC Research Press. Dodson, P. 1996. The Horned Dinosaurs, a Natural History. Princeton: Princeton University Press. Dodson, P., C. A. Forster, and S. D. Sampson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, Second Edition, pp. 494–513. Berkeley: University of California Press. Eberth and Getty. 2005. Ceratopsian bonebeds: Occurrence, origins, and significance. In P. J. Currie and E. B. Koppelhus, eds., Dinosaur Provincial Park, a Spectacular Ecosystem Revealed, pp. 501–536. Bloomington: Indiana University Press. Gardner, A. M. 2007. Digging for Dinos in the Land of Genghis Khan. Discover Magazine. Hatcher, J. 2007. Why Triceratops did not live in herds: evidence from the geological component. In D.R. Braman, compiler, Ceratopsian Symposium, Short Papers, Abstracts and Programs, pp. 75–77. Drumheller: Royal Tyrrell Museum of Palaeontology.

Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The Ceratopsia. U.S. Geological Survey Monograph 49. Lull, R. S. 1933. A revision of the Ceratopsia or horned dinosaurs. Peabody Museum of Natural History, Memoirs 3, part 3. Sampson, S. D. 2008. Foreword. In P. J. Currie, W. Langston Jr., and D. H. Tanke, eds., A New Horned Dinosaur from an Upper Cretaceous Bone Bed in Alberta, pp. v–viii. Ottawa: NRC Research Press. Sternberg, C. M. 1970. Comments on dinosaurian preservation in the Cretaceous of Alberta and Wyoming. National Museums of Canada, National Museum of Natural Sciences, Publications in Palaeontology No. 4, 9 pp. You, H.-L., and P. Dodson. 2004. Ceratopsidae. In D. B. Weishampel, P. Dodson, and H. Osmólska, eds., The Dinosauria, Second Edition, pp. 478–493. Berkeley: University of California Press.

afterword

567

Index

Italicized page numbers refer to illustrations. Aalenian stage: ceratopsian cladistics and, 400 Aardvarks: in arid environments, 323 Abelisaurs: caudal pathologies in, 368–369 Abrasion: of Hilda mega-bonebed fossils, 495, 502–503, 503, 504, 505; of KikakTegoseak Quarry fossils, 469 Abscesses, 346 Absolute ages: of ceratopsian species, 413, 414, 419–422, 422–423, 428 Abundance: Big Bend microsites, 522–524; ceratopsian, 429; Hell Creek and Lance Triceratops, 555, 558–562; Kaiparowits ceratopsid B, 484–485; Kaiparowits Formation faunal elements, 481; psittacosaur, 329 Academy of Natural Sciences: Peter Dodson’s work at, 7–8 Acanthophis: nocturnal lifestyle of, 320 Accessory antorbital fenestra: in ceratopsians, 132–134. See also Antorbital fenestrae Accessory dermal elements: psittacosaur, 42–43 Achelousaurus: in centrosaurine cladistics, 164; in centrosaurine evolution, 163, 165; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; in ceratopsid species diversity and turnover, 422; Diabloceratops eatoni n. gen. & sp. versus, 128; frill, 132; Kikak-Tegoseak Quarry Pachyrhinosaurus versus, 467; Medusaceratops lokii n. gen. & sp. versus, 183; from Montana, 181; MOR 449 versus, 161; in Pachyrhinosaurus n. sp. cladistics, 153; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153–154; Pachyrhinosaurus n. sp. versus, 141, 142, 145, 146, 147, 150; phylogeny, 411; provenance, 431; Rubeosaurus ovatus n. gen. &

comb. versus, 158, 161, 162, 163; stratigraphy, 163; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216 Achelousaurus horneri: bonebeds, 448, 449, 450, 450; in ceratopsian cladistics, 409; ceratopsid species diversity and turnover and, 421–422; frill, 164; occurrence of, 413; Rubeosaurus ovatus n. gen. & comb. versus, 159, 161, 166; skull, 406; skull and cladogram, 295, 296; skull strengths and measurements, 299, 300, 301, 302, 303, 304–305, 304; stratigraphy, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 442; taxonomy, 408; from Two Medicine Formation, 156 Acoustic cues: for nocturnal birds, 323 Acromegaly: Centrosaurus, 356 Actinomyces: osteomyelitis and, 348 Acuity. See Visual acuity Adobe Photoshop Elements 4.0: in chasmosaurine intraspecific interaction analysis, 284 Adocids: in Rattlesnake Mountain microsites, 525 Adocus: in Rattlesnake Mountain microsites, 524, 525 Adult skulls: Liaoceratops yanzigouensis, 245; psittacosaur, 40–45, 43; Tatankaceratops sacrisonorum n. gen. & sp., 203, 204, 204, 205–211; Triceratops, 271, 272–273, 272, 552. See also Skulls Adults: in bonebeds, 452, 456; distinguishing via bone surface texture, 251, 252, 253; Hell Creek Triceratops, 556; Kaiparowits ceratopsid B, 488; in Triceratops intraspecific interactions, 288 Aeolian settings. See Eolian settings Africa: extant sympatric bovids in, 293; Late Cretaceous/early Tertiary, 522;

Mesozoic paleogeography of, 395, 396, 397; semi-aquatic animals in, 199; semiaquatic dinosaurs from, 329; William Cutler’s death in, 541, 542; William Cutler’s expedition to, 542 Afternoon Delight Triceratops locality, 558, 562 Agathaumas: Cope’s description of, 565 Age of Dinosaurs: Triceratops and, 555 Ages: of ceratopsian species, 413, 414, 419–422, 422–423, 428, 439–446; of Kikak-Tegoseak Quarry palynomorphs, 461; of Prince Creek Formation, 458, 459 Agonistic behavior: among ceratopsids, 374–379 Aguillón-Martínez, Martha C., xvii, 99 Aguja Formation: ankylosaurs from, 527– 528, 529, 530; in Big Bend National Park geology, 521, 523; bonebeds, 448, 450, 451, 520–537; ceratopsian paleoenvironmental associations and taphonomy in, 444; ceratopsians from, 530; in ceratopsid stratigraphy, 412; dating, 413; dromaeosaurids from, 529, 531–532, 532; hadrosaurids from, 528–530, 529, 530; lithostrigraphy, 523; in North American paleogeography, 415; ornithomimids from, 525, 529, 531; tyrannosaurids from, 529, 530–531, 530 Agujaceratops: bonebeds, 186, 520–537, 566; in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 284, 285, 286; Coahuilaceratops magnacuerna n. gen. & sp. versus, 99, 104, 106, 108, 109, 111; Diabloceratops n. gen. versus, 133; distribution, 100; frill function, 283; frills as sexual display structures, 6; Medusaceratops lokii n. gen. & sp. versus, 185; taxonomy, 411

569

Agujaceratops mariscalensis: from Aguja Formation microsites, 530; in Big Bend microsite formation, 534; in Big Bend National Park geology, 521; biogeography, 111; bonebeds, 448, 449, 450, 451, 520–537; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 444; in chasmosaurine intraspecific interaction analysis, 285, 286; CMN 8547 versus, 195–196; Coahuilaceratops magnacuerna n. gen. & sp. versus, 109; Diabloceratops n. gen. versus, 133; Kaiparowits ceratopsids versus, 491; occurrence of, 413, 416; in ‘‘Purple Hill’’ section, 527; in Rattlesnake Mountain microsites, 525; skull, 407; skull and cladogram, 295, 296; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412; taxonomy, 520. See also Chasmosaurus mariscalensis Alag Teeg: ceratopsian paleoenvironmental associations and taphonomy in, 442 Alamo Wash, 170, 170 Alamo Wash individual (Ojoceratops n. gen.), 176 Alamosaurus: in dating Naashoibito Member, 171; occurrence of, 416 Alamosaurus sanjuanensis: from Ojo Alamo Formation, 170; Ojoceratops fowleri n. gen. & sp. and, 178 Alaska: basal ceratopsians from, 387, 390; bonebeds, 448, 450, 452, 456–477, 458, 459; centrosaurines from, 141; in ceratopsian biogeography, 111; in ceratopsian dispersal, 399; ceratopsian paleoenvironmental associations and taphonomy in, 428, 431, 434, 443; in ceratopsid fossil record, 418; ceratopsid stratigraphy in, 412; pachyrhinosaurs from, 408 Alaska pachyrhinosaur: biogeography, 111. See also Pachyrhinosaurus n. sp. (Alaska); Prince Creek pachyrhinosaur Alaskan dinosaurs, 6 Albanerpeton: in Rattlesnake Mountain microsites, 525 Albatrosses: eye sizes, 312; nocturnal, 320 Alberta, 142, 190, 496, 497; Aguja Formation versus geology of, 527, 533–534; Albertaceratops from, 181; Albertaceratops nesmoi from, 407–408; basal ceratopsians from, 222, 390; basal neoceratopsians from, 83–90, 431; Big Bend dinosaurs versus those of, 521, 533; bonebeds, 448, 449, 450, 451, 452–453; centrosaurine

570 index

bonebeds from, 495; centrosaurine mega-bonebed from, 495–508, 496, 497, 498, 499, 500, 503, 504; centrosaurines from, 130; Centrosaurus brinkmani from, 408; ceratopsian bonebeds in, 186; ceratopsian distribution in, 100, 110; ceratopsian paleoenvironmental associations and taphonomy in, 441, 443, 444, 445, 446; ceratopsians from, xiv; ceratopsid biostratigraphy in, 187; ceratopsid distribution in, 99, 111; in ceratopsid fossil record, 419; ceratopsid species diversity and turnover in, 419, 422; ceratopsid stratigraphy in, 412; ceratopsid teeth from, 530; Charles M. Sternberg in, 545; chasmosaurine bonebeds in, 356; chasmosaurines from, 408; Chasmosaurus russelli from, 187; CMN 8547 from, 189– 202; dating formations of, 413; Eoceratops collected in, 541, 542; geologic correlation in, 84; Kaiparowits Formation and, 478; Leptoceratops gracilis from, 243; letters from Cutler to Woodward from, 548; locality map, 85; Medusaceratops lokii n. gen. & sp. material in, 182; Montanoceratops cerorhynchus from, 69, 76– 77; neoceratopsian bonebeds in, 433; neoceratopsian taphonomy in, 433; pachyrhinosaurs from, 408; Pachyrhinosaurus n. sp. from, 141–155; paleoenvironments of, 414–416, 415, 417; paleopathologies in ceratopsids from, 355–384; pathological specimens from, 355, 356, 378, 379; Peter Dodson’s work in, 7; sympatric ceratopsians from, 300– 303, 303; Triceratops from, 551; Triceratops horridus from, 417 Albertaceratops: bonebeds, 566; in centrosaurine cladistics, 164; Ceratops montanus versus, 182; in ceratopsian family tree, 134; Diabloceratops eatoni n. gen. & sp. versus, 124, 131; frill, 132, 132; immature versus mature bone surface texture in, 252; Kaiparowits ceratopsid C versus, 489; long postorbital horns, 283; Medusaceratops lokii n. gen. & sp. versus, 181, 183, 185; Medusaceratops n. gen. versus, 182; in Pachyrhinosaurus n. sp. cladistics, 153; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153–154; Pachyrhinosaurus n. sp. versus, 146; phylogeny, 405, 411; provenance, 431; Rubeosaurus ovatus n. gen. & comb. versus, 160; squamosal pathologies in, 361; taxonomy, 418; Zuniceratops christopheri versus, 96 Albertaceratops nesmoi, 565; bonebeds, 448,

449, 450, 450; in ceratopsian cladistics, 409; in ceratopsid fossil record, 418, 419, 422; Medusaceratops lokii n. gen. & sp. versus, 185; Rubeosaurus ovatus n. gen. & comb. versus, 159, 161, 166; skull, 406; skull and cladogram, 295, 296, 297–298; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; taxonomy and distribution, 407–408 Albertosaurines: species diversity and turnover among, 422 Albertosaurus: skull collected by J. B. Tyrrell, 542 Albian stage: Archaeoceratops yujingziensis n. sp. from, 60; basal ceratopsians from, 387, 390, 391, 429; Bering Land Bridge established by, 396; ceratopsian cladistics and, 400; ceratopsian family tree and, 134; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440, 441; dating Utah fossils from, 394; in dinosaur paleobiogeography, 389; distribution of ceratopsians during, 389–391; Kikak-Tegoseak Quarry palynomorphs from, 461; Mesozoic paleogeography during, 398; neoceratopsians from, 96; psittacosaurs from, 22–23, 28, 32, 34, 329 Albuquerque: Zuniceratops christopheri and, 92 Alcedinids: eye sizes, 312 Alcelaphus: chasmosaurines versus, 283 Alcids: eye sizes, 312 Algae: in Kikak-Tegoseak Quarry bonebed, 456; Prince Creek Formation, 473 Algor FEMPRO 20 software, 266 All About Dinosaurs (Andrews), 3 Allied Art Centre: Calgary Public Museum collection at, 544 Alligator: eye size, 311 Alligator mississippiensis: ontogenetic bone surface texture changes in, 252 Alligators: juvenile, 8; Peter Dodson’s work on, 4, 5 Allometric growth: of Triceratops postorbital horns, 287–288 Allometry: among bird eyes, 313–314, 315, 316; among dinosaur eyes, 313–314, 315, 316, 317, 318; of eye size versus body mass, 312–313, 313, 315, 316, 318 Allosaurus: diaphyseal microfractures in, 349; osteomyelitis in, 347; pedal pathology in, 346

Allosaurus jimmadseni: eye size and body mass of, 317, 318 Alluvial channels: Big Bend microsites and, 533 Alluvial plains: basal ceratopsians in, 429, 431; in Big Bend National Park geology, 521; bonebeds in, 449, 451, 456; as ceratopsid paleoenvironments, 414–416; Hilda mega-bonebed deposited in, 503– 505; Kaiparowits Formation and, 479– 480; neoceratopsians in, 431; Prince Creek Formation and, 458, 460 Almond Formation: ceratopsian distribution in, 111 Alvarezsaurids: collected from Bayn Dzak, 320, 322 Alveolar pedestal: psittacosaur, 41 Amber: in Kikak-Tegoseak Quarry bonebed, 456 American buffalo: Pachyrhinosaurus versus, 377 American Carrion beetles: Protoceratops skeleton insect scavenging versus that by, 510–511 American Journal of Science John Ostrom issue, 9–10 American Museum of Natural History (AMNH), 4; ceratopsian research by, 566; discovery of psittacosaurs and, 21; Protoceratops andrewsi collection at, 5; Protoceratops skeletons collected by, 510, 511; William Cutler with, 542 Ammonites: in Wahweap Formation stratigraphy/ dating, 129 AMNH 5244: described, 76; in Montanoceratops cladistics, 78, 80, 81 AMNH 5464: in Montanoceratops cladistics, 78, 80, 81 Amniotes: immature versus mature bone texture among, 252; sclerotic rings among, 311–312 Amphibians: from Big Bend National Park, 521; caudal fins of, 335; Grand Staircase– Escalante localities and distribution, 119; in Rattlesnake Mountain microsites, 524, 525 Anagenesis: in centrosaurines, 163–165; among ceratopsids, 421 Anaktuval River: Kikak-Tegoseak Quarry and, 459 Analysis: in studies of eye size versus ecology, 312–313, 313 Anaplognathus: bone modification by, 516 Anas: eye size, 312 Anastomosed fluvial systems: KikakTegoseak Quarry representing, 471

Anastomosing splay channel deposits: at Kikak-Tegoseak Quarry, 461–465 Anatomy: ceratopsian, xiii, 221, 222–223, 406, 566 Ancell, Carrie: Sierra Skull Triceratops locality prepared by, 558; Trike II Triceratops elements prepared by, 559 Anchiceratops: aquatic behavior, 199, 200; behavior, 566; in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 284, 286, 287; CMN 8547 previously assigned to, 189–202; CMN 8547 versus, 197, 198, 199; Coahuilaceratops magnacuerna n. gen. & sp. versus, 99, 104, 106, 109, 110–111, 113; Medusaceratops lokii n. gen. & sp. versus, 185; nasal pathologies in, 359; phylogeny, 405, 411; provenance, 431; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216; taxonomy, 408 Anchiceratops longirostris: CMN 8547 possibly referable to, 189; CMN 8547 versus, 191, 198; skull cast with CMN 8547, 191–192, 192 Anchiceratops ornatus: bonebeds, 448, 449; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 444; in CMN 8547 stratigraphy, 197; CMN 8547 versus, 191; distribution, 100; occurrence of, 413; skull, 407; skull and cladogram, 295, 296; skull/bite strength, 303; skull strengths and measurements, 299, 300, 301, 302, 303, 303, 304, 305; stratigraphy, 412 Andersen, Art, xvii, 264 Andrews, Roy Chapman: discovery of psittacosaurs and, 21; Peter Dodson and, 3; Protoceratops andrewsi collection by, 5 Angulars: Archaeoceratops oshimai, 240, 241; Archaeoceratops yujingziensis n. sp., 63, 63; Auroraceratops rugosus, 241, 242; Chaoyangsaurus youngi, 236; Hongshanosaurus houi, 238; Leptoceratops gracilis, 243, 244; Liaoceratops yanzigouensis, 244, 245; measurements of basal ceratopsian, 236; Pachyrhinosaurus n. sp., 150, 151; Protoceratops andrewsi, 246; Psittacosaurus, 239, 240; Tatankaceratops sacrisonorum n. gen. & sp., 210, 211; Yinlong downsi, 237, 237 Angusticanaliculate eggshells: in Rattlesnake Mountain microsites, 524 Anhingids: eye sizes, 312 Animation: in chasmosaurine horn and

frill morphology analysis, 282, 284, 285–290, 285, 286, 287, 288, 289; in modeling locomotion studies, 351–352 Anisotropic bone: in finite element modeling, 269 Ankylosaurids: KBP discoveries of, 479, 481; from Rattlesnake Mountain microsites, 528; species diversity and turnover among, 422 Ankylosaurs: from Aguja Formation, 521, 524, 525, 527–528, 529, 530; in Agujaceratops bonebeds, 520; from Dinosaur Park Formation, 421; semi-aquatic, 329; species diversity and turnover among, 422. See also Nodosaurids/nodosaurines Anolis: studies, 293 Anserines: eye sizes, 312 Antarctica: Late Cretaceous/early Tertiary, 522; Mesozoic paleogeography of, 395, 396, 397 Antelopes: chasmosaurines versus, 283; studies of sympatric extant, 293 Anterolateral maxillary foramen: psittacosaur, 41 Antorbital fenestrae: in ceratopsids, 132– 134; Diabloceratops eatoni n. gen. & sp., 117, 122, 123–124, 123, 132–134; Diabloceratops n. gen., 129, 132–134 Antorbital fossa: psittacosaur, 41 Anurans: Grand Staircase–Escalante localities and distribution, 119 Anxi, 60 Apatosaurus: model at Calgary Zoo, 545 Aperture: eyesight resolution and, 309– 310, 310; in measuring dinosaur eye size allometry, 313, 314, 315, 316, 317, 318; nocturnality and, 320; of Protoceratops eyes, 314, 320, 323 Apodids: eye sizes, 312 Appalachia: in ceratopsid fossil record, 419; ceratopsids and, 414–415 Appendicular elements: Zuniceratops christopheri, 91, 95–96 Appendicular skeleton: Archaeoceratops yujingziensis n. sp., 64–65, 66; Montanoceratops cerorhynchus, 69–70, 75–76 Apteryx: as burrower/digger, 323; nocturnal lifestyle of, 320, 323 Aptian stage: Archaeoceratops yujingziensis n. sp. from, 60; basal ceratopsians from, 387, 390, 429; Bering Land Bridge established before, 395; ceratopsian cladistics and, 400; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440; distribution of ceratopsians during, 389, 391; earliest North American cera-

index

571

Aptian stage (continued) topsians from, 396; Kikak-Tegoseak Quarry palynomorphs from, 461; Mesozoic paleogeography during, 398, 401; neoceratopsians from, 96; psittacosaurs from, 23, 29, 32, 34, 38 Aquatic animals: rarity of Djadokhta, 322– 323 Aquatic behavior: for ceratopsids, 189, 199, 200; for Protoceratops, 309; sclerotic rings and, 329. See also Semi-aquatic behavior Aquatic plants: Prince Creek Formation, 473 Aquatic predators: eye sizes of avian, 312, 314 Aquatic taxa: from Kaiparowits Formation, 481 Aquatic vertebrates: in Big Bend microsites, 533; in Hilda mega-bonebed, 502, 505 Aquilan Land Vertebrate Age (LVA): in Wahweap Formation stratigraphy/ dating, 131 Aquillapollenites decorus: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Aquillapollenites fusiformis: as KikakTegoseak Quarry palynomorph, 461, 462 Aquillapollenites quadrilobatus: as KikakTegoseak Quarry palynomorph, 462 Aquillapollenites quadrilobus: as KikakTegoseak Quarry palynomorph, 461 Aquillapollenites reticulatus: as KikakTegoseak Quarry palynomorph, 461 Aquillapollenites scabridus: as KikakTegoseak Quarry palynomorph, 461 Arboreality: terrestriality versus, 293 Archaeoceratops: as basal ceratopsian, 222, 222; basicranium and palate, 228; basioccipital, 223; basisphenoid, 224, 225; in ceratopsian cladistics, 392, 400; ceratopsian mandibles versus that of, 247; discovery, 12; from Gongpoquan Basin, 59–60, 65; mandible, 240–241, 241; Montanoceratops cerorhynchus versus, 76, 78; in Montanoceratops cladistics, 78, 80, 81; palate, 225; palatines, 228; provenance, 387, 389, 390, 430; Psittacosaurus versus, 52; pterygoids, 226, 227; redescription of, 11, 14; skull and mandible, 234; systematic paleontology, 60; taxonomy, 429; vomers, 230 Archaeoceratops oshimai: Archaeoceratops yujingziensis n. sp. versus, 59, 60, 62, 65, 66–67; basicranium and palate, 228; coronoid/mandible measurements, 248;

572 index

from Gongpoquan Basin, 59; mandible, 240–241, 241; mandibular element measurements for, 236; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; studied specimens, 222, 223, 235 Archaeoceratops yujingziensis n. sp., 59–67, 565; appendicular skeleton, 64–65, 66; axial skeleton, 64, 65; dentition, 61, 63– 64, 63, 64; diagnosis, 60; localities, 60; mandible, 62–63, 63; skull, 60–64, 61; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; systematic paleontology, 60; taxonomy, 65–66, 66–67 Archaeopteryx: eye size versus body mass, 316, 318; John Ostrom’s work on, 4 Archaeopteryx lithographica: eye size and body mass of, 317, 318 Archaeopteryx recurva: eye size and body mass of, 317, 318 Archosaurs: eye sizes among, 311, 315; ontogenetic bone surface texture changes in extant, 252 Arctic Circle: Kikak-Tegoseak Quarry and, 459; in pachyrhinosaur biogeography, 417–418 Arctic Ocean: Mesozoic paleogeography of, 397 Arctic region: Troodon from, 311 Argon/argon dating: Prince Creek Formation, 458; in Wahweap Formation stratigraphy/dating, 130 Argon isotopes: in dating Cedar Mountain Formation, 394 Argon-argon analysis: in dating Kaiparowits formation, 411–413 Arid environments: challenges to faunas in, 323; nocturnal lifestyles in, 308 Arid paleoenvironments: ceratopsids in, 416, 428, 431; Djadokhta Formation as, 322–323, 432–433; Late Cretaceous Big Bend, 521; Protoceratops in, 309, 322– 323, 323–324; Psittacosaurus from, 329 Arizona, 119, 480; Zuniceratops christopheri and, 92, 92 Arrhinoceratops: in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 286–287, 288; and CMN 8547 as Anchiceratops, 198, 199; CMN 8547 possibly referable to, 189; CMN 8547 versus, 197–198, 195; Coahuilaceratops magnacuerna n. gen. & sp. versus, 99, 104, 106, 109, 110–111, 113; humerus, 195; Ojoceratops fowleri n.

gen. & sp. versus, 177; phylogeny, 405, 411; provenance, 431; publication, 198; RFTRA systematics of, 9; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216 Arrhinoceratops brachyops: bonebeds, 449; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 444; in ceratopsid fossil record, 419; in CMN 8547 stratigraphy, 197; CMN 8547 versus, 191, 191; distribution, 100; occurrence of, 413; Ojoceratops fowleri n. gen. & sp. versus, 177; skull, 407; skull/bite strength, 303; stratigraphy, 412 Arrhinoceratops utahensis: Ojoceratops fowleri n. gen. & sp. versus, 177 Articular gout, 349 Articular platform: psittacosaur, 43 Articulars: Archaeoceratops oshimai, 241; Archaeoceratops yujingziensis n. sp., 63, 63; Auroraceratops rugosus, 241, 242; Chaoyangsaurus youngi, 236; Hongshanosaurus houi, 238, 239; Leptoceratops gracilis, 243, 244; Liaoceratops yanzigouensis, 245; measurements of basal ceratopsian, 236; modeling, 294, 296; Pachyrhinosaurus n. sp., 150, 151; Protoceratops andrewsi, 246; Psittacosaurus, 239, 240; Tatankaceratops sacrisonorum n. gen. & sp., 210, 211, 214; Yinlong downsi, 237 Articulated macrosites: Kaiparowits Formation, 480, 482, 483 Articulated specimens: abundance of Kaiparowits, 481 Arundel Formation: basal ceratopsians from, 387, 391, 393, 401 Ash deposits: Kaiparowits Formation, 481 Asia: arid paleoenvironments of, 322–323; basal ceratopsians from, 387, 390, 391, 429–430; basal neoceratopsian taphonmy in, 432–433, 434; basal neoceratopsians from, 69, 430–431; biogeography of, 416; Cedar Mountain fossils versus those of, 394; ceratopsian bonebed taphonomy in, 431–432, 434; ceratopsian bonebeds in, 566; in ceratopsian dispersal, 399, 401; ceratopsian distribution in, 389–392, 398–399; in ceratopsian paleobiogeography, 388, 389; ceratopsian paleoenvironments in, 428; ceratopsian stratigraphy, paleogeography, and diversity in, 430; ceratopsians related to Diabloceratops eatoni n. gen. & sp. from, 117; ceratopsids from, 391; discovery of psittacosaurs in, 21; fossil

insect-modified skeletons in, 518; Mesozoic paleogeography of, 394–396, 395, 396, 397, 398; monophyly of protoceratopsids from, 133–134; Prince Creek Formation and, 457; psittacosaurs from, 23, 26, 329; Psittacosaurus assemblages in, 452; Psittacosaurus from, 328; Psittacosaurus species from, 238; recent basal ceratopsian discoveries in, 221–222 Asiaceratops: in Montanoceratops cladistics, 78, 80, 81; provenance, 431 Asiaceratops salsopaludalis: stratigraphy, paleoenvironmental associations, and taphonomic studies, 440 Asiatic Expeditions: dinosaurs collected by, 320; discovery of psittacosaurs and, 21; Protoceratops andrewsi discovered by, 308, 317 Aspergillus: osteomyelitis and, 348 Assays: standardless, 273 Associated elements: abundance of Kaiparowits, 481 Associated macrosites: Kaiparowits Formation, 480, 482, 483 Astragali: Montanoceratops cerorhynchus, 76 Astragalocalcanea: CMN 8547, 196 Asymmetry: of Pentaceratops postorbital horns, 286, 287 Atlantic Ocean: Mesozoic paleogeography of, 394–396, 395, 396, 397, 401 Atlas: Montanoceratops cerorhynchus, 74 Attachment ridge: psittacosaur, 41 Auks: eye sizes, 312 Auroraceratops, 7; Archaeoceratops yujingziensis n. sp. versus, 66; as basal ceratopsian, 222, 222; basicranium and palate, 228; basisphenoid, 223–224; in ceratopsian cladistics, 400; discovery, 11; mandible, 241–243, 241; Montanoceratops cerorhynchus versus, 73; palatines, 228, 229; provenance, 387, 389, 390, 430–431; pterygoids, 225, 226, 227; skull and mandible, 234; vomers, 230 Auroraceratops rugosus, 66; basicranium and palate, 228; basioccipital, 224; coronoid/mandible measurements, 248; from Gongpoquan Basin, 59; mandible, 241–243, 241; mandibular element measurements for, 236; skull, 13; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; studied specimens, 222, 223, 235 Australasia: Mesozoic paleogeography of, 395, 396, 397 Australia: arid environments of, 322; basal ceratopsians from, 387, 390, 391, 399,

428, 429, 431; ceratopsian paleoenvironmental associations and taphonomy in, 442; Mesozoic paleogeography of, 395, 396, 397; pupation chambers from, 517. See also Serendipaceratops Australian Desert: challenges to faunas in, 323 Australian weevils: pupation chambers, 517 Australopithecus africanus, 8 Autapomorphies: in basal ceratopsian basicranium and palate, 231–232; chasmosaurine, 198; Diabloceratops eatoni n. gen. & sp., 131; Ojoceratops fowleri n. gen. & sp., 169; Prenoceratops sp., 88; psittacosaur, 221; Psittacosaurus, 54; Sweden neoceratopsian, 399 Avaceratops, 12; in ceratopsid cladistic analysis, 110, 112; CPC 279 versus, 103; Medusaceratops lokii n. gen. & sp. versus, 185; from Montana, 181; in Pachyrhinosaurus n. sp. cladistics, 153; phylogeny, 411; provenance, 431; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 214, 215, 216, 216; Tatankaceratops sacrisonorum n. gen. & sp. versus, 203, 205, 208, 209; taxonomy, 429; Zuniceratops christopheri versus, 94 Avaceratops lammersi, 6–7; bonebeds, 449; in ceratopsian cladistics, 409; occurrence of, 414; Peter Dodson’s work on, 7–8; phylogenetic analysis, 8; Rubeosaurus ovatus n. gen. & comb. versus, 159, 160; skeleton, 8; stratigraphy, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; Tatankaceratops sacrisonorum n. gen. & sp. versus, 207 Aves. See Birds Avians. See Birds Avimimus: from Nemegt Formation, 322 Axial skeleton: Archaeoceratops yujingziensis n. sp., 64, 65; Montanoceratops cerorhynchus, 69–70, 74–75, 75 Axis: Montanoceratops cerorhynchus, 74 Axolotl: caudal fin of, 335 Azonia cribrata: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Azonia parva: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Azonia pulchella: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Azonia strictiparva: as Kikak-Tegoseak Quarry palynomorph, 461, 462 ‘‘B. rex’’ site: Hell Creek Formation, 553– 554, 553

Baby dinosaurs: at Afternoon Delight Triceratops locality, 558, 562; in Big Bend microsites, 533; from Big Bend National Park, 521; distinguishing via bone surface texture, 253; Psittacosaurus, 330; Triceratops, 551, 556, 558. See also Hatchlings; Immature individuals; Juveniles Backward burrowing: by Protoceratops andrewsi, 324 Bacterial pathogens: osteomyelitis and, 348 Baculatisporites comaumensis: as KikakTegoseak Quarry palynomorph, 462 Bader, Kenneth, xvii, 509 Badlands: Eoceratops collected in, 542; Hilda mega-bonebed in, 496; Rattlesnake Mountain microsites in, 524, 525 Bagaceratops: as basal ceratopsian, 222; in ceratopsian cladistics, 392, 400; in ceratopsian family tree, 134; in ceratopsian paleobiogeography, 398; choanae, 231; Diabloceratops eatoni n. gen. & sp. versus, 117; Diabloceratops n. gen. versus, 132– 133; eyes, 314; Montanoceratops cerorhynchus versus, 72; in Montanoceratops cladistics, 78, 80, 81; palatines, 229; provenance, 387, 389, 390, 431; RFTRA systematics of, 9; vomers, 230; Zuniceratops christopheri versus, 96 Bagaceratops rozhdestvenskyi: discovery/ description, 231; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440 Bainoceratops: collected from Bayn Dzak, 322; provenance, 431 Bainoceratops efremovi: stratigraphy, paleoenvironmental associations, and taphonomic studies, 441 Baja Peninsula ceratopsian: biogeography, 111 Bajocian stage: ceratopsian cladistics and, 400 Bakker, Robert T.: on ceratopsian posture, 10; Peter Dodson and, 4, 7 Ballast: gastroliths as, 333 Bangladesh: Dinosaur Park Formation paleoenvironment like, 499, 534 Barkas, Vaia, xvii, 495 Barlowe, Wayne: Peter Dodson and, 10 Barremian stage: basal ceratopsians from, 387, 390; ceratopsian cladistics and, 400; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440; distribution of ceratopsians during, 389; Mesozoic paleogeography during, 398; neoceratopsian bonebeds from, 433; psittacosaurs from, 22–23, 28

index

573

Barriers: to ceratopsids, 414, 416–417 Barrow: Kikak-Tegoseak Quarry and, 459 Barun Goyot Formation: ceratopsian paleoenvironmental associations and taphonomy in, 440, 441; ceratopsians from, 132; dinosaurs from, 322, 322 Baryonyx: as semi-aquatic, 329 Basal ceratopsians: Archaeoceratops yujingziensis n. sp. among, 59, 60; basicranial and palatal anatomy, 221– 233; in ceratopsian paleobiogeography, 388; in ceratopsid fossil record, 418, 419; Chaoyangsaurus youngi versus, 235–236; from China, 11, 12; choanae, 231; distinctive cranial morphology, 221, 222; distribution, 66, 389; earliest known, 387, 389, 428; eyes of, 314; forelimb movement of, 332–333; functions of cranial ornamentation among, 282–283; horns and taxonomy of, 418; Leptoceratops gracilis versus, 243; Liaoceratops yanzigouensis versus, 245; mandibular anatomy and evolution, 234–250; mandibular element measurements for, 236; Montanoceratops cerorhynchus among, 68, 69, 70, 73, 74–75; North American, 181; from Oldman Formation, 83–90; origin and evolution of horns and frills among, 290; paleoenvironmental associations of, 429–430, 430–431; phylogenetic analysis, 221, 222–223, 222; Prenoceratops sp. among, 85; Psittacosaurus versus, 238; recent discoveries in China, 221– 222, 231; RFTRA systematics of, 9–10, 9; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; skulls and cladogram, 295, 296, 296–298, 299; species diversity and turnover among, 421; stratigraphy, paleogeography, and diversity of, 430; studied specimens, 222– 223; table of genera and species, stratigraphy, paleoenvironmental associations, and taphonomic studies, 439–440; taxonomy, 429; from Xinminpu Group, Mazongshan, 59–60 Basal ceratopsids: osteomyelitis in, 347 Basal neoceratopsians. See Basal ceratopsians; Neoceratopsians Basal plate: psittacosaur, 42 Basal sandstone member: in Aguja Formation geology, 523 Basicrania: basal ceratopsian, 221, 222– 223, 223–225, 224, 225, 226, 227, 228, 231–232, 231 Basilemys: in Rattlesnake Mountain microsites, 524, 525

574 index

Basioccipitals: basal ceratopsian, 221, 223, 224, 225; Diabloceratops eatoni n. gen. & sp., 127 Basipterygoids: basal ceratopsian, 221. See also Basisphenoids Basisphenoids: basal ceratopsian, 221, 223–225; Diabloceratops eatoni n. gen. & sp., 125, 127 Bathonian stage: ceratopsian cladistics and, 400 Battle Formation: in ceratopsid stratigraphy, 412; in CMN 8547 stratigraphy, 191; geologic correlation, 84 Bayan Gobi Formation: ceratopsian paleoenvironmental associations and taphonomy in, 439; psittacosaurs from, 26 Bayan Mandahu, 510; ceratopsian paleoenvironmental associations and taphonomy in, 441; ceratopsians from, 132; Protoceratops from, 308, 317, 321–322, 322 Bayan Zag. See Bayn Zag (Bayn Dzak) Bayn Dzak Member: Protoceratops from, 511 Bayn Shiree Formation: ceratopsian paleoenvironmental associations and taphonomy in, 441; dinosaurs from, 322, 322 Bayn Zag (Bayn Dzak), 510; ceratopsian paleoenvironmental associations and taphonomy in, 441, 442; dinosaur expeditions to, 509; dinosaurs from, 320, 322; Protoceratops andrewsi collection from, 5, 308, 316–317, 321; Protoceratops from, 511 Beaks: ceratopsian, 232, 377 Bearpaw Formation: in ceratopsid stratigraphy, 412; geologic correlation, 84; Hilda mega-bonebed and, 497, 497; pachyrhinosaur evolution and, 408; Pachyrhinosaurus n. sp. and, 152; paleoenvironments of, 415, 416; Prenoceratops sp. and, 88; Rubeosaurus ovatus n. gen. & comb. and, 161; stratigraphy, 143, 163 Bearpaw Sea: centrosaurine evolution and, 165; Pachyrhinosaurus n. sp. and, 151– 152; Prenoceratops sp. and, 85, 88 Bearpaw transgression: ceratopsids and, 414, 416, 422; pachyrhinosaur evolution and, 408 Beetles: bone modification by, 515, 516, 517–518; carrion-eating, 433; chicken humerus damage by, 517; in KikakTegoseak Quarry fossils, 470, 472, 473; pupation chambers, 513–515, 513, 514, 515

Behavior: centrosaurine mega-bonebed and, 495–508; centrosaurine versus chasmosaurine, 431; ceratopsian, xiii, 566; ceratopsian bonebeds and, 447– 455; ceratopsid paleopathologies and, 355–384; chasmosaurine, 189, 199, 200; chasmosaurine locomotor, 340–354; dinocephalian, 356; Hilda megabonebed and ceratopsid, 495, 497, 502, 505; horns and ceratopsian, 271–272; Kaiparowits ceratopsids, 491, 492; paleoenvironments and ceratopsian, 428; psittacosaur taphonomy and, 432, 434; relative eye size and, 310–311; of sympatric ceratopsians, 293, 294; Zuniceratops christopheri, 93. See also Bipedality; Brooding behavior; Burrowing; Defense; Diurnal entries; Functional biology; Gregariousness; Group behavior; Head butting; Herding behavior; Interspecies recognition; Interspecific competition; Intraspecies recognition; Intraspecific entries; Mating entries; Migration/migrations; Nesting; Nocturnal entries; Semi-aquatic behavior; Social behavior; Sparring Behrensmeyer, Kay: Peter Dodson and, 7 Beijing, 510; psittacosaurs at, 22 Beipiao: psittacosaurs from, 29 Belgium neoceratopsian, 390, 391, 392, 393, 399. See also Craspedodon entries Belly River Group: basal neoceratopsians from, 83–90; ceratopsid bonebeds in, 496; geologic correlation, 84; Hilda mega-bonebed and, 497, 497, 499 Bentonite, 554, 557, 560 Bentonite dates: Pachyrhinosaurus n. sp. and, 141, 143 Bering Land Bridge: ceratopsian dispersal across, 387, 388, 391, 394, 399–401; Mesozoic paleogeography and, 394, 395, 396, 398 Berriasian stage: basal ceratopsians from, 390; ceratopsian cladistics and, 400; Kikak-Tegoseak Quarry palynomorphs from, 461 Betonnie Tsosie Wash, 170, 170; Ojoceratops n. gen. from, 176 Betonnie Tsosie Wash individual (Ojoceratops n. gen.), 176 Betulapollenites: as Kikak-Tegoseak Quarry palynomorph, 462 Biases: in bonebed studies, 447; in dinosaur paleobiogeography, 389; historical, in collecting Montana Triceratops, 551– 563

Big Bend National Park, 520, 521; Agujaceratops bonebeds in, 520–537; ceratopsian bonebeds in, 186, 451; Ojoceratops fowleri n. gen. & sp. and, 178 Big Valley, Alberta, 190 Big Water, 119 Binocular vision: Protoceratops, 308, 315, 319, 320–321, 323 Biocoenoses: of neoceratopsian bonebeds, 433 Biogeography: ceratopsian, xiii, 387–404, 566; ceratopsid, 405–427; Mesozoic, 416. See also Paleobiogeography Biological features: of Kikak-Tegoseak Quarry floodplain paleosols, 468 Biomechanics: finite element modeling in, 265; Protoceratops eyes, 308–327, 309, 310, 319; psittacosaur, 329; of skulls, horns, frills, and jaws, 294–296, 296– 298, 297, 298–300, 299, 300–305, 301, 303, 304 Biostratigraphy: Agujaceratops bonebeds, 520; Alberta ceratopsians and, 152–153; Big Bend National Park, 521; ceratopsian, 566; ceratopsid bonebeds, 375; Late Cretaceous New Mexico, 170–171, 177; Medusaceratops lokii n. gen. & sp., 187; Tatankaceratops sacrisonorum n. gen. & sp., 205 Bioturbated groundmass: at KikakTegoseak Quarry, 466, 468 Bioturbation: Kaiparowits Formation, 481. See also Trampling Bipedality: pedes and, 331–332; psittacosaur, 331 Birds: in diagnosing dinosaur pathologies, 341; eye sizes, 311, 312; eyes, 308; gastroliths found with, 333; gout in, 349; long-grained bone surface texture in, 258; nocturnal lifestyles among, 317, 320, 323; olfactory bulbs, 323; osteomyelitic infections in, 348; osteomyelitis in, 348; in Rattlesnake Mountain microsites, 524, 525; relative aperture size among, 320, 321; sclerotic rings and eye sizes among, 312, 313, 314, 315, 316; visual acuity of, 309–310, 313–314 Birds of prey: studies of sympatric extant, 293 Birefringence fabric (b-fabric): at KikakTegoseak Quarry, 466, 468 Biretisporites: as Kikak-Tegoseak Quarry palynomorph, 462 Bison bison, 204; Pachyrhinosaurus versus, 377

Bisti/De-na-zin Wilderness, 170; Ojoceratops n. gen. from, 176 Bite forces: in ceratopsian skulls, 294–296, 296–298, 297, 298–305, 298, 299, 301, 304 Bite marks: on Eotriceratops horn, 359; on Hilda mega-bonebed fossils, 495; on Kaiparowits ceratopsid B bones, 487; on Pachyrhinosaurus n. sp. (Alaska) bone, 472 Black, David E.: Calgary Public Museum collection and, 544 Black Hills Institute (BHI): Tatankaceratops sacrisonorum n. gen. & sp. at, 204–205 Black Peaks Formation: lithostrigraphy, 523 Black shore skink: nocturnal lifestyle of, 320 Blakey, Ron C.: Mesozoic paleogeography reconstructions by, 394–395, 395, 396, 397 Blastomycosis: osteomyelitis and, 348 Blood Reserve Formation: geologic correlation, 84 Blood vessels: in thermoregulatory function of horns, 280; vascular bone and, 258 Blows, William T.: report on British Museum Cutler documents, 547–549; William Cutler’s Eoceratops and, 546 Blues Ceratopsian, 485. See also Kaiparowits ceratopsid B; UMNH VP 12198 Boas: nocturnal, 320 Body mass: eye size versus, 312–313, 313, 315, 316, 317, 318 Body proportions: psittacosaur, 333, 335 Body size: ceratopsid, 418; eyes and, 308; relevance to ontogeny of fossil vertebrates, 252–253; among sympatric animals, 293 Boiga: nocturnal lifestyle of, 320 Bone: preparing thin sections of, 253–254 Bone concentration: in Hilda megabonebed, 502–503, 503, 504, 505–506 Bone cysts: diagnosing in fossils, 349 Bone damage: insect-caused, 509–520, 512, 513, 514, 515, 516, 517 Bone growth: surface textures associated with, 258 Bone growths: on ceratopsian phalanges, 374, 374; on ceratopsian vertebrae, 366, 366, 367. See also Bone spurs Bone infections: among ceratopsians, 355 Bone injuries: among ceratopsians, 355, 374–379 Bone lesions: diagnosing, 340–341;

osteomyelitic infections and, 348; pathology in Chasmosaurus irvinensis and, 343, 344 Bone microstructure: relevance to ontogeny of fossil vertebrates, 252 Bone phosphate studies: thermoregulation demonstrated via, 272 Bone remodeling, 255, 257, 258, 259–260 Bone resorption, 255, 258–259 Bone spurs: osteomyelitic infections and, 348. See also Bone growths Bone structure: in diagnosing dinosaur pathologies, 341 Bone surface texture: within bonebeds, 258–259; in diagnosing paleopathologies, 346; histology of ontogenetic changes in ceratopsian, 251–263; hyperostotic, 346; immature versus mature, 252; pathology in Chasmosaurus belli and, 345–346; pathology in Chasmosaurus irvinensis and, 343, 344; relevance to ontogeny of fossil vertebrates, 252 Bonebeds: Agujaceratops, 520–537, 522, 523, 526, 527; Alberta ceratopsid, 495– 496; in Big Bend microsites, 533; in Big Bend National Park geology, 521, 522; Big Bend versus Dinosaur Park, 533–534; bone surface textures found within, 258–259; Centrosaurus, 566; Centrosaurus apertus, 413; ceratopsian, 6, 566; ceratopsian behavior and, 447–455; ceratopsian cranial material from, 185–187; ceratopsian paleoenvironments and, 428; ceratopsid caudal pathology specimens from, 368; ceratopsid migrations and, 417; ceratopsid rib pathology specimens from, 367; in Cerro del Pueblo Formation, 102; comprising Hilda megabonebed, 495, 496, 498, 499, 500, 501; depositional environments of, 447, 448– 451, 452; in Dinosaur Provincial Park, 377; evidence for agonistic ceratopsid behavior from, 374–376; future work on, 453; high-latitude, 456–477, 458, 459; Hypacrosaurus, 161; Kaiparowits ceratopsid B, 486, 487–488; in Kaiparowits Formation, 478, 480, 482, 483–484; in Kika-Tegoseak Quarry, 456–477, 458, 459; locations of North American, 450; Medusaceratops lokii n. gen. & sp., 181– 182, 183, 183, 185, 186, 187; minimum numbers of individuals in, 450; Pachyrhinosaurus, 141–142; pathological specimens from, 355, 356; pathological squamosals from, 361–362; Pipestone

index

575

Bonebeds (continued) Creek Pachyrhinosaurus, 357; social behaviors inferred from, 451–452, 452– 453; taphonomic patterns in ceratopsian, 431–432, 433, 434; taxonomic and geographic distributions of ceratopsid, 448, 449, 450; in Two Medicine Formation, 156; Zuniceratops christopheri, 91, 92–93, 92, 95–96. See also Megabonebeds Borings: in ornithischian bone, 515, 516, 517; in Pinacosaurus skull, 512; in Protoceratops skeletons, 509, 512, 513–515, 513, 514, 515–517, 515, 517–518 Botryococcus: as Kikak-Tegoseak Quarry palynomorph, 462 Boulder, Colorado, 119 Bovids: chasmosaurines versus, 283, 284; studies of sympatric extant, 293; thermoregulatory function of horns, 280; Triceratops versus, 288 Brachyceratops: caudal pathologies in, 372; MOR 449 versus, 162; stratigraphy, 163 Brachyceratops montanensis: bonebeds, 449, 452; as nomen dubium, 156, 161; Rubeosaurus ovatus n. gen. & comb. versus, 161; from Two Medicine Formation, 156 Brachychampsa: in Rattlesnake Mountain microsites, 525 Brain thermoregulation: horns and ceratopsian, 272, 280 Braincases: Diabloceratops eatoni n. gen. & sp., 125, 126–127, 126; Diabloceratops n. gen., 129, 130; Montanoceratops cerorhynchus, 69, 76; neonate psittacosaur, 51; psittacosaur, 42; Psittacosaurus major, 30; Shenandoah University Triceratops, 273, 278, 279, 280; Zuniceratops christopheri, 91, 93, 94, 95 Brandlen, Erik, xvii, 456 Branta: immature versus mature bone surface texture in, 253 Breakage: of Hilda mega-bonebed fossils, 502–503, 503, 504, 505; of Kaiparowits ceratopsid B bones, 487 Breeding seasons: bonebed formation during, 451, 452 Breviceratops: Diabloceratops n. gen. versus, 132; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440 Bridges: in ceratopsid fossil record, 418– 419; Prince Creek Formation as, 457 Brinkman, Donald B., xiii, xvii, xxi, 141, 495

576 index

Bristle-like integumentary structures: psittacosaur, 335, 337; Psittacosaurus, 328, 329 British Columbia: Hilda mega-bonebed and, 497 British Museum of Natural History (BMNH): Chasmosaurus skeleton at, 546, 547, 549; correspondence from William Cutler to, 547–549; William Cutler with, 542; William Cutler’s Eoceratops at, 546– 547 Britt, Brooks, xxi Brontosaurus, 544 Brooding behavior: relevance to ontogeny of fossil vertebrates, 252 Brooks, Alberta, 190 Brooks Range: Kikak-Tegoseak Quarry and, 459, 473; Prince Creek Formation and, 458 Brown, Barnum, 565; ceratopsian research by, 566; on finding Triceratops in the field, 555; juvenile and subadult Triceratops collected by, 556; on New Mexico ceratopsians, 169; Triceratops studied by, 551; William Cutler and, 542, 544 Brownie Butte: Triceratops skull and, 272, 273 Buccal emargination: psittacosaur, 41 Buenos Aires: William Cutler’s Eoceratops skeleton in, 542 Buffalo: Pachyrhinosaurus versus, 377 Buffalo Lake: Montanoceratops cerorhynchus from, 70 Buffalo, South Dakota: Tatankaceratops sacrisonorum n. gen. & sp. from, 203, 204– 205 Bug Creek: Russell Basin Triceratops locality along, 560 Bureau of Land Management (BLM): Kaiparowits Formation joint survey with UMNH, 478–479; San Juan Basin survey by, 169–170; Wahweap Formation explorations and, 118 Burhinus: eye size, 312; nocturnal lifestyle of, 320 Burial. See Taphonomy Burke Museum: studied specimens at, 312 Burrowing: in arid envirnments, 323– 324; by Protoceratops, 323–324, 432– 433, 434 Burrows: with Fox Protoceratops, 513; insect, 510, 511, 516, 517–518; morphotypes of Djadokhta, 511; with Pinacosaurus skull, 512 Buteo jamaicensis: sclerotic rings, 311

Caanan Peak Formation: in ceratopsid stratigraphy, 412 Calgary, 190; Calgary Public Museum collection and, 544, 545; Eoceratops stored in, 541; letters from Cutler to Woodward from, 548; Medusaceratops lokii n. gen. & sp. material in, 182; William Cutler in, 542; William Cutler’s Eoceratops skeleton in, 542 Calgary Allied Arts Council for the Allied Art Centre: Calgary Public Museum collection at, 544 Calgary Public Library: Calgary Public Museum collection and, 546 Calgary Public Museum: alternative names of, 543; William Cutler and, 542; William Cutler’s fossil collection and, 542, 543–546, 546–547 Calgary Syndicate for Prehistoric Research: William Cutler with, 542, 543–544 Calgary Zoo: Calgary Public Museum collection and, 544–545, 545–546, 547; William Cutler’s Eoceratops skeleton at, 542 Calgary Zoological Society: Calgary Public Museum collection and, 545–546 Calhoun, Alexander: William Cutler and, 543 Caliches: table of ceratopsian, 439–446 Callistopollenites radiostriatus: as KikakTegoseak Quarry palynomorph, 461, 462 Callovian stage: ceratopsian cladistics and, 400 Camarasaurids: in Mesozoic paleogeography, 398 Camarasaurus lentus: eye size and body mass of, 317, 318 Campanian/Maastrichtian boundary: in Aguja Formation geology, 523, 524; Prince Creek Formation and, 458, 460 Campanian stage: Agujaceratops bonebeds from, 520–537; Albertaceratops nesmoi from, 407–408; basal ceratopsians from, 390, 391, 392, 429; in Big Bend National Park geology, 521, 523; biogeography during, 416, 417; bonebeds from, 456– 477, 458, 459; centrosaurine evolution during, 163–165; centrosaurine megabonebed from, 495–508, 496, 497, 498, 499, 500, 503, 504; centrosaurine squamosals from, 135; Centrosaurus brinkmani from, 408; ceratopsian bonebeds from, 356; ceratopsian cladistics and, 400, 409, 410; ceratopsian distribution during, 100, 110, 111; ceratopsian family tree and, 134; ceratopsian paleo-

environmental associations and taphonomy in, 440, 441, 442–444, 445; ceratopsian paleoenvironments during, 428; ceratopsid cladistic analysis and, 110; ceratopsid evolution during, 406; in ceratopsid fossil record, 418, 419; ceratopsid occurrences in, 413; ceratopsid species diversity and turnover during, 419–422, 422–423; ceratopsid stratigraphy during, 412; chasmosaurines from, 408–411; in CMN 8547 stratigraphy, 191; Coahuilaceratops magnacuerna n. gen. & sp. from, 104; CPC 278 from, 102; CPC 279 from, 103; dating formations of, 413; Diabloceratops eatoni n. gen. & sp. from, 117–140; earliest known ceratopsid from, 413, 414; geologic correlation of western North American, 84; insect trace fossils with Protoceratops skeletons from, 509–520, 512, 513, 514, 515, 516, 517; KikakTegoseak Quarry palynomorphs from, 461; Medusaceratops lokii n. gen. & sp. from, 181; Mexican ceratopsians from, 99–116; in Mexican stratigraphy, 100, 101; Montana ceratopsians from, 181– 182; Montanoceratops cerorhynchus from, 68, 70; neoceratopsians from, 96, 431, 434; North American basal neoceratopsians from, 88; North American ceratopsids during, 414, 415, 416; North American paleogeography during, 415; in North Slope stratigraphy, 460; pachyrhinosaurs from, 408; Pachyrhinosaurus n. sp. from, 141–155; Prenoceratops sp. from, 83; Protoceratops from, 321; Rattlesnake Mountain microsites and, 525; taphonomy of Kaiparowits ceratopsids from, 478–494; Two Medicine Formation stratigraphy and, 163; Wahweap centrosaurine from, 408; in Wahweap Formation stratigraphy/dating, 129– 131 Canaan Peak formation: ceratopsian family tree and, 134 Canada, 183; Albertaceratops from, 181; basal ceratopsians from, 390; basal neoceratopsians from, 83–90; bonebeds, 448, 449, 450, 451; centrosaurine megabonebed from, 495–508, 496, 497, 498, 499, 500, 503, 504; ceratopsian bonebed taphonomy in, 432; ceratopsian paleoenvironments in, 428, 431, 434; ceratopsians from, xiv, 428–429, 566; dinosaur discoveries, 555; Medusaceratops lokii n. gen. & sp. material in, 182a;

Pachyrhinosaurus n. sp. from, 141–155; paleopathologies in ceratopsids from, 355–384; Peter Dodson’s work in, 7; sympatric ceratopsians from, 300–303, 303; Triceratops from, 551; Triceratops studies in, 552 Canada Fossils, Ltd.: Medusaceratops lokii n. gen. & sp. at, 182 Canadian Arctic: Peter Dodson’s fieldwork in, 4 Canadian Cordillera: Hilda mega-bonebed and, 499 Canadian Museum of Nature (CMN). See CMN 8457; National Museum of Canada Canadian Pacific Railway Company, 542, 544 Caniniform teeth: Protoceratops, 315 Canis latrans: femoral fracture in, 373 Cannonball Sea: ceratopsids and, 414 Cantilevered beams: skull anatomy as, 294 Cañon del Tule Formation: in ceratopsid stratigraphy, 412; stratigraphy, 101 Cape buffalo: chasmosaurines versus, 283 Capillaries: vascular bone and, 258 Capitol Reef National Park, 119 Capra: injuries from play behavior in, 375 Caprimulgiforms: eye sizes among, 311, 312; nocturnal lifestyle of, 320 Capybara: nares and orbits of, 334 Carbon, Alberta: William Cutler at, 541– 542, 543 Carbonaceous mudstone lithofacies: Kaiparowits Formation, 480, 480, 481, 482 Carbonaceous siltstone/claystone facies: of Hilda mega-bonebed, 500, 500 Carbonate soil nodules: in Big Bend microsites, 533; in ‘‘Purple Hill’’ section, 525, 527 Careless Creek bonebed, 6 Careless Creek Ranch: ceratopsids from, 7–8 Cariama: eye size, 312 Carnivore scavenging: of Kaiparowits ceratopsid B bones, 487 Carnivores: biogeography of, 418; Nemegt and Djadokhta, 322; Protoceratops as, 317; studies of sympatric extant, 293. See also Predators Carnivorous theropods: dentition, 316 Carnosaurs: in Mesozoic paleogeography, 398 Carotid rete: in thermoregulatory function of horns, 280 Carpals: CMN 8547, 195, 197 Carpenter, Kenneth, xxi Catastrophic death events: bonebed for-

mation during, 452. See also Death events Cathartids: eye sizes, 312 Caudal fins: psittacosaur, 335, 337 Caudal neural spines: Protoceratops, 323; Psittacosaurus, 334. See also Tails Caudals: Archaeoceratops yujingziensis n. sp., 64, 65; CMN 8547, 192, 193, 193, 194, 198, 199; isotopic analysis of, 377– 379; Montanoceratops cerorhynchus, 70, 75, 77; pathologies in ceratopsid, 368– 372, 370, 371; pathologies in hadrosaur, 363; protoceratopsian, 329; of semiaquatic ceratopsids, 199; Zuniceratops christopheri, 95, 96 CD-ROM: included with this book, xiv, 565–566 Cedar Mountain Formation: argon isotope dating of, 394; in Mesozoic paleogeography, 396–397, 399 Cedarosaurus weiskopfae: in Mesozoic paleogeography, 398 Cedripites canadensis: as Kikak-Tegoseak Quarry palynomorph, 461 Cenomanian stage: basal ceratopsians from, 390; ceratopsian cladistics and, 400; ceratopsian family tree and, 134; in dinosaur paleobiogeography, 389; neoceratopsians from, 96 Cenozoic ecosystems, 317 Central Asia: arid paleoenvironments of, 322–323 Central Asiatic Expedition. See Asiatic Expeditions Centrosaurine bonebeds. See Bonebeds Centrosaurines: from Alberta and Montana, 152; Alberta mega-bonebed, 495– 508, 503, 504s; anatomical features, 406; Avaceratops lammersi as basal, 8; biogeography, 111, 417–418; bone growth and resorption in, 258–259; bonebeds, 186, 447, 448, 449, 450, 456–477, 458, 459, 495; bonebeds in Alberta, 356; in ceratopsian biostratigraphy, 152; in ceratopsian cladistics, 409, 410; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 109–110, 110; in ceratopsid fossil record, 418; ceratopsid species diversity and turnover and, 419, 420, 422; from Cerro del Pueblo Formation, 100, 102, 103, 103; characters uniting subclades of, 164; chasmosaurine intraspecific interaction analysis and, 289; CMN 8547 versus, 196–197; Coahuilaceratops magnacuerna n. gen. & sp. versus, 109–110; CPC 279 versus, 103,

index

577

Centrosaurines (continued) 109–110; depositional bonebed environments for, 448–451, 452–453; Diabloceratops eatoni n. gen. & sp. versus, 117, 118, 120, 123, 125, 126, 127, 128, 131, 135–136; distribution, 99–100; diversity and taxonomy of, 406, 407–411, 409; early, 96; evolution of Two Medicine, 163–165; frills, 132; frills as sexual display structures in, 6; genera included among, 181, 442–444; Grand Staircase– Escalante localities and distribution, 119; histology of ontogenetic bone surface texture changes in, 251–263; from Horseshoe Canyon Formation, 196–197; immature versus mature bone surface texture among, 252–253, 258; indeterminate, 183, 186; jugal/epijugal pathologies in, 359–360, 360, 361; Kaiparowits ceratopsid C among, 489; KBP discoveries of, 479, 481; at Kikak-Tegoseak Quarry, 467; life habits, 428; longgrained and mottled bone surface texture in, 258–259; in Mansfield Bonebed, 185–186; marginal frill ossifications, 102; Medusaceratops lokii n. gen. & sp. versus, 185; Medusaceratops n. gen. versus, 182; Montanoceratops cerorhynchus versus, 68; from Mexico, 99, 100– 102, 417; MOR 449 versus, 161; occurrences, 141, 413–414; ontogeny, 470– 471; in Pachyrhinosaurus n. sp. cladistics, 153; Pachyrhinosaurus n. sp. versus, 141, 150, 151, 152, 153–154; paleoenvironmental associations of, 431; parietal pathologies in, 362–364, 362, 363; pathological skuls of, 356; Peter Dodson’s studies of, 9; phalangeal stress fractures in, 373–374, 374; phylogeny, 411; RFTRA systematics of, 9–10, 9; rib paleopathologies among, 376–377; Rubeosaurus ovatus n. gen. & comb. among, 156, 157, 158; sexual selection among, 165; similarities and sympatry among, 293–294; skull roof pathologies in, 359; skull shapes and niche partitioning by sympatric, 293–307; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; skulls, 406; skulls and cladogram, 295, 296, 296– 298, 298, 299; sparring among, 375, 376–377; species diversity, 405, 406, 407–411, 409, 418–419; squamosal pathologies in, 360–362; stratigraphy, paleogeography, and diversity of, 430; in studies of ceratopsian sympatry, 294;

578 index

synapomorphies, 160; table of genera and species, stratigraphy, paleoenvironmental associations, and taphonomic studies, 442–444; taphonomy, 433–434; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 213, 214, 216–217, 216; taxonomy, 406, 429; from Two Medicine Formation, 156; Zuniceratops christopheri versus, 96 Centrosaurus: acromegaly in, 356; behavior, 566; bone histology, 259; bonebeds, 356, 451–452, 505, 566; caudal pathologies in, 369, 370, 371; in centrosaurine cladistics, 164; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; ceratopsid species diversity and turnover and, 419, 422; cervical pathologies in, 366; eye size, 313; fibular pathologies in, 372, 373; frill, 132, 132; frills as sexual display structures in, 6; immature versus mature bone surface texture in, 252; mandibular and dental pathologies in, 364–365, 364; Medusaceratops lokii n. gen. & sp. versus, 185; modeling frill of, 269; in Montanoceratops cladistics, 78, 80, 81; nasal pathologies in, 357, 359; osteopathies of, 355; osteopathy in bonebed specimens, 377; paleopathologies in bonebed specimens, 376; parietal pathologies in, 362, 363, 362; pathological cervicals of, 356; pathological jugals/ epijugals of, 360, 360; pathological parietal of, 355–356; pathological squamosal of, 356; pelvic pathologies in, 372; Peter Dodson’s systematics of, 9; phalangeal stress fractures in, 373–374; phylogeny, 405, 411; postorbital pathologies in, 360; Prenoceratops sp. versus, 85; provenance, 431; RFTRA systematics of, 9; rib pathologies in, 367, 369; Rubeosaurus ovatus n. gen. & comb. versus, 157, 160; severe fractures in, 375; skull excavated by William Cutler, 547; squamosal pathologies in, 361; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216; taxonomy, 429; Triceratops versus, 264; ulnar pathologies in, 372 Centrosaurus apertus: bonebeds, 186, 377, 448, 449–450, 450, 451, 452, 496, 566; in centrosaurine cladistics, 164; in ceratopsian biostratigraphy, 152; in ceratopsian cladistics, 409; ceratopsid species diversity and turnover and, 419; CMN 8547 versus, 192, 193, 194, 195, 195, 196, 198; Diabloceratops eatoni n.

gen. & sp. versus, 135; eye size and body mass of, 317, 318; eyes, 314; Hilda megabonebed, 495–508, 496, 497, 498, 499, 500, 503, 504; histology of ontogenetic bone surface texture changes in, 251– 263; humerus, 195; immature versus mature bone surface texture in, 253, 258, 259–260; manus, 343; occurrence of, 413–414; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153–154; Peter Dodson’s systematics of, 9; Peter Dodson’s work on, 8; range/duration of, 413, 414; Rubeosaurus ovatus n. gen. & comb. versus, 158; Rubeosaurus ovatus n. gen. & comb. versus, 159, 161, 161, 163, 166; skull, 309, 406; skull and cladogram, 295, 296, 298, 298; skull strengths and measurements, 298, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 303, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; in studies of ceratopsian sympatry, 296, 298 Centrosaurus brinkmani: bonebeds, 377, 448, 449, 450, 450, 566; in centrosaurine cladistics, 164; in ceratopsian cladistics, 409; in ceratopsid fossil record, 419; Medusaceratops lokii n. gen. & sp. versus, 183; occurrence of, 414; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153– 154; Rubeosaurus ovatus n. gen. & comb. versus, 159, 161, 161, 166; skull, 406; skull and cladogram, 295, 296, 298; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; taxonomy and distribution, 408 Centrosaurus dawsoni: Peter Dodson’s systematics of, 9 Centrosaurus flexus: Diabloceratops eatoni n. gen. & sp. versus, 135; Peter Dodson’s systematics of, 9 Centrosaurus longirostris: Peter Dodson’s systematics of, 9 Centrosaurus nasicornus: Diabloceratops eatoni n. gen. & sp. versus, 127, 135; Pachyrhinosaurus n. sp. versus, 146; Peter Dodson’s systematics of, 9 Cerasinops, 83; in ceratopsian cladistics, 392, 400; in ceratopsian paleobiogeography, 398; from Montana, 181; in Montanoceratops cladistics, 78, 80, 81; Montanoceratops cerorhynchus versus, 68, 70, 71, 72, 74, 76, 77, 78, 79; Montanoceratops versus, 70; Prenoceratops sp. versus,

87, 88; provenance, 387, 390, 391, 431; skull and mandible, 234; taxonomy, 429; from Two Medicine Formation, 88 Cerasinops hodgkissi, 565; Diabloceratops eatoni n. gen. & sp. and, 118; stratigraphy, paleoenvironmental associations, and taphonomic studies, 441 Ceratohyals: psittacosaur, 44 Ceratophaga: bone modification by, 516 Ceratophile, 566; defined, 3 Ceratophilia: Peter Dodson on, 3–17 Ceratops: New Mexico frill material versus, 169 ‘‘Ceratops’’ beds: Triceratops from, 551 Ceratops montanus: bonebeds, 187; from Montana, 182; Medusaceratops lokii n. gen. & sp. versus, 182 Ceratopsia, 25; early systematics, 10. See also Basal ceratopsians; Ceratopsians Ceratopsia monograph (Marsh, Hatcher & Lull), 565 Ceratopsian Symposium, xiii, 565; ceratopsian genera and species named since, 565 Ceratopsians: Aguja Formation microsite, 530; in Agujaceratops bonebeds, 520; anatomy, 566; Archaeoceratops oshimai versus, 240; Archaeoceratops yujingziensis n. sp.among, 60; from Big Bend National Park, 521, 525; in Calgary Public Museum collection, 546; ceratopsid evolution among, 421; chronostratigraphy of, 390; CMN 8547 among, 190; cornual sinuses among, 279–280; CPC 278 among, 102; Cretaceous radiation of, 565; current research, 566; dental batteries, 234; dentition, 316; Diabloceratops eatoni n. gen. & sp. among, 120, 131; discovery of and research on, 406–407; dispersal events among, 387, 388; distribution of, 389–392, 398–399; earliest known basal, 387, 428; evolutionary interactions between horn and frill morphology among, 282–292; family tree, 134; finite element modeling of, 265; functions of cranial ornamentation among, 282–283, 289–290; future research, 566; high-latitude bonebed, 456–477, 458, 459; histology of ontogenetic bone surface texture changes in, 251–263; history of research on, 565; horns and frills in sparring among, 282– 283, 283–290, 285, 286, 287, 288, 289; at Kikak-Tegoseak Quarry, 467, 469; Leptoceratops gracilis versus, 243; literature on, 429; mandibular evolution, 234,

246–248; Medusaceratops lokii n. gen. & sp. among, 182; Mesozoic paleogeography and, 394–396; Montanoceratops cerorhynchus among, 68, 69; nasal functions, 232; new genera and species of, xiii, 565; nocturnal, 308–327, 432–433; number of researchers of, 565; number of specimens of, 566; Ojoceratops fowleri n. gen. & sp. among, 171; paleobiogeography of, xiii, 387–404; paleoenvironmental associations and taphonomy of, 428–446; phylogenetic analysis, 222; posture, 10; Prenoceratops sp. among, 85; Protoceratops in the study of, 308–309; Psittacosaurus versus, 41, 55–56; recent publications and reviews of, 565; relationships among, 398–399, 400, 409, 410; Rubeosaurus ovatus n. gen. & comb. among, 157; similarities and sympatry among, 293–294; species diversity of, 565; stress fractures in, 349, 352; structural horn anatomy among, 279; Sweden neoceratopsian versus, 399; table of genera and species, stratigraphy, paleoenvironmental associations, and taphonomic studies, 439–446; taphonomic patterns among, 431–433, 433–434; taxonomy, 429, 566; thermoregulation in, 272; thermoregulation via horns of, 271–272, 280. See also Basal ceratopsians; Ceratopsids; Centrosaurines; Chasmosaurines; Neoceratopsians; Psittacosaurs Ceratopsid bonebeds. See Bonebeds Ceratopsid localities: Kaiparowits Formation, 480, 481, 482, 483 Ceratopsid remains: from New Mexico, 169–170 Ceratopsids, 271: agonistic behavior among, 374–379; from Alberta and Montana, 152; anatomical features, 406; from Asia, 391; bauplan of, 422; biogeography, 416–418; biostratigraphy, 187; bone surface texture in, 259; bonebeds, 447–455; in ceratopsian cladistics, 392, 400, 409, 410; ceratopsian dispersal and, 399; in ceratopsian family tree, 134; in ceratopsian paleobiogeography, 398; from Cerro del Pueblo Formation, 100, 102, 103–108, 103, 105, 106, 107, 108; chronostratigraphy, 390; cladistic analysis, 109–111, 110, 112; cladogram, 295, 296; CMN 8547, 189–202; CMN 8547 versus, 190, 191, 194, 195, 195, 196–197; CPC 278 among, 102; Diabloceratops eatoni n. gen.

& sp. among, 120, 125, 124, 127, 131; in Dinosaur Provincial Park bonebeds, 377; dispersal events among, 387; distribution, 99–100; diversity, stratigraphic distribution, biogeography, and evolution of, 405–427; earliest known, 413, 414; early, 96; evolution, 221; eyes, 314; forelimbs in locomotion among, 341–342; fossil record, 418–419, 419–422, 422– 423; frill function, 283; genera included among, 181, 442–446; from Hell Creek Formation, 203; high-latitude bonebed, 456–477, 458, 459; horns and frills in sparring among, 282–283, 283–290, 285, 286, 287, 288, 289; from Horseshoe Canyon Formation, 196–197; immature versus mature bone surface texture among, 252–253, 258; KBP discoveries of, 479, 481; locomotor behavior inferred from pathology in, 340–354; mammals versus, 356; mandibular evolution, 246–248; Medusaceratops lokii n. gen. & sp. among, 182; in Mexican stratigraphy, 100; from Mexico, 100–102; migrations of, 417; Montanoceratops cerorhynchus versus, 68, 69, 74; North American, 7; North American paleoenvironments for, 414–416; Ojoceratops fowleri n. gen. & sp. among, 169–180; Ojoceratops fowleri n. gen. & sp. among, 171; origin and evolution of horns and frills among, 290; Pachyrhinosaurus n. sp. among, 141–155; paleobiogeography of, 389; paleopathologies in Albertan, 355– 384; phylogeny, 411; Prenoceratops sp. versus, 85, 87–88; Protoceratops versus, 324; psittacosaur/neoceratopsian mandibles versus those of, 247–248; in Rattlesnake Mountain microsites, 525; Rubeosaurus ovatus n. gen. & comb. among, 157; semi-aquatic behavior, 189, 199, 200; skulls, 406, 406, 407; skulls and cladogram, 295, 296, 296–298, 298, 299; species diversity, 405, 406, 407–411, 407, 409, 410, 565; species diversity and turnover among, 419–422, 422–423; systematic phylogeny of indeterminate Mexican, 102–104; systematics, 9–10, 9; table of genera and species, stratigraphy, paleoenvironmental associations, and taphonomic studies, 442–446; taphonomy of Kaiparowits, 478–494; Tatankaceratops sacrisonorum n. gen. & sp. among, 204, 213, 216–217, 216; taxonomy, 406, 429; Triceratops versus, 264; Turanoceratops versus, 391; undiagnosed

index

579

Ceratopsids (continued) Kaiparowits, 490, 491; vertebrae, 192, 196, 199; youngest known, 413, 414; Zuniceratops christopheri versus, 91–98, 392; Zuniceratops versus, 187 Ceratopsomorphs: in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 213, 216–217, 216 Cerro del Pueblo chasmosaurine: bonebeds, 448, 449, 450; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 445; occurrence of, 416; stratigraphy, 412. See also Coahuilaceratops entries Cerro del Pueblo Formation: bonebeds in, 448, 450; ceratopsian paleoenvironmental associations and taphonomy in, 445; ceratopsians from, 99–116; in ceratopsid stratigraphy, 412; chasmosaurine from, 408; CPC 278 from, 102; CPC 279 from, 103, 109–110; fossils from, 100; in North American paleogeography, 415; stratigraphy, 100, 101 Cerro Grande Formation: in ceratopsid stratigraphy, 412; stratigraphy, 101 Cerro Huerta beds, 100–102, 101 Cerro Huerta Formation: in ceratopsid stratigraphy, 412 Cervicals: CMN 8547, 192–194, 192, 193, 194; Montanoceratops cerorhynchus, 74– 75, 76; pathological Centrosaurus, 356; pathologies in ceratopsid, 365–366, 365; pathology in Chasmosaurus irvinensis, 344; Psittacosaurus mongoliensis, 50–51; Tatankaceratops sacrisonorum n. gen. & sp., 211, 215; Zuniceratops christopheri, 96 Champsosaurs: in Hilda mega-bonebed, 502, 505 Chandler River: Kikak-Tegoseak Quarry and, 459; Prince Creek Formation and, 458 Changma, 60 Channel deposits: neoceratopsian taphonomy in, 433; in ‘‘Purple Hill’’ section, 526, 527 Channel facies association: at KikakTegoseak Quarry, 460–461 Channel fills: in and around KikakTegoseak Quarry, 460 Channel lag deposits: Triceratops in, 551, 554, 558 Chaoyangsaurus: as basal ceratopsian, 221, 222, 222; in ceratopsian cladistics, 392, 400; ceratopsian mandibles versus that of, 247; discovery, 11; Liaoceratops

580 index

yanzigouensis versus, 245; mandible, 235–236, 237; in Montanoceratops cladistics, 78, 80, 81; provenance, 387, 389, 390; Psittacosaurus versus, 41, 54, 56, 238; skull and mandible, 234; Sweden neoceratopsian versus, 399; taxonomy, 429 Chaoyangsaurus youngi: coronoid/mandible measurements, 248; mandible, 235–236, 237; mandibular element measurements for, 236; provenance, 391; Psittacosaurus versus, 53; stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; studied specimens, 235 Chapman, Ralph E., xvii, 264 Character correlation: in psittacosaur taxonomy, 23–24 Character states: in basal ceratopsian basicranium and palate, 231–232; in basal ceratopsian skulls, 234–235; in basal ceratopsians, 222–223; ceratopsid, 411, 421; in Coahuilaceratops magnacuerna n. gen. & sp. cladistics, 109, 112; in Montanoceratops cladistics, 79–81; in Pachyrhinosaurus n. sp. cladistics, 153– 154; problematic, 398; in psittacosaur taxonomy, 23–24, 28, 52–56; in Rubeosaurus ovatus n. gen. & comb. phylogenetic analysis, 159, 162–163, 164, 166–167; in Tatankaceratops sacrisonorum n. gen. & sp., 213, 214, 216–217 Charles M. Russell Wildlife Refuge: High Triceratops locality in, 560 Chasmosaurine bonebeds. See Bonebeds Chasmosaurine postcranium. See CMN 8547 Chasmosaurines: anatomical features, 406; behavior, 189, 199, 200; biogeography, 111–113; bone surface texture in, 259; bonebeds, 186, 187, 447, 448, 449, 450; bonebeds in Alberta, 356; Ceratops montanus among, 182; in ceratopsian biostratigraphy, 152; in ceratopsian cladistics, 409, 410; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110–111, 110; in ceratopsid fossil record, 418; ceratopsid species diversity and turnover and, 419, 420, 422; from Cerro del Pueblo Formation, 102, 103–108, 105, 106, 107, 108; CMN 8547, 189– 202; CMN 8547 among, 190, 191, 191, 197–198; Coahuilaceratops magnacuerna n. gen. & sp. versus, 106, 108–109, 110– 111; cornual sinuses, 276–278; depositional bonebed environments for, 448–

451, 452–453; Diabloceratops eatoni n. gen. & sp. versus, 124, 131, 135; Diabloceratops n. gen. versus, 133, 134; diagnoses, 198; distribution, 100; diversity and taxonomy of, 407–411, 407, 410; dyschondroplasia in, 348; evolutionary interactions between horn and frill morphology among, 282–292; frills as sexual display structures in, 6; genera included among, 181, 444–446; from Horseshoe Canyon Formation, 197, 198; indeterminate, 183, 186; intraspecific interaction analysis, 283–290, 285, 286, 287, 288, 289; juvenile horns, 556; Kaiparowits ceratopsid B among, 484–485; in Kaiparowits Formation bonebed, 478; in Mansfield Bonebed, 185–186; KBP discoveries of, 479, 481; locomotor behavior inferred from pathology in, 340–354; Medusaceratops lokii n. gen. & sp. among, 181–188; Medusaceratops lokii n. gen. & sp. versus, 184, 185, 187; from Mexico, 99, 100–102; occurrence of, 413; Ojoceratops fowleri n. gen. & sp. among, 169–180; Ojoceratops fowleri n. gen. & sp. versus, 171, 172, 176–177, 178; ontogeny, 185; in Pachyrhinosaurus n. sp. cladistics, 153; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153–154; paleoenvironmental associations of, 431; paleoenvironments for Kaiparowits, 491; phylogeny, 411; RFTRA systematics of, 9–10, 9; rib pathologies in, 367–368, 376–377; Rubeosaurus ovatus n. gen. & comb. versus, 159, 161, 166; similarities and sympatry among, 293–294; skull shapes and niche partitioning by sympatric, 293–307; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; skulls, 407; skulls and cladogram, 295, 296, 296–298, 299; sparring among, 283–290, 285, 286, 287, 288, 289, 375, 376–377; species diversity, 405, 407–411, 407, 410; stratigraphy, paleogeography, and diversity of, 430; in studies of ceratopsian sympatry, 294; table of genera and species, stratigraphy, paleoenvironmental associations, and taphonomic studies, 444–446; taphonomy, 433; Tatankaceratops sacrisonorum n. gen. & sp. among, 204, 205, 206–207, 210; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 213, 214–215, 216– 217, 216; taxonomy, 406, 429; Triceratops versus, 264; Zuniceratops christopheri versus, 95, 96

Chasmosaurus: biostratigraphy, 187; bonebeds, 433, 451–452; in ceratopsian family tree, 134; in ceratopsian frill stress analysis, 268; in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 285, 285, 286; CMN 8547 versus, 195; Coahuilaceratops magnacuerna n. gen. & sp. versus, 99, 104, 106, 108, 109; Diabloceratops n. gen. versus, 133; distribution, 100; dorsal pathologies in, 367; forelimb, 350; locomotor behavior, 340–354; manual pathologies in, 342, 372; manual phalangeal pathologies in, 374; manus, 343; Medusaceratops lokii n. gen. & sp. versus, 184; model at Calgary Zoo, 545; modeling frill of, 269; nasal pathologies in, 357; osteopathies of, 355; phylogeny, 405, 411; posture, 10; provenance, 431; RFTRA systematics of, 9, 10; rib pathologies in, 367–368; skeleton at British Museum, 546, 547, 549; skull from Dinosaur Park Formation Centrosaurus bonebed, 186; species of, 413; squamosal pathologies in, 361– 362; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216; taxonomy, 411, 429, 542; Triceratops versus, 264; Wayne Barlowe painting of, 10 Chasmosaurus belli: bonebeds, 449; in ceratopsian biostratigraphy, 152; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 444; in chasmosaurine intraspecific interaction analysis, 284, 285, 286; CMN 8547 versus, 192, 193, 195, 196; Coahuilaceratops magnacuerna n. gen. & sp. versus, 109; Diabloceratops n. gen. versus, 133; distribution, 100; dyschondroplasia in, 348; locomotor behavior, 340–354; Medusaceratops lokii n. gen. & sp. versus, 184, 185; occurrence of, 413; pathologies in specimen of, 344–346, 347; Peter Dodson’s work on, 8; skull, 407; skull and cladogram, 295, 296, 298; skull strengths and measurements, 299, 300, 301, 301, 302, 303, 304, 305; stratigraphy, 303, 412; studied specimens of, 342–343; ulna, 197 Chasmosaurus irvinensis: from Alberta, 152–153; bonebeds, 449; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 444; in chasmosaurine intraspecific interaction analysis, 285; CMN 8547 versus, 195; Coahuilaceratops

magnacuerna n. gen. & sp. versus, 109; Diabloceratops n. gen. versus, 133; distribution, 100; locomotor behavior, 340–354; Medusaceratops lokii n. gen. & sp. versus, 185; occurrence of, 413; pathological manus, 342, 343–344; pathologies in holotype specimen of, 343, 344, 345, 345, 346, 347; skull, 407; skull and cladogram, 295, 296, 298; skull strengths and measurements, 299, 300, 301, 301, 302, 303, 304, 305; stratigraphy, 303, 412; studied specimens of, 342; in studies of ceratopsian sympatry, 294, 297 Chasmosaurus kaiseni: RFTRA systematics of, 10 Chasmosaurus mariscalensis: from Aguja Formation microsites, 530; distribution and taxonomy, 100; growth series in, 259; taxonomy, 411, 520. See also Agujaceratops mariscalensis Chasmosaurus russelli: biostratigraphy, 187; bonebeds, 449; in ceratopsian biostratigraphy, 152; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 444; in chasmosaurine intraspecific interaction analysis, 284, 285, 286; Coahuilaceratops magnacuerna n. gen. & sp. versus, 109; Diabloceratops n. gen. versus, 133; distribution, 100; frill function, 283; humerus, 195; Medusaceratops lokii n. gen. & sp. versus, 184, 185, 187; occurrence of, 413; skull, 407; skull and cladogram, 295, 296, 298; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 303, 412 Cheeks: psittacosaur, 41 Chelonians: Grand Staircase–Escalante localities and distribution, 119 Chemical analysis: of Triceratops matrix, 273–275, 274, 275–276 Chevrons: Montanoceratops cerorhynchus, 69; pathologies in ceratopsid, 368, 370, 371, 372; Psittacosaurus, 333–334 Chi-square analysis: of chasmosaurine intraspecific interactions, 284 Chicago Field Museum: Protoceratops skeletons collected by, 510 Chicken humerus: dermestid damage to, 517 Chickens: osteomyelitic infections in, 348 Chimedtseren, Anadin: Fox Protoceratops and, 510 Chimeras: psittacosaurs versus, 335 China, 510; Archaeoceratops oshimai from,

240; Archaeoceratops yujingziensis n. sp. from, 59–67, 60; Auroraceratops rugosus from, 241; basal ceratopsians from, 390, 429–430; basal neoceratopsian taphonomy in, 432, 434; basal neoceratopsians from, 431; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440, 441, 442; ceratopsian paleoenvironments in, 428; ceratopsian taphonomy in, 432; ceratopsians from, 132, 429, 566; Chaoyangsaurus youngi from, 235; Hongshanosaurus houi from, 238; Liaoceratops yanzigouensis from, 245; Mesozoic paleogeography of, 397a; new dinosaur discoveries in, 11, 12; Peter Dodson in, 10–11, 12; Protoceratops andrewsi from, 246, 321; psittacosaurs from, 21–22, 22, 22–23, 26, 28, 29, 32, 38, 329, 330, 335; Psittacosaurus from, 328; Psittacosaurus species from, 238; Psittacosaurus taphonomy in, 432; recent basal ceratopsian discoveries in, 221– 222, 231; Yinlong downsi from, 236 Chinese Academy of Sciences, 59 Chinnery-Allgeier, Brenda J., xiii, xiv, xvii, xxi, 83, 91, 387; Peter Dodson and, 6, 7 Chionis: eye size, 312 Choanae: basal ceratopsian, 225, 231 Chondrichthyes: Grand Staircase– Escalante localities and distribution, 119 Chronic degenerative joint disease: in chasmosaurine manus, 342 Chronic osteomyelitis: in dinosaurs, 347, 352 Chronic periosteal inflammation, 346 Chronostratigraphy: of ceratopsians, 390; Kaiparowits Formation, 479 Cicatricosisporites: as Kikak-Tegoseak Quarry palynomorph, 462 Ciconiids: eye sizes, 312 Ciconiiforms: orbit shapes, 311 Circle Cliffs, 119 Clades: paleobiogeography of ceratopsian, 388–389; psittacosaur, 54–55, 55 Cladistics: of Avaceratops lammersi, 8; in ceratopsian paleobiogeography, 388– 389, 392, 398–399, 400, 409, 410; of ceratopsids, 109–111, 110, 112, 294, 295, 296; of Coahuilaceratops magnacuerna n. gen. & sp., 109, 110–111, 110, 112; in dinosaur paleobiogeography, 389; of Montanoceratops cerorhynchus, 77–79, 79–81; Pachyrhinosaurus n. sp., 150– 153, 152, 153–154; Peter Dodson’s ceratopsian systematics and, 9–10; Tatankaceratops sacrisonorum n. gen. & sp., 213,

index

581

Cladistics (continued) 214, 216–217, 216. See also Phylogenetic analyses Cladogenesis: in centrosaurines, 163–165; in ceratopsids, 421 Claggett Shale: in ceratopsid stratigraphy, 412 Clam shells: in Big Bend microsites, 533 Clay coatings/infillings: at Kikak-Tegoseak Quarry, 466, 468, 469 Clay concentrations: at Kikak-Tegoseak Quarry, 469 Clay illuviation: at Kikak-Tegoseak Quarry, 467, 472 Clay-rich papules: at Kikak-Tegoseak Quarry, 466, 467, 468, 469 Claystone facies: of Hilda mega-bonebed, 500, 500 Claystone layer: on Triceratops horns, 271, 280 Cleridae: Protoceratops skeleton insect scavenging versus that by, 510–511 Cleveland Museum of Natural History, xiii Climate: ceratopsids and, 414, 428, 439– 446; for high-latitude dinosaurs, 456– 457; during Hilda mega-bonebed formation, 505; fossil insect-modified skeletons and, 518; Kaiparowits Formation and, 481; Late Cretaceous Big Bend, 521; Prince Creek Formation and, 472; during Rattlesnake Mountain microsite deposition, 524 Cloverly Formation: basal ceratopsians from, 387, 389–391; John Ostrom’s fieldwork in, 4 Cluster diagrams: in dinosaur paleobiogeography, 389 CMN 8547, 189–202; chasmosaurine behavior and, 199; description, 190– 196; discovery, 189; locality, 190, 190; manus, 197; significance, 199–200; stratigraphy, 191, 196–197; systematic paleontology, 190; taxonomy, 196–198; ulna, 197. See also National Museum of Canada CNM 41357: pathologies in, 340–354, 342, 345, 346, 347 Coahuila, 101; bonebeds in, 448, 450; ceratopsian paleoenvironmental associations and taphonomy in, 445; ceratopsians from, 99–116, 566; in ceratopsid fossil record, 419; chasmosaurine from, 408 Coahuilaceratops magnacuerna n. gen. & sp., 565; biogeographic implications, 111–113; in ceratopsid cladistic analysis,

582 index

110, 112; circumnarial region, 104–107, 105, 106; circumorbital region, 107– 108; description, 103–104, 104–108; diagnosis, 104; holotype material, 103– 104; material, 99; parietosquamosal frill, 108, 109; phylogenetic analysis, 108– 109; postcranial elements, 104; referred material, 104; skull, 407; systematic paleontology, 104–108; taxonomy, 99, 109–111 Coal Creek, 170 Coalified tree limbs: in Big Bend microsites, 533; in ‘‘Purple Hill’’ section, 527 Coals: table of ceratopsian, 439–446. See also Z coal complex Coastal plains: in Big Bend National Park geology, 521; bonebeds in, 449–450, 456; as ceratopsid paleoenvironments, 414–416, 420, 428; Hell Creek Formation and, 552; Hilda mega-bonebed deposited in, 495, 497, 503–505; neoceratopsians in, 431; Prince Creek Formation and, 458; Zuniceratops christopheri and, 92 Coccidioidomycosis: osteomyelitis and, 348 Coelophysis bauri: eye size and body mass of, 317, 318 Coelophysis kayentakatae: eye size and body mass of, 317, 318 Coelurosaurs: in Mesozoic paleogeography, 398 Colbert, Edwin H.: Peter Dodson and, 3, 4 Cole, Ava: Peter Dodson and, 7 Cole, Eddie: Peter Dodson and, 7 Coleopterans: in Kikak-Tegoseak Quarry fossils, 470, 472, 473; Protoceratops skeleton insect scavenging versus that by, 510–511; pupation chambers, 513–515, 513, 514, 515. See also Beetles Collection: of ceratopsians, xiii–xiv Colorado: Big Bend dinosaurs versus those of, 521; ceratopsian paleoenvironmental associations and taphonomy in, 446; ceratopsid distribution in, 99; ceratopsid stratigraphy in, 412; paleoenvironments of, 417; Triceratops from, 551; Triceratops horridus from, 417; Zuniceratops christopheri and, 92 Columbiforms: eye sizes, 312 Colville Basin: Kikak-Tegoseak Quarry in, 457, 458, 473 Colville River: Kikak-Tegoseak Quarry along, 457, 458, 459 Communication-based evolution: of ceratopsids, 421

Competition: ceratopsian horns in, 271; among sympatric ceratopsians, 293, 294 Compression: of manus in ceratopsid walking, 351–352 Compressional strength: in ceratopsian skulls, 294–296, 296–298, 297, 298– 305, 298, 299, 301, 3040 Computerized tomography (CT): of ceratopsian phalanges, 343; in diagnosing paleopathologies, 356, 377; of Triceratops skull, 273, 276, 278, 278 Computers: Peter Dodson and, 4, 8 Comrey Sandstone Zone: in Medusaceratops lokii n. gen. & sp. stratigraphy, 182 Concentration values: in Hilda megabonebed, 502–503 Conchae: ceratopsian, 232 Conchoidal fractures: on dentition, 316 Coniacian stage: basal ceratopsians from, 390; ceratopsian cladistics and, 400; ceratopsian family tree and, 134; ceratopsid cladistic analysis and, 110 Conifers: in Kikak-Tegoseak Quarry bonebed paleoenvironment, 456, 472 Continental landmasses: in biogeography, 416; in ceratopsian paleobiogeography, 389, 392–396, 395, 396, 397; ceratopsids and, 414 Continuoolithus: in Rattlesnake Mountain microsites, 524 Converse County: Triceratops from, 551 Cope, Edward Drinker, 565; on Agathaumas, 565; Peter Dodson and, 4, 10 Coprolites: in Big Bend microsites, 533, 534; in Rattlesnake Mountain microsites, 524 Coracoids: CMN 8547, 193, 193, 194, 195; Montanoceratops cerorhynchus, 75; Zuniceratops christopheri, 95, 96 Cormorants: eye sizes, 312 Cornual sinus matrix: of Triceratops horns, 276–278, 279–280 Cornual sinuses: among ceratopsians, 279–280; Shenandoah University Triceratops, 273, 276–278, 278, 279–280; Tatankaceratops sacrisonorum n. gen. & sp., 205; Triceratops, 271 Coronoids: Archaeoceratops oshimai, 241, 241; Archaeoceratops yujingziensis n. sp., 63, 63; Auroraceratops rugosus, 241, 242; basal versus advanced ceratopsian, 247, 248; Calgary Public Museum ceratopsian, 546; Chaoyangsaurus youngi, 235; Hongshanosaurus houi, 238; Leptoceratops

gracilis, 243, 244; Liaoceratops yanzigouensis, 244, 245; measurements of basal ceratopsian, 236; Montanoceratops cerorhynchus, 69, 70, 73, 74, 74, 76; Ojoceratops fowleri n. gen. & sp., 175, 175; osteomyelitis in Triceratops, 347; Pachyrhinosaurus n. sp., 150, 151; pathologies in, 364–365, 364; Protoceratops andrewsi, 246; Psittacosaurus, 214, 239, 240; Tatankaceratops sacrisonorum n. gen. & sp., 210–211, 212, 213; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 213; Yinlong downsi, 236; Zuniceratops christopheri, 93, 96 Coronosaurians: in ceratopsian paleobiogeography, 398 Cortical resorption, 255, 258–259 Corvids: eye sizes, 312 Corvus corax: eye size, 312; sclerotic rings, 311 Corythosaurus: in ceratopsian biostratigraphy, 152; ceratopsid species diversity and turnover and, 419; psittacosaurs versus, 44; sexual dimorphism in, 5, 9; skeleton at Calgary Zoo, 546 Corythosaurus casuarius, 8; eye size and body mass of, 317 Corythosaurus zone: Kritosaurus zone versus, 521 Coste House: Calgary Public Museum collection at, 544, 545 Cottontails: in arid environments, 323 Covered deposits: in ‘‘Purple Hill’’ section, 526 Cow Creek Valley: Ceratops montanus from, 182 Coyotes: osteomyelitis in, 348 CPC 276, 99, 103–104, 104–106, 105, 106, 107, 108, 113; locality, 101. See also Coahuilaceratops magnacuerna n. gen. & sp. CPC 277, 99, 104, 106–108, 113; locality, 101. See also Coahuilaceratops magnacuerna n. gen. & sp. CPC 278, 99, 102, 103; taxonomy, 110 CPC 279, 99. 103, 103, 113; biogeography, 111; locality, 101; taxonomy, 109–110 Cranial material. See Skull entries Cranial nerves: Diabloceratops eatoni n. gen. & sp., 126, 127 Cranial ontogeny: psittacosaur, 21 Cranial orientation: chasmosaurine, 282, 284–290, 285, 286, 287, 288, 289; psittacosaur, 40–52 Cranial ornamentation: bone surface texture of centrosaurine, 259–260; chasmo-

saurine, 282, 284–290, 285, 286, 287, 288, 289; functions of ceratopsian, 282– 283, 289–290 Cranial ornamentation development: relevance to ontogeny of fossil vertebrates, 252 Cranioquadrate passage: basal ceratopsian, 226–227 Cranwellia rumseyensis: as Kikak-Tegoseak Quarry palynomorph, 461 Cranwellia striata: as Kikak-Tegoseak Quarry palynomorph, 462 Craspedodon: in ceratopsian cladistics, 400; dentition, 393, 399; provenance, 390, 399 Craspedodon lonzeensis: dentition, 391, 393, 399; provenance, 391, 399 Creagrus furcatus: nocturnal lifestyle of, 320 Crepuscular lifestyle: eyes and, 311; Troodon, 311 Cretaceous period: bonebeds from, 448– 451, 450, 452, 456–477, 458, 459; ceratopsian cladistics and, 400, 409, 410; ceratopsian paleobiogeography during, 388; ceratopsian radiation during, 565; distribution of ceratopsians during, 389– 392; in Mexican stratigraphy, 101; in North Slope stratigraphy, 457, 460; paleogeography during, 392, 394–396, 396, 397, 401; recent basal ceratopsian discoveries from, 221–222 Cretaceous Western Interior Seaway (KWIS): ceratopsids and, 414–416, 415, 417, 420, 421–422. See also Western Interior Seaway Cretaceous-Paleogene boundary: Triceratops horridus at, 413. See also Maastrichtian-Paleocene boundary Cretaceous-Tertiary (KT) boundary. See Cretaceous-Paleogene boundary; Maastrichtian-Paleocene boundary Crevasse splay deposits: at Kikak-Tegoseak Quarry, 465 Crocodylians: basal ceratopsians versus, 227; in Big Bend microsites, 533; from Big Bend National Park, 521; caudal pathologies in Pleistocene, 368; in diagnosing dinosaur pathologies, 341; Grand Staircase–Escalante localities and distribution, 119; in Hilda mega-bonebed, 502, 505; juvenile, 8; nares and orbits of, 334; nutritional stress in, 365; ontogenetic bone surface texture changes in extant, 252; psittacosaurs versus, 335; Psittacosaurus versus, 333–334; in Rat-

tlesnake Mountain microsites, 524, 525; sclerotic rings among, 311; semi-aquatic ceratopsids and, 199; sprawling pose of, 330; with undiagnosed Kaiparowits ceratopsid, 491 Crocodylomorphs: gastroliths found with, 333 Cross sections (anatomical): of ceratopsian skulls, 294–296, 296–298, 297, 301 Cross sections (geological): Aguja Formation, 523, 524; Hilda mega-bonebed, 497, 497, 498, 499, 500; ‘‘Purple Hill,’’ 526 Cross-taxic faunal turnover: in ceratopsid evolution, 421 Crows: eye sizes, 312 Crowsnest Pass: Montanoceratops cerorhynchus from, 69, 76–77 Cruising speed: avian eye size allometry and, 313–314 Crustal rifting: psittacosaur evolution and, 329 Crypsis: in ceratopsids, 420 Cryptobranchius: caudal fin of, 335 Cryptobranchoid salamanders: in Late Jurassic paleogeography, 394 CTi-GE scanner, 273 Curculionidae: pupation chambers, 513, 517 Curlews: eye sizes, 312; nocturnal, 320 Currie, Philip J., xiii, xvii, xxi, 83, 141, 565; discovery of Pachyrhinosaurus n. sp. and, 142; 1979 fieldwork led by, 542; Peter Dodson and, 7, 8 Cutbank, 85 Cutler, William Edmund, 543; biography, 541–542; death, 541, 542; Eoceratops and other fossils collected by, 541–550; hadrosaur excavated by, 543–544 Cyanoliscus patagonus: as burrow dweller, 323 Cycadopites: as Kikak-Tegoseak Quarry palynomorph, 462 Cyclura cornuta: pushing contests among, 290 Cystic lesions: diagnosing in fossils, 349 Cysts: diagnosing in fossils, 349 Dakota Formation: ceratopsian family tree and, 134; Prince Creek Formation versus, 471 Dalanzadgad, 510 Danian stage: in Big Bend National Park geology, 523; Kikak-Tegoseak Quarry palynomorphs from, 461 Danis, Gilles: Peter Dodson and, 4

index

583

Dart, Raymond, 8 Dasyatids: in Rattlesnake Mountain microsites, 525 Data collection: for studies of eye size versus ecology, 312 Dating: Cedar Mountain Formation, 394; ceratopsian-bearing formations, 129– 131, 411–413; Kaiparowits Formation, 411–413, 479; Naashoibito Member, 171; Prince Creek Formation, 458, 459; Wahweap Formation, 413 Dawson Creek, 522 Dawson Creek section: Agujaceratops from, 520; ankylosaurs from, 528; in Big Bend National Park geology, 521, 525 de Leon, Claudio A., xvii, 99 De-na-zin Member: ceratopsians from, 169, 177 De-na-zin Wash, 170, 170 Death adder: nocturnal lifestyle of, 320 Death assemblages: estimating size of Hilda mega-bonebed, 505–506, 506–507 Death events: basal ceratopsian, 429, 430; centrosaurine mega-bonebed and, 495– 508. See also Catastrophic death events DeBlieux, Donald D., xvii, 117, 118 Deciduous forests: in Kikak-Tegoseak Quarry bonebed paleoenvironment, 456, 472 Decomposition: insect damage during, 515–517; of Kaiparowits ceratopsid B bones, 487 Deep sea fish: binocular vision among, 321 Deep sea lifestyle: ichthyosaur eyes indicating, 311 Deer: in arid environments, 323 Defense: ceratopsian horns in, 271; in Triceratops intraspecific interactions, 288 Defensive structures: frills as, 269 Deformation: of stressed Triceratops frill, 267, 268 Degenerative joint disease: in chasmosaurine manus, 340, 342 Deinocheirus: from Nemegt Formation, 322 Deinonychus: discovery, 4 Deltopectoral crest: in ceratopsian posture, 10 Delunchan/Delunshan: psittacosaurs from, 38; Psittacosaurus xinjiangensis from, 330, 331 Demographics: in Triceratops intraspecific interactions, 287–288 Dental batteries: basal versus advanced ceratopsian, 234 Dental caries: absence in ceratopsids, 365 Dentaries: Archaeoceratops oshimai, 240,

584 index

241; Archaeoceratops yujingziensis n. sp., 62, 63; Auroraceratops rugosus, 241, 242, 243; Chaoyangsaurus youngi, 235–236; Coahuilaceratops magnacuerna n. gen. & sp., 104, 105, 106–107; at High Triceratops locality, 558; Hongshanosaurus houi, 238; Leptoceratops gracilis, 243, 244; Liaoceratops yanzigouensis, 244, 245; measurements of basal ceratopsian, 236; Montanoceratops cerorhynchus, 69, 73, 73, 76; Ojoceratops fowleri n. gen. & sp., 172, 175, 175; Pachyrhinosaurus n. sp., 149– 150, 151; pathologies in, 364–365, 364; Protoceratops andrewsi, 246; psittacosaur, 329; Psittacosaurus, 52, 239, 239; Psittacosaurus mongoliensis, 46; Psittacosaurus sattayaraki, 39–40; Tatankaceratops sacrisonorum n. gen. & sp., 210, 211, 212, 213; at Trike II Triceratops locality, 558, 559; Yinlong downsi, 237, 237; Zuniceratops christopheri, 92, 93, 94, 95 Dentary flange: psittacosaur, 43 Denticles: dromaeosaurid, 532; tyrannosaurid, 530–531, 530 Dentition: Archaeoceratops oshimai, 241, 241; Archaeoceratops yujingziensis n. sp., 61, 63–64, 63, 64; Arundel ceratopsian, 391, 399; Auroraceratops rugosus, 241, 241, 243; ceratopsian, 247, 421; Chaoyangsaurus youngi, 235, 236; Coahuilaceratops magnacuerna n. gen. & sp., 106; Craspedodon lonzeensis, 391, 393, 399; Diabloceratops eatoni n. gen. & sp., 123, 124; gastroliths and, 333; of highlatitude vertebrates, 457; Hongshanosaurus houi, 238; Leptoceratops gracilis, 244, 245; Liaoceratops yanzigouensis, 244, 246; measurements of basal ceratopsian, 236; Montanoceratops cerorhynchus, 70; neoceratopsian, 393, 396–397, 399; neonate psittacosaur, 51–52; Pachyrhinosaurus n. sp., 149–150, 151; pathologies in, 364–365, 364; Protoceratops, 309, 315–316, 319; Protoceratops andrewsi, 244, 246; psittacosaur, 39–40, 44, 45, 46, 329; Psittacosaurus, 239; Psittacosaurus major, 30, 31; Psittacosaurus mazongshanensis, 38–39; Psittacosaurus mongoliensis, 33, 46; Psittacosaurus sattayaraki, 39– 40; Psittacosaurus sp., 27; Psittacosaurus xinjiangensis, 38; replacement of, 365; Shenandoah University Triceratops, 273; Turanoceratops, 391; Yinlong downsi, 237; Zuniceratops christopheri, 92, 93, 94–95, 95, 96, 392

Denver: Zuniceratops christopheri and, 92 Denver Formation: in ceratopsid stratigraphy, 412; in North American paleogeography, 415 Depositional environments: in bonebed studies, 447; of ceratopsid bonebeds, 448–451, 452, 456; Hilda megabonebed, 497–500, 497, 498, 499, 500; of Kikak-Tegoseak Quarry bonebed, 456; in ‘‘Purple Hill’’ section, 527 Depositional history: of Hilda megabonebed, 503–505 Dermal elements: psittacosaur, 42–43. See also Osteoderms Dermatophytosis: osteomyelitis and, 348 Dermestes maculates: Protoceratops skeleton insect scavenging versus that by, 510– 511 Dermestid beetles: bone modification by, 515, 516, 517; chicken humerus damage by, 517; in Kikak-Tegoseak Quarry fossils, 470, 472, 473; Protoceratops skeleton insect scavenging versus that by, 510– 511. See also Insect scavenging Desert environments. See Arid environments Desert skink: nocturnal lifestyle of, 320 Desmoscaphites: in Wahweap Formation stratigraphy/dating, 129 Developmental Mass Extrapolation (DME), 5 Devil’s Coulee: TMP 87.89.8 from, 85, 85, 88 Devonian period: Peter Dodson’s fieldwork in, 4 Diabloceratops eatoni n. gen. & sp., 117– 140, 565; Asian protoceratopsid monophyly and, 133–134; braincase, 126– 127; circumorbital region, 124; description, 123–128, 128–129; diagnosis, 122; discovery and collection, 118, 120, 121; frill, 123, 127–128, 132, 132; holotype skull, 120–122, 122; life reconstruction, 137; ontogeny, 134–135; other ceratopsians versus, 117, 131–136; palatal region, 125–126; phylogenetic analysis, 131, 133–134; skull, 406; skull and skull elements, 117–140, 122, 123, 124, 125, 126, 132, 134, 135, 136, 137; skull roof, 124, 124; snout, 123–124, 123; stratigraphic context/occurrence, 118, 119, 120, 121, 122, 129–131; systematic paleontology, 120–122; taxonomy, 135–136; temporal region, 124–125. See also Wahweap centrosaurine Diabloceratops n. gen.: in ceratopsian fam-

ily tree, 134; frill, 132, 132; ontogeny, 134–135; skull material, 128–129, 130; taxonomy, 135–136 Diagnoses (pathologic): of Chasmosaurus paleopathologies, 346–349; importance of, 349; in paleopathology, 340–341, 346–349 Diagnoses (taxonomic): Archaeoceratops yujingziensis n. sp., 60; chasmosaurine, 198; Coahuilaceratops magnacuerna n. gen. & sp., 104; Diabloceratops eatoni n. gen. & sp., 122; Medusaceratops lokii n. gen. & sp., 182, 183; Montanoceratops cerorhynchus, 70; Ojoceratops fowleri n. gen. & sp., 171; Psittacosaurus lujiatunensis, 28; Psittacosaurus major, 28–29; Psittacosaurus meileyingensis, 32; Psittacosaurus mongoliensis, 26, 32; Psittacosaurus neimongoliensis, 32–34; Psittacosaurus sibiricus, 34; Psittacosaurus sinensis, 37; Psittacosaurus sp., 26; Psittacosaurus xinjiangensis, 38; Rubeosaurus ovatus n. gen. & comb., 157; Tatankaceratops sacrisonorum n. gen. & sp., 204. See also Differential diagnoses Diagnostic material: in psittacosaur taxonomy, 23–24, 25–26 Diagrams: in chasmosaurine intraspecific interaction analysis, 284 Diaphyseal microfractures: diagnosing in fossils, 349 Diapsids: extant nocturnal, 320 Diceratops: in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 284, 286, 288; Coahuilaceratops magnacuerna n. gen. & sp. versus, 104, 109, 111; frill function, 283; Medusaceratops lokii n. gen. & sp. versus, 185; occurrence of, 414–415; Ojoceratops fowleri n. gen. & sp. versus, 177; provenance, 431; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216 Diceratops hatcheri: bonebeds, 449; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 445; occurrence of, 413, 419; Ojoceratops fowleri n. gen. & sp. versus, 169, 178; skull, 407; skull and cladogram, 295, 296, 297, 298; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412 Diceratus. See Diceratops Dicynodonts: sclerotic rings among, 312 Differential diagnoses: of Chasmosaurus

paleopathologies, 346–349; importance of, 349; in paleopathology, 340–341, 346–349; in psittacosaur taxonomy, 24; Rubeosaurus ovatus n. gen. & comb., 157 Diffuse idiopathic skeletal hyperostosis (DISH): in ceratopsian vertebrae, 366, 370 Difunta Group: Coahuilaceratops magnacuerna n. gen. & sp. from, 104; CPC 278 from, 102; CPC 279 from, 103; fieldwork in, 100; stratigraphy, 100, 101 Digital modeling: in locomotion studies, 349, 350–351, 351–352 Digital scans: ceratopsid locomotion and, 341, 349, 350–351, 351–352 Digitized models: of ceratopsian skulls, 294–296, 296–298, 297, 301 Dinamation International Society (DIS): dinosaur expeditions to Mongolia by, 509 Dineley, David: Peter Dodson and, 4 Dinocephalians: head butting in, 356 Dinosaur babies. See Baby dinosaurs; Hatchlings; Immature individuals; Juveniles Dinosaur bone(s): burrowing through, 517; insect-modified, 509 Dinosaur Book, The (Colbert), 3 Dinosaur Park centrosaurine: in ceratopsian cladistics, 409; stratigraphy, 412. See also Pachyrhinosaurus n. sp. Dinosaur Park Formation (DPF): Aguja Formation versus, 527, 533; basal neoceratopsians from, 83; bonebeds in, 448, 449, 450, 451, 452; ceratopsian abundance in, 413; ceratopsian bonebeds in, 186; ceratopsian distribution in, 100, 111; ceratopsian paleoenvironmental associations and taphonomy in, 441, 443, 444; ceratopsid bonebeds in, 496, 497; ceratopsid species diversity and turnover in, 421, 422; in ceratopsid stratigraphy, 412; ceratopsids from, 7, 414; Chasmosaurus russelli from, 187; Chasmosaurus species in, 413; dating, 413; geologic correlation, 84; geology and paleoenvironment of, 499–500, 503– 504, 507; Hilda mega-bonebed in, 497– 500, 498, 499, 499, 500; Kaiparowits Formation versus, 478, 479, 491; Leptoceratops sp. from, 88; neoceratopsian bonebeds in, 433; in North American paleogeography, 415; pachyrhinosaur from, 408; Pachyrhinosaurus n. sp. from, 141–155; paleoenvironments of, 415, 417; Prenoceratops sp. from, 85; stratig-

raphy, 143, 144; Styracosaurus from, 156; sympatric ceratopsians from, 293, 300–303, 303; in Wahweap Formation stratigraphy/dating, 131 Dinosaur Provincial Park (DPP), 142, 190, 482; Aguja Formation versus geology of, 527, 533–534; bonebeds, 449, 450, 451, 452, 566; British Museum Chasmosaurus from, 546, 547, 549; ceratopsian bonebeds in, 356; ceratopsian paleoenvironmental associations and taphonomy in, 443; ceratopsian research in, 566; ceratopsid bonebeds in, 377, 495–496; cervical pathologies in ceratopsids from, 365–366; Charles M. Sternberg in, 545; Eoceratops collected in, 541, 542; hadrosaurs from, 5; Hilda mega-bonebed and, 495, 496, 496, 497, 497, 504–505, 505– 506, 507; neoceratopsian taphonomy in, 433; Pachyrhinosaurus n. sp. from, 141; paleopathologies in bonebed specimens from, 376; pathological Pachyrhinosaurus rostrals from, 357, 357; Peter Dodson’s fieldwork in, 4, 7; phalangeal stress fractures in ceratopsids from, 373– 374; Prenoceratops sp. from, 85; sympatric ceratopsians from, 293, 300–303, 303; William Cutler in, 542, 543–544, 547 Dinosaur Renaissance, 328 Dinosaur research: paleopathology in, 355–356 Dinosaur Systematics Conference: Peter Dodson at, 9 Dinosauria, 25. See also Dinosaurs Dinosauria, The: ceratopsians in, 8–9; publication of, 8 Dinosauroid prismatic eggshells: in Rattlesnake Mountain microsites, 524 Dinosauroid spherulitic eggshells: in Rattlesnake Mountain microsites, 524 Dinosaur-rich habitats: biogeography of, 416, 417 Dinosaurs, 271; in Agujaceratops bonebeds, 520–537; Alaskan, 456; basal ceratopsians among, 221; Bayn Dzak, 320, 322; in Big Bend microsites, 533, 534; from Big Bend National Park, 521, 524; bone surface texture in, 259; burrowing by, 323; ceratopsid species diversity and turnover and, 419, 421, 422; Cerro del Pueblo Formation, 100; collected by William Cutler, 542; diagnosing pathologies in, 341; display structures in, 6; eye sizes among, 311; eyes of, 308, 314; eyesight, 309; fossil insect-modified skele-

index

585

Dinosaurs (continued) tons of, 518; from Gongpoquan Basin, 59; Grand Staircase–Escalante localities and distribution, 119; herbivorous, 308, 314; in high-latitude localities, 456–457; immature versus mature bone texture among, 252; Kaiparowits Formation, 478–479; from Kaiparowits Plateau, 118; KBP discoveries of, 479, 481; at KikakTegoseak Quarry, 467, 469, 473; Medusaceratops lokii n. gen. & sp. and, 181–182; Mesozoic dispersal patterns among nonceratopsian, 396–398; nasal functions in, 232; Nemegt and Djadokhta, 322; nocturnal lifestyles among, 317–323; number of researchers of, 565; from Ojo Alamo and Kiryland Formations, 170; Ojoceratops fowleri n. gen. & sp. among, 169–180; ontogenetic patterns among, 252; paleobiogeography of, 388–389, 392–396; paleobiogeography of ceratopsian, xiii, 387–404; Peter Dodson’s work on, 4–5; Prince Creek Formation, 457, 458; prospecting Hell Creek Formation for, 554–555, 554, 555; Protoceratops in the study of, 308–309; public interest in, 555; in Rattlesnake Mountain microsites, 524, 525; relative aperture size among, 320, 321; rib damage due to scavenging, 367; sclerotic rings and eye sizes among, 312, 313–314, 315, 315, 316, 317, 318; semi-aquatic behavior among, 328–329; sympatric, 293–294; thermoregulation in, 272; trampling by, 433; from Two Medicine Formation, 156 Diomedea immutabilis: eye size, 312; nocturnal lifestyle of, 320 Diplodocus: eye size, 311; eye size and body mass, 318 Diplodocus longus: eye size and body mass of, 317 Diseases: diagnosing in fossils, 340–341 Dispersal events: among ceratopsians, 387, 388–389, 398–399, 399–401; among non-ceratopsians, 396–398 Display structures: in dinosaurs, 6; horns and frills as, 282, 377 Distance matrices: in dinosaur paleobiogeography, 389 Distortion: in ceratopsian skull and jaw analysis, 294, 296 Distribution: basal ceratopsian, 66, 389; ceratopsian, 99–100, 110, 111, 113, 389–392, 398–399, 429; ceratopsid, 406; neoceratopsian, 389–392; psittacosaur, 389. See also Stratigraphic distribution

586 index

Diurnal birds: eyes of, 310, 321 Diurnal feeding times: nocturnal feeding times versus, 293, 432 Diurnal lifestyle: eyes and, 308, 310 Divers: eye sizes of avian, 312, 314, 321 Diversity: in Agujaceratops bonebeds, 520– 537; of Bayn Dzak dinosaurs, 320, 322; ceratopsian, 429, 430; ceratopsid, 405– 427, 428; in ceratopsid fossil record, 418–419, 419–422, 422–423; of Nemegt dinosaurs, 322y; of Protoceratops fauna, 322–323; psittacosaur, 329; Psittacosaurus, 328; of sympatric ceratopsians, 293–294, 300–305 Djadokhta Formation: as arid paleoenvironment, 322–323; ceratopsian paleoenvironmental associations and taphonomy in, 440, 441, 442; ceratopsians from, 132; insect trace fossils with Protoceratops skeletons from, 509–520, 512, 513, 514, 515, 516, 517; insect-modified skeletons in, 518; Protoceratops from, 308, 309, 316–317, 321, 322–323, 322, 324; pupation chambers from, 513–515, 513, 514, 515; taphonomy, 432–433; trace fossils in, 511–513, 512, 513 Dodson, Dawn, 4, 13 Dodson, Peter, xiii, xvii, xxi, 59, 221, 234, 565, 566; on ceratophilia, 3–17; early career, 4; early influences on, 3–4; first studies of ceratopsians by, 4; students, 6–7, 7, 10–12 Dolphins: psittacosaurs versus, 332, 334 Domestic fowl: osteomyelitic infections in, 348 Dongshen: psittacosaurs from, 32 Dorsal skull roof: neonate psittacosaur, 46–48; psittacosaur, 41–42, 43 Dorsals: Archaeoceratops yujingziensis n. sp., 64, 65; CMN 8547, 192, 192, 193, 193, 194, 194; Montanoceratops cerorhynchus, 74–75, 75; pathological Styracosaurus, 356; pathologies in ceratopsid, 366–367, 366, 367, 368; Tatankaceratops sacrisonorum n. gen. & sp., 211, 215; Zuniceratops christopheri, 95, 96 Dorsosacrals: CMN 8547, 194 Doushan Formation: psittacosaurs from, 37 Drawings: in chasmosaurine intraspecific interaction analysis, 284 Drip Tank Member: Diabloceratops n. gen. and, 122, 128; stratigraphy/dating, 129 Dromaeosaurids: Aguja Formation, 525, 529, 531–532, 532; in Agujaceratops bonebeds, 520; from Nemegt Formation,

322; in Rattlesnake Mountain microsites, 525, 529, 531–532, 532. See also Dromaeosaurs Dromaeosaurs: dentition, 316; eye sizes and body masses of, 317; in KikakTegoseak Quarry bonebed, 456 Dromaeosaurus albertensis: eye size and body mass of, 317, 318; at KikakTegoseak Quarry, 467 Dromornithids: relative eye sizes, 311 Drought: bonebed formation during, 451, 505; Hilda mega-bonebed formation during, 505 Drowning: ceratopsid bonebeds and, 449– 450 Drumheller, 190; Calgary Public Museum collection at, 544; Ceratopsian Symposium in, 565; ceratopsid bonebeds at, 495; Edmontosaurus excavated at, 543; Montanoceratops cerorhynchus from, 76; William Cutler at, 542 Drumheller Marine Tongue: ceratopsian distribution and, 100; semi-aquatic ceratopsids and, 199 Drumheller Valley: Pachyrhinosaurus from, 141 Dry Island Buffalo Jump Provincial Park, 190 Duck-billed dinosaurs. See Hadrosaurs Ducks: eye sizes, 312; osteomyelitic infections in, 348 Dunhuang, 60 Duration: of ceratopsian species, 413, 414, 420, 422–423, 428–429 Dyschondroplasia: diagnosing in fossils, 348 Eagle Formation: in ceratopsid stratigraphy, 412 Eagles: eye sizes, 313 Earless monitor: nocturnal lifestyle of, 320 Early Cretaceous epoch: Archaeoceratops oshimai from, 240; Archaeoceratops yujingziensis n. sp. from, 59, 60, 65, 66; Auroraceratops rugosus from, 241; basal ceratopsians from, 222, 387, 390, 391, 429, 430–431, 430; ceratopsian cladistics and, 400; ceratopsian paleobiogeography during, 388; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440, 441, 442; Chaoyangsaurus youngi from, 235; crustal rifting during, 329; in dinosaur paleobiogeography, 389; distribution of ceratopsians during, 389; Hongshanosaurus houi from, 238; Liaoceratops yanzigouen-

sis from, 245; paleogeography during, 394–395, 396–398, 396; psittacosaurs from, 23, 330; Psittacosaurus from, 328; Psittacosaurus species from, 238; semiaquatic dinosaurs from, 329 Early Jurassic epoch: ceratopsian paleoenvironmental associations and taphonomy in, 439 East Coulee, Alberta, 190 Eaton, Jeffery G., 122 Eberth, David A., xiii, xiv, xvii, xxi, 99, 141, 428, 495 Echolocation: in nocturnal birds, 323 Ecological niches: among ceratopsids, 420–421 Ecological traits: in ceratopsids, 420–421 Ecology: insect scavenging in, 517; for Kaiparowits ceratopsids, 491–492; relative eye size and, 310–311, 311 Ecosystems: for high-latitude dinosaurs, 456–457; Mesozoic versus Cenozoic, 317; species diversity and turnover in ceratopsid, 422–423 Ecotypes: avian eye size and, 312, 313, 314, 315, 316, 321 Ectopterygoids: Archaeoceratops yujingziensis n. sp., 61, 62; basal ceratopsian, 225; Diabloceratops eatoni n. gen. & sp., 124, 126; Diabloceratops n. gen., 130; in Diabloceratops n. gen. ontogeny, 135; Pachyrhinosaurus n. sp., 145 Edmonton Formation: bonebeds in, 453 Edmonton Group: ceratopsids from, 198; CMN 8547 from, 189; geologic correlation, 84 Edmontonia: in Rattlesnake Mountain microsites, 525, 527–528 Edmontonia rugosidens: in Rattlesnake Mountain microsites, 528 Edmontonian Land Vertebrate Age (LVA): in Big Bend National Park, 521; Ojoceratops fowleri n. gen. & sp. from, 178 Edmontosaurus: from De-na-zin Member, 170; psittacosaurs versus, 44; skeleton excavated by Loris Russell, 543; skull, 555 Edmontosaurus annectens: eye size and body mass of, 317 Edmontosaurus regalis, 8 Eels: psittacosaurs versus, 335 Egernia inornata: nocturnal lifestyle of, 320 Egernia striata: nocturnal lifestyle of, 320 Egg clutches: in bonebed formation, 453. See also Nesting Egg Mountain fauna, 416 Eggs: pupation chambers mistaken for, 513

Eggshells: in Agujaceratops bonebeds, 520; in Big Bend microsites, 533, 534; from Big Bend National Park, 521, 524; in Rattlesnake Mountain microsites, 524–525, 528 Egypt: Peter Dodson in, 10 Einiosaurus: bone histology, 259; bonebeds, 451, 452, 566; in centrosaurine evolution, 163, 165; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; ceratopsid species diversity and turnover and, 419; Diabloceratops eatoni n. gen. & sp. versus, 128; frill, 132; immature versus mature bone surface texture in, 252; Medusaceratops lokii n. gen. & sp. versus, 183; from Montana, 181; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153–154; Pachyrhinosaurus n. sp. versus, 147, 152; paleopathologies in bonebed specimens, 376; phylogeny, 405, 411; provenance, 431; Rubeosaurus ovatus n. gen. & comb. versus, 157, 158, 159, 160, 161, 162– 163; stratigraphy, 163; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216; taxonomy, 408 Einiosaurus procurvicornis: bonebeds, 448, 449, 450, 450; in centrosaurine evolution, 163; in ceratopsian cladistics, 409; ceratopsid species diversity and turnover and, 419; frill, 164; occurrence of, 413, 414; Rubeosaurus ovatus n. gen. & comb. versus, 156, 159, 160, 161, 162–163, 165, 166; skull, 406; skull and cladogram, 295, 296, 297; skull strengths and measurements, 299, 300, 301, 302, 303, 304–305, 304; stratigraphy, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; from Two Medicine Formation, 156 Ejin Qi, 60 Ejinhoro Formation: ceratopsian paleoenvironmental associations and taphonomy in, 440; psittacosaurs from, 32 El Picacho chasmosaurine: Coahuilaceratops magnacuerna n. gen. & sp. versus, 111 Elbows: in Chasmosaurus locomotion, 350, 350, 351–352; modeling in locomotion studies, 351–352 Elephants: step cycle of, 341 Elevator Flats ceratopsian: Zuniceratops christopheri and, 92 Elmisaurus: from Nemegt Formation, 322 Embryonic dinosaurs: in Agujaceratops bonebeds, 525

Endemicity/endemism: ceratopsid, 405, 417, 419 Endotherms: biogeography of, 418 Endothermy: ceratopsid forelimbs and, 341 Endplate pathologies, 368, 369, 370–371 Energy dispersive X-ray detector (EDX), 273; chemical analysis results via, 273– 275, 274, 275–276 England: Calgary Public Museum collection in, 547; Mesozoic paleogeography of, 398; William Cutler’s move from, 541 Enthesophytes, 348 Environmental context: Protoceratops, 321–322, 322 Environmental fluctuations: as driving ceratopsid evolution, 420–421 Eoceratops: fate of skeleton after collection, 542–546, 546–547; recovery of Cutler’s ‘‘missing,’’ 541–550; taxonomy, 542 Eolian settings: basal ceratopsians in, 429; in ceratopsian paleoenvironments, 428; for Djadokhta Formation, 432–433; for ‘‘fighting dinosaurs’’ fossil, 432; insect trace fossils with Protoceratops skeletons from, 509 Eotriceratops: CMN 8547 versus, 197; in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 287; Coahuilaceratops magnacuerna n. gen. & sp. versus, 109; Medusaceratops lokii n. gen. & sp. versus, 185; nasal pathologies in, 359; Ojoceratops fowleri n. gen. & sp. versus, 177; provenance, 431 Eotriceratops xerinsularis, 565; bonebeds, 449; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 445; in ceratopsid fossil record, 419; in CMN 8547 stratigraphy, 197; occurrence of, 416; Ojoceratops fowleri n. gen. & sp. versus, 169, 177, 178; skull, 407; skull and cladogram, 295, 296; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412; taxonomy, 408 Epijugals: Diabloceratops eatoni n. gen. & sp., 117, 122, 124, 131; Diabloceratops n. gen., 129, 134; Montanoceratops cerorhynchus, 69, 71, 71, 77; Pachyrhinosaurus n. sp., 145, 147, 148; pathologies in, 359– 360, 360, 361; Tatankaceratops sacrisonorum n. gen. & sp., 204, 209; Triceratops, 272, 557

index

587

Epinasals: Diabloceratops eatoni n. gen. & sp., 117, 123, 131; in Diabloceratops n. gen. ontogeny, 134–135; Ojoceratops fowleri n. gen. & sp., 174; at Sierra Skull Triceratops locality, 558; Tatankaceratops sacrisonorum n. gen. & sp., 206, 207; Triceratops, 557 Epiparietals: defined, 102; Kaiparowits ceratopsid B, 488; Kaiparowits ceratopsid C, 489; from Mansfield Bonebed, 185–186; Medusaceratops lokii n. gen. & sp., 181; Medusaceratops n. gen., 182, 183–185, 184; in Ojoceratops fowleri n. gen. & sp. taxonomy, 177, 178 Epiphyseal fusion: relevance to ontogeny of fossil vertebrates, 252 Episquamosals: CMN 8547, 190, 191; Coahuilaceratops magnacuerna n. gen. & sp., 108; defined, 102; Ojoceratops fowleri n. gen. & sp., 172, 173; in Ojoceratops fowleri n. gen. & sp. taxonomy, 177, 178 Epoccipitals: centrosaurine, 163, 164; Diabloceratops eatoni n. gen. & sp., 117, 122, 128, 131, 132, 132; Diabloceratops n. gen., 128–129; in Diabloceratops n. gen. ontogeny, 134, 135; in Diabloceratops n. gen. taxonomy, 135–136; in Ojoceratops fowleri n. gen. & sp. taxonomy, 177, 178; Kaiparowits ceratopsid B, 488; Kaiparowits ceratopsid C, 489; Ojoceratops fowleri n. gen. & sp., 171– 172; Rubeosaurus ovatus n. gen. & comb., 159, 160–161, 160, 162; Tatankaceratops sacrisonorum n. gen. & sp., 208, 209, 209; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 214. See also Exoccipitals Equijubus normani: from Gongpoquan Basin, 59 Erdtmanipollis procumbentiformis: as KikakTegoseak Quarry palynomorph, 461, 462 Erect stance: in ceratopsid locomotion, 341–342, 349–350, 350–351 Erickson, Greg: on Developmental Mass Extrapolation, 5 Escalante, 119 Escalante River, 119 Escherichia coli: osteomyelitis and, 348 Estuaries: semi-aquatic animals in, 199 Etiology: of Chasmosaurus irvinensis holotype pathologies, 343, 344, 346–349; of fossil diseases, 340–341 Euoplocephalus: from Rattlesnake Mountain microsites, 528 Euoplocephalus tutus: from Dinosaur Park Formation, 421 Eurasia: ceratopsian distribution in, 389–

588 index

392, 398–399; Mesozoic paleogeography of, 394–396, 395, 396–398, 396, 397 Europe: basal ceratopsians from, 387; Cedar Mountain fossils versus those of, 394; in ceratopsian dispersal, 399, 401; ceratopsian distribution in, 389–392, 399; in ceratopsian paleobiogeography, 388; in dinosaur paleobiogeography, 389; Mesozoic paleogeography of, 394– 396, 395, 396–398, 396, 397; neoceratopsians from, 392, 393; Triceratops museum collections in, 552 Eutherians: Grand Staircase–Escalante localities and distribution, 119 Evans, David C., xxi; on dinosaur sexual dimorphism, 5 Evanston Formation: ceratopsian paleoenvironmental associations and taphonomy in, 446; Triceratops from, 551 Evaporative cooling: arid environments and, 323 Evolution: basal ceratopsian jaws, 234– 250; ceratopsian frill, 269; ceratopsian paleobiogeography and, 388; ceratopsid, 405–427, 409, 410; ceratopsid fossil record and, 418–419; ceratopsid species diversity and turnover and, 419–422, 422–423; large eye size, 310–311; paleoenvironments and, 414; psittacosaur, 329 Evolutionary plasticity: of sympatric ceratopsians, 293 Exoccipitals: Diabloceratops eatoni n. gen. & sp., 127; Ojoceratops fowleri n. gen. & sp., 172; Pachyrhinosaurus n. sp., 148. See also Epoccipitals Exostosis: osteomyelitic infections and, 348 Extant animals: surviving limb fractures/injuries among, 372–373, 373 External mandibular fenestra: psittacosaur, 54 Extinction: biogeography and, 418; of ceratopsids, 420–421; neoceratopsian, 431 Eye sockets. See Orbits Eyes: advantages and disadvantages of large, 310–311; body mass versus sizes of, 312–313, 313, 315, 316, 317, 318; evolution of large, 310–311; of highlatitude vertebrates, 457; nocturnal lifestyle and, 310, 311, 311, 315, 317–323; paleobiological significance of relative size of, 309–311; of predators, 310, 311, 311, 315–317; Protoceratops, 308–327, 309, 310, 319; Troodon, 311; Tyrannosaurus rex, 311

Facies associations (FAs): at Kikak-Tegoseak Quarry, 460–467 Facies-related bias: in collecting Montana Triceratops, 551–563 Faguspollenites granulatus: as KikakTegoseak Quarry palynomorph, 461, 462 Falcarius utahensis: in Mesozoic paleogeography, 398 Falconiforms: eye sizes, 312, 313; studies of sympatric extant, 293 Family groupings: in neoceratopsian bonebeds, 433 Fantastia (Disney), 3 Farke, Andrew A., xvii, xxi, 99, 127, 264, 447; Peter Dodson and, 6, 7 Farlow, James O.: Peter Dodson and, 6, 10 Faunal provincialism: in Late Cretaceous North American biogeography, 417 Faunas: in Agujaceratops bonebeds, 520– 537; hadrosaur-ornithomimid, 322; at Kikak-Tegoseak Quarry, 467–471; Protoceratops, 322–323, 322 Fe-oxide depletion coatings: at KikakTegoseak Quarry, 466, 467, 469 Fe-oxide depletion zones: at KikakTegoseak Quarry, 467 Fe-oxide nodules: at Kikak-Tegoseak Quarry, 466, 467 Feces. See Coprolites Feeding: ceratopsian mandibular evolution and, 234, 246–248 Felids: studies of sympatric extant, 293 Femora: Archaeoceratops yujingziensis n. sp., 65, 66; CMN 8547, 193, 193, 194, 196; fractured and healed Canis latrans, 373; Montanoceratops cerorhynchus, 76; ornithopod, 331, 332; psittacosaur, 331, 332, 334; in psittacosaur swimming, 335; Zuniceratops christopheri, 95, 96 Fenestra of the lacrimal canal: psittacosaur, 41 Fenestrae: in ceratopsian frill stress analysis, 268; development of, 269; horns and frills in ceratopsian sparring and, 283– 290, 285, 286, 287, 288, 289 Fennec foxes (Fennecus zerda): in arid environments, 323 Fenno-Scandinavian Shield: Mesozoic paleogeography and, 394 Ferns: Prince Creek Formation, 471 Ferruginous features: of Kikak-Tegoseak Quarry floodplain paleosols, 469 Ferruginous vascular casts: on Triceratops braincase, 279; on Triceratops horns, 271, 275–276, 276–278, 277, 278, 279–280

Fibro-osseous tumors: diagnosing in fossils, 349 Fibrolamellar bone tissue, 254, 257, 258 Fibrous bone surface texture, 252 Fibulae: CMN 8547, 193, 193, 194, 196; Montanoceratops cerorhynchus, 76; osteomyelitis in, 347; pathologies in Centrosaurus, 372, 373; Zuniceratops christopheri, 95, 96 Fictovichnus gobiensis: pupation chambers, 513, 517 Fictovichuns parvus: pupation chambers, 513, 517 Field Experience Program, 153 ‘‘Fighting dinosaurs’’ fossil (Protoceratops and Velociraptor), 432 Filter feeders: eye sizes of avian, 312, 314, 315 Finches: sympatric Galapagos, 294 Fining upward succession (FUS): at KikakTegoseak Quarry, 460 Finite element modeling (FEM), 265; limitations of, 268–269; of Triceratops frill, 264, 265–267, 268–269 Fiorillo, Anthony R., xvii, 45; on Avaceratops lammersi, 7–8; Peter Dodson and, 6, 76 Fish: binocular vision among, 321; Grand Staircase–Escalante localities and distribution, 119; in Hilda mega-bonebed, 502, 505; in Kikak-Tegoseak Quarry bonebed, 456, 467; psittacosaurs versus, 335; in Rattlesnake Mountain microsites, 524, 525 Fitness-related traits: in ceratopsids, 420, 421, 422 FitzGerald, Gerry: Peter Dodson and, 4 Flagstaff: Zuniceratops christopheri and, 92 Flaig, Peter P., xvii, 456 Flaming Cliffs: dinosaur expeditions to, 509; dinosaurs from, 320; Protoceratops andrewsi collection from, 5, 308 Flank-butting behavior: ceratopsian, 355, 377 Flexor tendons: of semi-aquatic ceratopsids, 200 Flooding: in Big Bend microsite formation, 533, 534; ceratopsian bonebeds and, 452, 456; Hilda mega-bonebed and, 495, 497, 504, 505; Kikak-Tegoseak Quarry bonebed and, 456; in neoceratopsian taphonomy, 433 Floodplain: as Kikak-Tegoseak Quarry bonebed paleoenvironment, 456 Floodplain deposits: in Big Bend National Park geology, 521, 533, 534; Kaiparowits

Formation and, 479–480, 483, 484; Triceratops skull from, 272, 278 Floodplain fine deposits: at Kikak-Tegoseak Quarry, 465–466 Floodplain paleosols: at Kikak-Tegoseak Quarry, 466–467, 468–469 Florida International University: David Weishampel at, 8 Fluvial channels: Kaiparowits Formation, 478 Fluvial systems: anastomosed, 471; Hell Creek Formation and, 552; Kaiparowits Formation and, 479–480, 483 Foodstuffs: biogeography and, 418, 423; psittacosaur, 329; of sympatric ceratopsians, 293, 294 Footprints: ceratopsian, 10. See also Trackways Foramen magnum: Diabloceratops eatoni n. gen. & sp., 127 Forces: in ceratopsian skulls, 294–296, 296–298, 297, 298–305, 298, 299, 301, 304 Ford, Tracy L., xiii, xiv, xvii, 328, 566 Forelimbs: of burrowers, 323; ceratopsid, 196, 199, 341–342; Chasmosaurus, 350; CMN 8547, 193, 193, 194, 195, 198, 199; digital reconstruction of, 350–351, 351– 352; Kaiparowits ceratopsid C, 489–490; modeling in locomotion studies, 351– 352; Montanoceratops cerorhynchus, 69; pathologies in ceratopsid, 372; Protoceratops displaying insect damage, 514; psittacosaur, 332–333, 334, 335; in psittacosaur swimming, 335; Psittacosaurus, 328, 335; scale models for studying, 349–350, 351–352; of semi-aquatic ceratopsids, 199, 200; of swimming animals, 332–333, 334 Foremost Formation: in ceratopsid stratigraphy, 412; geologic correlation, 84; Hilda mega-bonebed and, 497; Kaiparowits Formation versus, 479 Forests: in Kikak-Tegoseak Quarry bonebed paleoenvironment, 456, 472 Forster, Catherine A., xii; Peter Dodson and, 6, 7 Fossil bone: preparing thin sections of, 253–254 Fossil bone damage: insect-caused, 509– 520, 512, 513, 514, 515, 516, 517 Fossil record: ceratopsid, 418–419; historical biases in Montana Triceratops, 551– 563; Triceratops, 555–556 Fossil Research and Development Center, 59

Fossil vertebrates: distinguishing ontogenetic stages among, 251–252 Fossils: Kaiparowits Formation lithofacies bearing, 480–481, 480, 482–484; in ‘‘Purple Hill’’ section, 525–527, 526, 527 Fossorial animals: bone modification by, 517 Fowler, Denver, xxi, 560; Ojoceratops fowleri n. gen. & sp. discovered by, 171 Fox, Ed: Protoceratops skeleton excavated by, 509–511, 510, 511, 513–515, 513, 515–517, 515, 517–518 Fox Hills Formation: Hell Creek Formation versus, 553, 553, 554; in Tatankaceratops sacrisonorum n. gen. & sp. biostratigraphy, 205 Fox, Richard C.: Peter Dodson and, 4, 7 Foxes: in arid environments, 323; as burrow dwellers, 323 Fracture(s): in ceratopsians, 355, 374–379; in ceratopsid bonebed bones, 375; in ceratopsid cervicals, 365–366; diagnosing in fossils, 348–349; in Hell Creek Formation fossils, 554–555; in Pachyrhinosaurus parietal, 363; of stressed Triceratops frill, 268; survival among extant animals with, 372–373 Fragmentary specimens: in ceratopsian paleobiogeography, 388–389, 401 Fratercula: as burrow dwellers, 323 Frenchman Formation: ceratopsian paleoenvironmental associations and taphonomy in, 446; juvenile chasmosaurine horns from, 556; Torosaurus latus from, 417; Triceratops from, 551 Freshwater environments: Psittacosaurus from, 329 Frill function. See Behavior; Functional biology Frills: as acromegalic structures, 356; at Afternoon Delight Triceratops locality, 558, 562; centrosaurine, 132, 164; in ceratopsian sparring, 282–283, 283– 290, 285, 286, 287, 288, 289; in ceratopsian systematics, 9–10, 9; CMN 8547, 190–191, 191, 196–197; Coahuilaceratops magnacuerna n. gen. & sp., 103–104, 105, 108, 109; Diabloceratops eatoni n. gen. & sp., 117, 131, 132, 132; Diabloceratops n. gen., 128–129, 129; in Diabloceratops n. gen. ontogeny, 135; evolution of, 420, 421; evolutionary interaction between chasmosaurine horns and, 282–292; fractures in, 375; function, 566; at High Triceratops locality, 558; histology of ontogenetic bone

index

589

Frills (continued) surface texture changes in ceratopsian, 251–263; marginal ossifications of, 102; measurements of chasmosaurine, 284; mechanical properties, 264–265, 266, 267, 268–269, 268; Montanoceratops cerorhynchus, 77; mottled bone surface texture of centrosaurine, 258–259, 259– 260; musculature related to, 364; of New Mexico ceratopsians, 169; Ojoceratops fowleri n. gen. & sp., 171–172; origin and evolution, 290; Pachyrhinosaurus n. sp., 144, 145, 146, 148–149; pathological Monoclonius lowei, 356, 364; pathologies in, 359, 360–362, 362–364, 362, 363; protective function, 282–283; Protoceratops, 309, 323; resorption of, 375; Rubeosaurus ovatus n. gen. & comb., 162; as sexual display structures, 6; structural models of Triceratops, 264–270; Tatankaceratops sacrisonorum n. gen. & sp., 208; Triceratops, 264, 551, 554, 554, 556; Triceratops horridus, 265–266, 265, 266, 267, 268–269, 268; Triceratops versus other neoceratopsian, 264; at Trike II Triceratops locality, 558; Turanoceratops, 391 Frontals: Diabloceratops eatoni n. gen. & sp., 117, 122, 123, 124, 124; Leptoceratops, 87; Montanoceratops cerorhynchus, 69, 72, 72; Ojoceratops fowleri n. gen. & sp., 173; Pachyrhinosaurus n. sp., 147; Prenoceratops sp., 85–88, 86, 87; Tatankaceratops sacrisonorum n. gen. & sp., 205 Fruitland Formation, 170; ceratopsian paleoenvironmental associations and taphonomy in, 445; in ceratopsid stratigraphy, 412; dating, 413; in North American paleogeography, 415 Fukui Prefectural Dinosaur Museum: Medusaceratops lokii n. gen. & sp. material in, 182, 186 Functional anatomy: psittacosaur, 329– 330, 331, 432 Functional biology: ceratopsian, xiii; finite element modeling in, 269; of Triceratops horns, 271–281 Functional morphology: chasmosaurine horns and frill, 282–292; Protoceratops, 309; sclerotic rings, 309–311, 311–312, 311 Fungi: osteomyelitis and, 348 Fusion: of ceratopsian vertebrae, 366, 367, 368–369, 369–370, 370, 371–372, 371 Fusion of epiphyses: relevance to ontogeny of fossil vertebrates, 252

590 index

Gait: ceratopsid, 342, 351–352 Galapagos Islands: sympatric species on, 294 Galapagos swallowtail gull: nocturnal lifestyle of, 320 Galliforms: eye sizes, 312 Gallimimus: from Nemegt Formation, 322 Game birds: eye sizes, 312 Gangloff, Roland A., xvii, 456 Gansu Province: Archaeoceratops localities in, 60; Archaeoceratops oshimai from, 240; Auroraceratops rugosus from, 241; basal ceratopsians from, 222; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440; new dinosaur discoveries in, 11, 59–60 Gansu Provincial Museum, 59 Gaps: in ceratopsid fossil record, 418–419 Gar: in Big Bend microsites, 533; in Rattlesnake Mountain microsites, 524 Garbani, Harley: baby Triceratops skull found by, 556 Garfield County: Hell Creek Formation and, 552; Triceratops from, 552; Triceratops skull from, 272, 272 Garudimimus brevipes: eye size and body mass of, 317, 318 Gastonia burgei: in Mesozoic paleogeography, 398 Gastroliths: psittacosaur, 329, 333, 432; Psittacosaurus, 328, 333; Psittacosaurus mongoliensis, 32 Gastropod shells: in Big Bend microsites, 533 Gaviiforms: eye sizes, 312 Gazella: chasmosaurines versus, 283 Geckos: binocular vision among, 321; nocturnal, 320 Geese: eye sizes, 312 Geist, Valerius: on display structures, 6 Gekkota: nocturnal, 320 Genera: table of ceratopsian, 439–446 Generic diagnoses: Rubeosaurus ovatus n. gen. & comb., 157. See also Diagnoses (taxonomic) Genetic variation: among ceratopsids, 421 Geococcyx: eye size, 312 Geographic ranges: ceratopsid, 405, 418; of ceratopsid bonebeds, 448, 449, 450 Geologic correlation chart: for the western North American Campanian/ Maastrichtian, 84 Geologic settings: Hilda mega-bonebed, 497–500, 497, 498, 499, 500; Kaiparowits Formation, 479–482, 480; of KikakTegoseak Quarry, 457–458, 458; Mongolian Protoceratops, 511

Geological Society of America, 198 Geology: Big Bend National Park, 521, 523; Hell Creek Formation, 552–553, 552, 553–555, 553, 554, 555 Germany: Mesozoic paleogeography of, 398 Getaway Trike Triceratops locality, 558, 561 Getty, Michael A., xvii, xxi, 99, 478 Gilmore, Charles W.: on New Mexico ceratopsians, 169 Gingerich, Philip: Peter Dodson and, 9 Glacier County: Montanoceratops cerorhynchus from, 70; Rubeosaurus ovatus n. gen. & comb. from, 157 Glen Canyon National Recreation Area, 119 Glyptodontopelta mimus: from Ojo Alamo Formation, 170 Goat horns: thermoregulatory function, 280 Gobi: ceratopsian paleoenvironmental associations and taphonomy in, 442; ceratopsians from, 566; dinosaur expeditions to, 509; discovery of Psittacosaurus in, 21; fossil insect-modified skeletons in, 518; insect trace fossils with Protoceratops skeletons from, 509. See also Bayan Gobi Formation Gobiceratops minutus, 565 Gobiconodontids: in Mesozoic paleogeography, 398 Gobititan shenzhouensis: from Gongpoquan Basin, 59 Goethite: in Triceratops horn vascular casts, 275–276 Gonadal development: relevance to ontogeny of fossil vertebrates, 251 Gongpoquan Basin: Archaeoceratops yujingziensis n. sp. from, 59, 65; dinosaurs from, 66; joint dinosaur expeditions in, 59–60 Goniometer, 344 Goniopholidids: in Rattlesnake Mountain microsites, 525 Goodwin, Mark B., xiv, xvii, 551; historical research by, 565; Triceratops squamosal found by, 554; Trike II Triceratops locality and, 557, 558 Gorgosaurus libratus: at Kikak-Tegoseak Quarry, 467 Gout: diagnosing in fossils, 349 Graciliceratops: in ceratopsian cladistics, 392, 400; in ceratopsian paleobiogeography, 398; in Montanoceratops cladistics, 78, 80, 81; provenance, 387, 389, 390, 431

Graciliceratops mongoliensis: stratigraphy, paleoenvironmental associations, and taphonomic studies, 441 Grand Junction: Zuniceratops christopheri and, 92 Grand Staircase, 119 Grand Staircase–Escalante National Monument (GSENM), 119, 480; ceratopsian family tree and, 134; Diabloceratops eatoni n. gen. & sp. from, 117–140; Kaiparowits ceratopsid A from, 408; taphonomy of ceratopsids from, 478–494; Wahweap centrosaurine from, 408 Grande Prairie: ceratopsian paleoenvironmental associations and taphonomy in, 444; ceratopsid bonebeds at, 495; Pachyrhinosaurus from, 141, 146, 147, 150, 151, 152, 153–154; Pachyrhinosaurus n. sp. bonebed near, 377, 378, 379. See also Pachyrhinosaurus n. sp. (Grande Prairie) Great Depression: Calgary Public Museum collection during, 544 Great Falls, 85 Grebes: eye sizes, 312 Greenhouse episodes: in Big Bend National Park geology, 521 Gregariousness: ceratopsian, 566; Hilda mega-bonebed and ceratopsid, 495, 497, 502, 505; inferred from bonebed studies, 447–448, 451–452; Kaiparowits ceratopsids, 491, 492; in neoceratopsian taphonomy, 433, 434 Gregory, W. K.: Peter Dodson and, 4 Grooves: as cornual sinus vascular structure, 276 Ground squirrels: taphonomy of, 324 Group behavior: inferred from bonebed studies, 451–452, 452–453 Growth plate dysplasia, 348 Growth rates: studies of dinosaur, 5 Growth series: Triceratops, 551 Growths: on ceratopsian phalanges, 374, 374; on ceratopsian vertebrae, 366, 366 Gryposaurus: excavataed by William Cutler, 543–544, 545 Gryposaurus latidens: Diabloceratops eatoni n. gen. & sp. and, 118 Gubik Formation: Prince Creek Formation versus, 457, 460 Gulf of Mexico: Mesozoic paleogeography of, 397 Gulls: eye sizes, 312; nocturnal, 320 Hadrosaur eggshells: in Rattlesnake Mountain microsites, 524

Hadrosaurids: Aguja Formation microsite, 528–530, 529, 530; in ceratopsian biostratigraphy, 152, 153; from De-na-zin Member, 170; KBP discoveries of, 479, 481; nasal functions in, 232; in Rattlesnake Mountain microsites, 525; species diversity and turnover among, 422 Hadrosaurines: from De-na-zin Member, 170; Diabloceratops eatoni n. gen. & sp. and, 118; KBP discoveries of, 479, 481; nasal functions in, 232; species diversity and turnover among, 422 Hadrosauroids: from Gongpoquan Basin, 59; Zuniceratops christopheri and, 92 Hadrosaurs: from Aguja Formation, 521; in Agujaceratops bonebeds, 520; basal ceratopsians versus, 222; in Big Bend microsites, 533, 534; from Big Bend National Park, 521, 525; bone surface texture in, 259; caudal injuries in, 368; dentition, 316; Diabloceratops eatoni n. gen. & sp. and, 118; excavataed by William Cutler, 543–544; eye sizes among, 311; isotopic analysis of, 377–379; in Kikak-Tegoseak Quarry bonebed, 456, 467, 469, 473; in Late Cretaceous North American biogeography, 417; in Mexican stratigraphy, 100; Nemegt, 322; nutritional stress in, 365; pathologies in, 347, 363; Peter Dodson’s work on, 4, 5, 7; in Protoceratops fauna, 322; short species durations among, 420; shoving matches among, 305; skin impressions, 490; supposed aquatic lifestyles of, 328; teeth grinding in hatchling, 528 Hager, Mick: MORT Triceratops and, 555– 556 Hallux valgus, 340, 349, 351, 352 Hanna, Alberta, 190 Happ, John W., xvii, 271 Harding County, South Dakota: Tatankaceratops sacrisonorum n. gen. & sp. from, 204–205 Harmon, Bob: ‘‘B. rex’’ site discovered by, 553–554, 553; at Getaway Trike Triceratops locality, 561 Harris, Jerry, xxi; Peter Dodson and, 10 Hartman, Scott, xvii, 181 Hatcher, John Bell, 565; on finding Triceratops in the field, 555; Triceratops collected and studied by, 551; Triceratops flabellatus discovered by, 271 Hatcher, Joseph, 565 Hatchling dinosaurs: in Agujaceratops bonebeds, 520, 524, 525

Hatchlings: in Big Bend microsites, 533, 534; from Rattlesnake Mountain microsites, 528, 529, 530, 531 Hauterivian stage: ceratopsian cladistics and, 400; ceratopsian paleoenvironmental associations and taphonomy in, 439, 441; Mesozoic paleogeography during, 398 Haversian bone, 255; in finite element modeling, 269; in Triceratops horns, 271, 275, 276 Havre, Montana: Medusaceratops lokii n. gen. & sp. from, 181, 183, 183 Hawks: eye sizes, 313; sclerotic rings, 311; studies of sympatric extant, 293 Haystack Butte bonebed; Zuniceratops christopheri and, 92 Head butting: in dinocephalians, 356; in pachycephalosaurids, 356; in Pachyrhinosaurus, 356 Head pushing contests: in origin and evolution of horns and frills, 290 Healed fractures. See Fractures Healed injuries: horns and frills in ceratopsian sparring and, 282, 283 Hebei: ceratopsian paleoenvironmental associations and taphonomy in, 439 Hedgehogs: in arid environments, 323 Hell Creek chasmosaurine: in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 445; occurrence of, 419–420; stratigraphy, 412. See also Tatankaceratops entries Hell Creek Formation, xiv; bonebeds in, 448, 450–451, 450, 452; ceratopsian localities, 552, 553, 554, 555, 556, 558, 557, 559, 560, 561, 562; ceratopsian paleoenvironmental associations and taphonomy in, 441, 445, 446; ceratopsid species diversity and turnover in, 419– 420, 422; in ceratopsid stratigraphy, 412; chasmosaurine from, 411; geology, 552– 553, 552, 553–555, 553, 554, 555; hatchling ankylosaurs from, 528; Hell Creek Project and, 552; juvenile and subadult Triceratops from, 556; Leptoceratops gracilis from, 243; neoceratopsians/ceratopsids from, 203; in North American paleogeography, 415; prospecting, 553– 555; representative dinosaur fossils from, 555; Tatankaceratops sacrisonorum n. gen. & sp. from, 203–218; Torosaurus latus from, 417; Triceratops from, 551, 552; Triceratops skull from, 271, 272, 272, 273

index

591

Hell Creek Project, 552–553, 552, 553, 554, 554, 555, 556, 557, 559, 560, 561, 562; results of, 558 Hellbenders: caudal fins of, 335 Helopanoplia: in Rattlesnake Mountain microsites, 524, 525 Henderson, Donald M., xvii, xxi, 293 Hepatozoa: osteomyelitis and, 348 Herbaceous shrubs: Prince Creek Formation, 471 Herbivores: biogeography of, 418; eye sizes of avian, 312, 313, 314, 315; eyes of, 308, 311, 314; Nemegt and Djadokhta, 322; Protoceratops as, 308, 315–316; species diversity and turnover among, 422–423; studies of sympatric extant, 293 Herding behavior: ceratopsian, 566; inferred from bonebed studies, 447–448, 451–452, 495–508 Herons: eye sizes, 312; nocturnal, 320 Heterolithic sediments: at Kikak-Tegoseak Quarry, 460, 461, 471 Hierarchal linked systems: for modeling ceratopsid limb movements, 351 High Triceratops locality, 558, 560 High-latitude bonebeds, 456–477, 458, 459 High-latitude localities: dinosaurs in, 456– 457 Hilda, Alberta: Centrosaurus apertus megabonebed near, 377, 495–508, 496, 497, 498, 499, 500, 503, 504 Hilton, Dick: Sierra Skull Triceratops locality found by, 558, 561 Hindlimbs: ceratopsid, 196, 199; CMN 8547, 193, 193, 194, 196, 198, 199; Kaiparowits ceratopsid C, 489–490; Montanoceratops cerorhynchus, 69, 76; pathologies in ceratopsid, 372–374, 373; in Protoceratops burrowing, 323–324; psittacosaur, 330, 331, 332, 333, 334, 335; in psittacosaur swimming, 335; Psittacosaurus, 328, 335; Psittacosaurus mongoliensis, 330, 330, 335; Psittacosaurus neimongoliensis, 334; Psittacosaurus ordosensis, 39; Psittacosaurus xinjiangensis, 330, 330; of semi-aquatic ceratopsids, 199, 200; versus ceratopsid forelimbs, 341 Hippopotamus: nares and orbits of, 334 Hippopotamus amphibious: chasmosaurine behavior as resembling that of, 189, 199, 200 Hirundinids: eye sizes, 312 Histology: ceratopsian, xiii, 566; of ceratopsian bone surface texture changes during ontogeny, 251–263, 256, 257; of

592 index

ceratopsian phalanges, 343; in diagnosing ceratopsid paleopathology, 377–379 Historical biases: in collecting Montana Triceratops, 551–563 Historical research: on ceratopsians, 565– 566 Hollow casts: as cornual sinus vascular structure, 276, 277 Holmes, Robert, xvii, 189, 340 Holocene epoch: Prince Creek Formation and, 457 Holtz, Thomas R., Jr.: on dinosaur paleobiogeography, 389 Hongshanosaurus, 7; as basal ceratopsian, 222, 222; basioccipital, 223; choanae, 231; discovery, 11, 21; mandible, 238, 239; palatines, 228; pterygoids, 227; skull and mandible, 234; systematic paleontology, 25; taxonomy, 23, 429; vomers, 230 Hongshanosaurus houi: anatomical studies, 221; basicranium and palate, 226; basioccipital, 224; coronoid/mandible measurements, 248; discovery, 21; mandible, 238, 239; mandibular element measurements for, 236; provenance, 430; stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; studied specimens, 222, 223, 235; systematic paleontology, 25–26; taxonomy, 23–24 Hoplochelys: in Rattlesnake Mountain microsites, 524, 525 Horn function. See Behavior; Functional biology Horn sheaths: in chasmosaurine intraspecific interaction analysis, 284; keratin in, 278, 280; Triceratops, 271, 273–275; Triceratops flabellatus, 271 Horned dinosaurs: this book and, xiii–xiv, 565–567. See also Ceratopsian entries; Ceratopsid entries Horned mammals: origin and evolution of horns among, 290; Triceratops versus, 288 Horner, John R., xiv, xvii, 156, 551, 565; Fox Protoceratops and, 510; historical research by, 565–566 Horns: at Afternoon Delight Triceratops locality, 558, 562; asymmetry of Pentaceratops, 286, 287; basal ceratopsid, 418; Centrosaurus apertus 502; Ceratops montanus, 182; in ceratopsian sparring, 282–283, 283–290, 285, 286, 287, 288, 289; in ceratopsian systematics, 9–10, 9; ceratopsid sparring and, 375; from Cerro del Pueblo Formation, 102, 103; Coahuilaceratops magnacuerna n. gen. & sp.,

103, 105, 105, 106, 108; CPC 278, 102, 103; Diabloceratops eatoni n. gen. & sp., 117, 120–122, 122, 123, 124, 124, 131; Diabloceratops n. gen., 130; evolution of, 420, 421; evolutionary interaction between chasmosaurine frills and, 282– 292; Fox Protoceratops, 513; frills as defenses against, 269; function, 566; juvenile chasmosaurine, 556; Kaiparowits ceratopsid B, 488; Kaiparowits ceratopsid C, 489, 489; Magnirostris, 133, 133; from Mansfield Bonebed, 186–187; measurements of chasmosaurine, 284; Montanoceratops cerorhynchus, 68; from Ojo Alamo Formation, 169; in Ojoceratops fowleri n. gen. & sp. taxonomy, 177; origin and evolution, 290; outer bone layer of, 271, 275, 275, 276, 277; paleopathologies in, 356; pathologies in, 357– 359, 359; pathology in Chasmosaurus belli, 344; pathology in Chasmosaurus irvinensis, 344; resorption of, 375; Rubeosaurus ovatus n. gen. & comb., 156, 157– 158, 158, 162; scanning electron microscopy of, 273; structure and function of Triceratops, 271–281; Styracosaurus ovatus, 156, 162; suggested functions, 271; Tatankaceratops sacrisonorum n. gen. & sp., 204, 205, 206–207, 206, 207, 208, 209; thermoregulatory function, 271– 272, 280, 323; Torosaurus, 356; Triceratops, 264, 271, 272, 356, 551, 554; at Trike II Triceratops locality, 558; Turanoceratops, 391; Zuniceratops christopheri, 91, 92–93, 93, 94, 94, 95, 95, 96, 96, 392 Horseshoe Canyon Formation: basal neoceratopsians from, 83; Big Bend microsites versus, 533; bonebeds in, 448, 450; ceratopsian distribution in, 100, 110, 111, 113; ceratopsian paleoenvironmental associations and taphonomy in, 443, 444, 445; in ceratopsid stratigraphy, 412; ceratopsids from, 196–197; chasmosaurines from, 408; CMN 8547 from, 189–202; in CMN 8547 stratigraphy, 191, 196–197; geologic correlation, 84; Montanoceratops cerorhynchus from, 69, 76; in North American paleogeography, 415; Pachyrhinosaurus canadensis from, 152; Pachyrhinosaurus from, 141; sympatric ceratopsians from, 303 Horsethief Formation/Sandstone: in ceratopsid stratigraphy, 412; geologic correlation, 84 Host beds: of Hilda mega-bonebed, 500– 502, 500

Houcheng Formation: ceratopsian paleoenvironmental associations and taphonomy in, 439 Humans: dinosaur diseases versus those of, 341; gout in, 349; joint pathology in, 340, 349, 351, 352 Humeri: in ceratopsian posture, 10; in Chasmosaurus locomotion, 350, 350, 351– 352; CMN 8547, 189, 193, 193, 194, 195, 195; dermestid damage to, 517; Horseshoe Canyon ceratopsid, 198; pathologies in ceratopsid, 372; psittacosaur, 332–333, 334; in tetrapod locomotion, 341; Zuniceratops christopheri, 95, 96 Hunt, ReBecca K., xviii, xxi, 447 Hunter Wash, 170, 170 Hyaenas: sympatric Miocene, 293 Hydromorphic soils: Kaiparowits Formation, 481 Hyoids: psittacosaur, 43–44 Hypacrosaurus: Rubeosaurus ovatus n. gen. & comb. and, 161 Hypacrosaurus stebingeri: Prenoceratops sp. and, 85, 88 Hyperflexed hindlimbs: of psittacosaur skeletons, 330, 330, 331 Hyperostosis: in ceratopsian vertebrae, 366 Hyperostotic bone surface texture, 346 Hypothesis testing: in chasmosaurine intraspecific interaction analysis. 283, 284, 285, 289–290 Hypsiglena: nocturnal lifestyle of, 320 Hypsilophodintids: in high-latitude ecosystems, 457 Hypsilophodon: basicranium and palate, 231; femur, 331, 332; in Montanoceratops cladistics, 78, 80, 81; Psittacosaurus versus, 333; RFTRA systematics of, 9 Hypsilophodon foxii: eye size and body mass of, 317, 318; femur, 332 Hypsilophodontians: burrowing by, 323; KBP discoveries of, 479, 481 Hypsilophodontids: osteomyelitis in, 347; pathology in, 347; species diversity and turnover among, 422 Ibex: injuries from play behavior in, 375 Ibises: eye sizes, 312 Ichthyosaurs: gastroliths found with, 333; immature versus mature bone texture among, 252; relative eye sizes, 311; sclerotic rings among, 312 Idaho ceratopsian, 391 Iddesleigh, Alberta: Hilda mega-bonebed and, 497; Pachyrhinosaurus n. sp. from, 141, 142, 143

Idiopathic skeletal hyperostosis: in ceratopsian vertebrae, 366 Iguanids: pushing contests among, 290 Iguanodon: femur, 331, 332; Psittacosaurus versus, 333 Iguanodon bernissartensis: femur, 332 Iguanodon hilli: in Mesozoic paleogeography, 398 Iguanodon ottingeri: in Mesozoic paleogeography, 398 Iguanodontians: Craspedodon versus, 399 Iguanodontids: in Mesozoic paleogeography, 398 Iguanodontoideans: from Xinminpu Group, Mazongshan, 59 Ilek Formation: ceratopsian paleoenvironmental associations and taphonomy in, 440; psittacosaurs from, 34 Ilia: CMN 8547, 193, 193, 194, 195; Montanoceratops cerorhynchus, 75, 76. See also Pelvic girdles Image brightness: eyesight resolution and, 309–310, 310 Imageware Surfacer software, 266 Immature bone: surface texture, 252, 255, 258 Immature individuals: in bonebeds, 452– 453; among Hell Creek neoceratopsians, 203; Hell Creek Triceratops, 556; psittacosaur, 45–52, 47, 48, 49, 50–51; in psittacosaur taxonomy, 23–24, 25–26; Psittacosaurus xinjiangensis, 38; Tatankaceratops sacrisonorum n. gen. & sp., 205, 207–208; in Triceratops intraspecific interactions, 288. See also Baby dinosaurs; Hatchlings; Juveniles Immersion Microscribe point digitizer, 266 In-life resting postures: of fossil terrestrial vertebrates, 330 Incisive foramen: psittacosaur, 42 Inclined heterolithic stratification (IHS): Dinosaur Park Formation, 499; at KikakTegoseak Quarry, 460, 461, 471 Incomplete data: in ceratopsian paleobiogeography, 388–389, 401 Indeterminate chasmosaurine postcranium. See CMN 8547 India: Mesozoic paleogeography of, 397 Indian Ocean: Mesozoic paleogeography of, 395, 396, 397 Individual variation: Triceratops, 551, 562; in Triceratops intraspecific interactions, 288 Indonesia: Mesozoic paleogeography of, 397 Infectious agents: osteomyelitis and, 348

Infectious synovitis: diagnosing in fossils, 346, 347–348 Inferential confidence hierarchy: in diagnosing dinosaur pathologies, 341 Inflammation: of bones and joints, 346 Inflammatory reactions: in nonmammalian vertebrates, 341 Ingenia: from Nemegt Formation, 322 Injuries: to centrosaurine squamosals, 361–362; among ceratopsians, 355, 356, 374–379; horns and frills in ceratopsian sparring and, 282, 283, 285; in origin and evolution of horns and frills, 290; from play behavior, 375; survival among extant animals with, 372–373; in Triceratops intraspecific interactions, 288 Inland seas: North American, 394–395, 397. See also Turgai Sea; Western Interior Seaway Inland terrestrial habitats: as ceratopsid paleoenvironments, 414–416, 420 Inner horncores: of Triceratops horns, 276, 278, 279 Inner Mongolia: Archaeoceratops localities in, 60; basal ceratopsians from, 222; ceratopsian paleoenvironmental associations and taphonomy in, 439, 442; ceratopsians from, 132; Protoceratops from, 308; Protoceratops hellenikorhinus from, 5. See also Nei Mongol Autonomous Region Insect scavenging: during decomposition, 515–517; future research on, 517–518; of Kaiparowits ceratopsid B bones, 487; of Protoceratops skeletons, 509–520, 512, 513, 514, 515, 516, 517. See also Dermestid beetles Insect trace fossils: Protoceratops skeletons with, 509–520, 512, 513, 514, 515, 516, 517 Institute of Vertebrate Paleontology and Paleoanthropology (IVPP): Peter Dodson at, 11 Integrative geological/paleontological studies: of neoceratopsian bonebed taphonomy, 433 Integricorpus: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Integumentary structures: among KBP dinosaur discoveries, 479; psittacosaur, 335, 337; Psittacosaurus, 328, 329 Interbedded sandstone and siltstone facies association: at Kikak-Tegoseak Quarry, 465 Intercoronoids: Archaeoceratops yujingziensis n. sp., 63, 63; Auroraceratops rugosus, 242; Leptoceratops gracilis, 243

index

593

Internal premaxillary foramen: psittacosaur, 42 Interpterygoid vacuity: psittacosaur, 42 Interspecies recognition: ceratopsian horns in, 271; ceratopsid horns and frills and, 304; chasmosaurine horns and frills and, 282 Interspecific competition: among sympatric ceratopsians, 293, 294 Intimidation: in origin and evolution of horns and frills, 290 Intraspecies recognition: ceratopsian horns in, 271; ceratopsid horns and frills and, 304, 323; chasmosaurine horns and frills and, 282 Intraspecific interactions: agonistic, 374– 379; chasmosaurine, 283–290, 285, 286, 287, 288, 289; inferred from bonebed studies, 447–448; injuries from, 355, 356, 374–379 Intraspecific variation: relevance to ontogeny of fossil vertebrates, 252 Invertebrates: in Calgary Public Museum collection, 546; ceratopsid evolution versus, 421; Kaiparowits Formation, 481, 481 Iren Dabasu: dinosaurs from, 322, 322 Iron concretions: in Hell Creek Formation, 554 Iron depletion: in ‘‘Purple Hill’’ section, 525, 526 Iron oxide: fossil insect burrows and, 510, 511, 512 Iron oxide depletion coatings: at KikakTegoseak Quarry, 466, 467, 469 Iron oxide depletion zones: at KikakTegoseak Quarry, 467 Iron oxide nodules: at Kikak-Tegoseak Quarry, 466, 467 Ischia: CMN 8547, 193, 193, 194, 195– 196; Montanoceratops cerorhynchus, 75; Zuniceratops christopheri, 95, 96, 96. See also Pelvic girdles Isolated elements: abundance of Kaiparowits, 481 Isolated macrosites: Kaiparowits Formation, 480, 482, 483 Isotopic analyses: in diagnosing paleopathologies, 377–379 Isotropic bone: in finite element modeling, 269 Israel: extant sympatric felids in, 293 Jackrabbits: in arid environments, 323 Jacobs, Louis, xviii, 456 Japan: psittacosaurs from, 26; Psittacosaurus species from, 238

594 index

Javelina Formation: Aguja Formation versus, 521, 524, 525; bonebeds in, 448, 450, 451; ceratopsian paleoenvironmental associations and taphonomy in, 446; in ceratopsid stratigraphy, 412; lithostrigraphy, 523; Torosaurus latus from, 417; tyrannosaurid teeth from, 531 Javkhlant: dinosaurs from, 322 Javkhlant Formation: ceratopsian paleoenvironmental associations and taphonomy in, 442 Jaws: Archaeoceratops yujingziensis n. sp., 61, 61–64, 63; basal versus advanced ceratopsian, 246–248; ceratopsian, 421; Coahuilaceratops magnacuerna n. gen. & sp., 104, 105, 106, 108; evolution of basal ceratopsian, 234–250; modeling, 294–296, 296–298, 297; Montanoceratops cerorhynchus, 73–74; neonate psittacosaur, 48; peccary, 323; Protoceratops, 309, 315–316, 319; psittacosaur, 43, 45, 329; Psittacosaurus mazongshanensis, 38– 39; Psittacosaurus sattayaraki, 39–40; Zuniceratops christopheri, 92, 93, 94, 95 Jehol fauna: basal ceratopsians from, 221 Jehol Group: dinosaurs from, 59 JEOL JSM-840 scanning electron microscope, 273 Ji, Qiang, 11 Jianshangou: psittacosaurs from, 22–23 Jiayuguan, 60 Jiufotang Formation: ceratopsian paleoenvironmental associations and taphonomy in, 439; psittacosaurs from, 29, 32 Jiuquan, 60 John Henry Member: stratigraphy/dating, 129, 130 Johnson, Gene: discovery of Pachyrhinosaurus n. sp. and, 142, 153 Johnson, H.: Calgary Public Museum collection and, 545, 547 Johnson Storage and Cartage Company Ltd.: Calgary Public Museum collection and, 542, 545 Joint damage: Fox Protoceratops, 513, 514, 515, 516 Joint disease: in chasmosaurine manus, 340, 342 Joints: inflammation of, 346–347, 349; modeling in locomotion studies, 351– 352 Jordan, Montana: Hell Creek Project at, 552; Triceratops skull from, 271, 272, 272 Judith River chasmosaurine: ceratopsian paleoenvironmental associations and taphonomy in, 445; in ceratopsian

cladistics, 410; in ceratopsid fossil record, 419; stratigraphy, 412. See also Medusaceratops entries Judith River Formation: bonebeds in, 448, 450; ceratopsian paleoenvironmental associations and taphonomy in, 443, 445; in ceratopsid stratigraphy, 412; ceratopsids from, 7, 181–182; chasmosaurine from, 411; dating, 413; Kaiparowits Formation versus, 479; Medusaceratops lokii n. gen. & sp. from, 181–188; in North American paleogeography, 415; paleoenvironments of, 415, 417 Judith River Group: ceratopsid teeth from, 530 Judithian dinosaur fauna: Peter Dodson’s work on, 7 Judithian Land Vertebrate Age (LVA): Avaceratops from, 203, 207; in Big Bend National Park, 521; ceratopsid species diversity and turnover during, 419–422; Kaiparowits Formation from, 479 Jugal fossa: psittacosaur, 42 Jugal horn: psittacosaur, 41, 46 Jugals: at Afternoon Delight Triceratops locality, 558, 562; Archaeoceratops yujingziensis n. sp., 61, 62; Diabloceratops eatoni n. gen. & sp., 117, 122, 123–124, 124– 125, 131; Diabloceratops n. gen., 129; Kaiparowits ceratopsid B, 488; Montanoceratops cerorhynchus, 69, 70–71, 71, 77; mottled bone surface texture of centrosaurine, 258–259; ornamentation on, 282; Pachyrhinosaurus n. sp., 145, 145, 147–148; pathologies in, 359–360, 360, 361; Protoceratops displaying insect damage, 515; psittacosaur, 329; Psittacosaurus sp., 26–28, 27; Shenandoah University Triceratops, 273; Tatankaceratops sacrisonorum n. gen. & sp., 209, 210; Zuniceratops christopheri, 91, 92–93, 93 Junggar Basin: psittacosaurs from, 38, 330 Jurassic period: ceratopsian cladistics and, 400; recent basal ceratopsian discoveries from, 221; sauropods from, 555 Juveniles: Aguja Formation microsite tyrannosaurid, 530–531, 530; of alligators and crocodiles, 8; Avaceratops lammersi as, 8; basal ceratopsian basicranium and palate. 225, 226, 228, 230–231, 231; in Big Bend microsites, 533; in bonebeds, 452–453; Brachyceratops montanensis, 161; from Cerro del Pueblo Formation, 102; distinguishing via bone surface texture, 251, 252, 253; excavataed by William Cutler, 543–544, 545; among Hell

Creek neoceratopsians, 203; Hell Creek Triceratops, 556; injuries from play behavior in, 375; Liaoceratops, 222, 223; Liaoceratops yanzigouensis, 245; Mexican ceratopsian, 99; Montanoceratops cerorhynchus, 69; Ojoceratops fowleri n. gen. & sp., 172, 174, 175; psittacosaur, 45–52, 47, 48, 49, 50–51; in psittacosaur taxonomy, 23–24; Psittacosaurus, 331; Psittacosaurus lujiatunensis, 223; Psittacosaurus mongoliensis, 330; tahonomy of neoceratopsian, 433; at Sierra Skull Triceratops locality, 558, 561; Tatankaceratops sacrisonorum n. gen. & sp., 205, 215; Triceratops, 551, 558; in Triceratops intraspecific interactions, 288. See also Immature individuals Kaiparowits Basin/Plateau, 119; Diabloceratops n. gen. skull from, 128–129; prospecting on, 118, 119 Kaiparowits Basin Project (KBP), 479 Kaiparowits ceratopsid A, 484. 485; biogeography, 111; discovery, 478, 484; in ceratopsian cladistics, 409; occurrence of, 413–414; stratigraphy, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; taxonomy, 418; taxonomy and distribution, 408 Kaiparowits ceratopsid B, 484–488, 486, 491; biogeography, 111; bonebeds, 448, 450, 450; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 445; ceratopsid species diversity and turnover and, 420; Coahuilaceratops magnacuerna n. gen. & sp. versus, 111; discovery, 478, 484; occurrence of, 413; stratigraphy, 412; taxonomy, 408–411 Kaiparowits ceratopsid C, 489–490, 489; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 445; ceratopsid species diversity and turnover and, 420; discovery, 478, 489; occurrence of, 413; skin impressions, 490; stratigraphy, 412; taxonomy, 411 Kaiparowits Formation, 119; bonebeds in, 448, 450, 450; ceratopsian family tree and, 134; ceratopsian paleoenvironmental associations and taphonomy in, 443, 445; in ceratopsid stratigraphy, 412; ceratopsids from, 413; chasmosaurines from, 408–411; fauna from, 422; new ceratopsids from, 478; in North American paleogeography, 415; radiometric

dating, 411–413; stratigraphy, 478–479, 479–482, 480; taphonomy of ceratopsids from, 478–494; undiagnosed ceratopsids from, 490, 491; in Wahweap Formation stratigraphy/dating, 131 Kakapo: nocturnal lifestyle of, 320, 323 Kanab, 119 Kemerovo Province: psittacosaurs from, 34 Kennedy Coulee: Medusaceratops lokii n. gen. & sp. from, 182, 183 Keratin: in ceratopsian horn sheaths, 278, 280 Key, Archie: Calgary Public Museum collection and, 544 Khamryn-Us: psittacosaurs from, 329, 330 Khermeen Tsav: ceratopsian paleoenvironmental associations and taphonomy in, 440, 441; dinosaurs from, 322 Khodzhakul Formation: ceratopsian paleoenvironmental associations and taphonomy in, 440, 441 Khukhtek Formation (Khukhtekskaya Svita): ceratopsian paleoenvironmental associations and taphonomy in, 439; psittacosaurs from, 32 Khulsan: ceratopsian paleoenvironmental associations and taphonomy in, 441; dinosaurs from, 322 Khulsyngoskaya Svita: ceratopsian paleoenvironmental associations and taphonomy in, 439 Kikak Creek: Kikak-Tegoseak Quarry near, 457 Kikak-Tegoseak Quarry, 457–458, 458, 459, 460; bonebed in, 456–477, 458, 459; discovery and excavation, 457; locality map, 459; palynology, 458–459, 461, 462, 463, 464 Killik River: Kikak-Tegoseak Quarry and, 459 Kimmeridgian stage: ceratopsian cladistics and, 400 Kingbird: eye size, 312 Kingfishers: eye sizes, 312 Kingsnakes: nocturnal lifestyle of, 320 Kinosternoidea: in Rattlesnake Mountain microsites, 525 Kirkland, James I., xiii, xviii, xxi, 91, 117, 387, 509 Kirtland Formation, 170; ceratopsian paleoenvironmental associations and taphonomy in, 445, 446; ceratopsians from, 169, 177; in ceratopsid stratigraphy, 412; dating, 413; dinosaur faunas from, 170; dinosaurs from, 170; in North American paleogeography, 415

Kirtlandian Land Vertebrate Age (LVA): in Big Bend National Park, 521 Kit foxes: in arid environments, 323 Kiwi: as burrower/digger, 323; nocturnal lifestyle of, 320, 323 Kneehills Creek: William Cutler at, 542 Knees: psittacosaur, 331 Kogosukruk Tongue: of Prince Creek Formation, 457 Krauss, David A., xviii, 282 Kritosaurus: from Aguja Formation, 521; from De-na-zin Member, 170 Kritosaurus zone (Aguja Formation): in Big Bend National Park, 521; dating, 413 Kulceratops: provenance, 431 Kulceratops kulensis: stratigraphy, paleoenvironmental associations, and taphonomic studies, 441 La Popa Basin, 101 Labriform fish: psittacosaurs versus, 335 Lacrimal canal: neonate psittacosaur, 48; psittacosaur, 41; Psittacosaurus, 52 Lacrimals: basal ceratopsian, 228; Diabloceratops eatoni n. gen. & sp., 117, 124; Ojoceratops fowleri n. gen. & sp., 175, 175; Pachyrhinosaurus n. sp., 145, 147; psittacosaur, 41; Tatankaceratops sacrisonorum n. gen. & sp., 209–210, 210 Lacustrine deposits: at Kikak-Tegoseak Quarry, 465; neoceratopsians in, 431; psittacosaurs in, 329, 330, 429–430, 430–431, 432; Psittacosaurus from, 328 Laevigatosporites: as Kikak-Tegoseak Quarry palynomorph, 462 Laevigatosporites perine: as Kikak-Tegoseak Quarry palynomorph, 462 Lag deposits: Triceratops in, 551, 554, 558 Laiyang beds: Psittacosaurus sinensis from, 330, 331 Lake Powell, 119, 480 Lamaceratops: Diabloceratops n. gen. versus, 132; provenance, 431; skull and mandible, 234 Lamaceratops tereschenkoi: stratigraphy, paleoenvironmental associations, and taphonomic studies, 441 Lambe, Lawrence: Peter Dodson and, 4 Lambeosaurines: bone surface texture in, 259; from De-na-zin Member, 170; in Mexican stratigraphy, 100; Peter Dodson’s work on, 5; species diversity and turnover among, 422 Lambeosaurus lambei, 8; in ceratopsian biostratigraphy, 152; eye size and body mass of, 317, 318

index

595

Lambeosaurus magnicristatus: from Alberta, 152–153 Lamellar bone tissue, 251, 254, 257, 258 Laminated siltstone facies association: at Kikak-Tegoseak Quarry, 465 Lammers family, 7 Lampropeltis: nocturnal lifestyle of, 320 Lancc Formation: ceratopsian paleoenvironmental associations and taphonomy in, 445, 446; bonebeds, 450; ceratopsians from, 215; Leptoceratops gracilis from, 243; sympatric ceratopsians from, 302–303; Triceratops flabellatus from, 271; Triceratops from, 551, 555, 558 Lancian Land Vertebrate Age (LVA): Naashoibito Member in, 170–171; Tatankaceratops sacrisonorum n. gen. & sp. from, 205 Landmasses: in biogeography, 416; in ceratopsian paleobiogeography, 389, 392– 396, 395, 396, 397; ceratopsids and, 414 Landslide Butte Field Area: Rubeosaurus ovatus n. gen. & comb. from, 157 Langston, Wann, Jr., 565; on centrosaurine skull pathology, 356; Hilda megabonebed studies by, 496; Pachyrhinosaurus collected by, 141; Peter Dodson and, 4 Laniids: eye sizes, 312 Lanius: eye size, 312 Lanthanotus borneensis: nocturnal lifestyle of, 320 Laramide orogeny: ceratopsids and, 416 Laramidia: ceratopsid distribution in, 99, 111; ceratopsids and, 414–415, 418, 419 Laramie Formation: ceratopsian paleoenvironmental associations and taphonomy in, 446; Triceratops from, 551 Large body size: biogeography and, 418 Laricoidites magnus: as Kikak-Tegoseak Quarry palynomorph, 462 Larines: eye sizes, 312 Larson, Peter L., xviii, 203 Las Encinas Formation: stratigraphy, 101 Las Imagenes Formation: stratigraphy, 101 Las Magenes Formation: in ceratopsid stratigraphy, 412 Laser fusion argon-argon analysis: in dating Kaiparowits formation, 411–413 Laser scanner, 266 Last Chance Canyon: Diabloceratops eatoni n. gen. & sp. from, 122 Last Chance Creek, 119, 121; Diabloceratops eatoni n. gen. & sp. from, 118, 121

596 index

Last Chance skull. See Diabloceratops eatoni n. gen. & sp.; UMNH VP 16699 Late Cretaceous epoch: Agujaceratops bonebeds from, 520–537; ankylosaurs from, 528; basal ceratopsians from, 387, 429, 430; basal neoceratopsians from, 431; in Big Bend National Park geology, 521, 522, 523; biogeography during, 416–418; bonebeds from, 448–451, 450, 452, 456–477, 458, 459; centrosaurine mega-bonebed from, 495–508, 496, 497, 498, 499, 500, 503, 504; centrosaurines during, 141; ceratopsian cladistics and, 400, 409, 410; ceratopsian paleoenvironmental associations and taphonomy in, 440, 441, 442–446; ceratopsid biogeography during, 405, 406; ceratopsid cladistic analysis and, 110; ceratopsid distribution during, 99–100; in ceratopsid fossil record, 418–419; ceratopsid species diversity and turnover during, 419–422, 422–423; ceratopsid stratigraphy during, 412; chasmosaurines during, 340–354; Diabloceratops eatoni n. gen. & sp. from, 117–140; in dinosaur paleobiogeography, 389; dinosaurs from, 322; distribution of ceratopsians during, 389–392; fossils in Calgary Public Museum collection, 546; insect trace fossils with Protoceratops skeletons from, 509–520, 512, 513, 514, 515, 516, 517; latest ceratopsians from, 428; Leptoceratops gracilis from, 243; Medusaceratops lokii n. gen. & sp. from, 181–182; Mexican ceratopsians from, 99–116; in Mexican stratigraphy, 101; neoceratopsian taphonomy in, 433, 434; neoceratopsians from, 96, 430; North American ceratopsids from, 181, 414–416; North American dinosaurian faunas from, 83; North American paleogeography during, 415; Ojoceratops fowleri n. gen. & sp. from, 169–180; paleogeography during, 394, 395–396, 397, 398, 401; paleopathologies in bonebed specimens from, 376; paleopathologies in ceratopsids from, 355–384; pathologies in abelisaurs from, 368–369; Protoceratops andrewsi from, 246, 308; Protoceratops from, 321; sympatric theropods from, 293; taphonomy of Kaiparowits ceratopsids from, 478–494; Tatankaceratops sacrisonorum n. gen. & sp. from, 203– 218; Triceratops, from, 551, 552, 556; Triceratops skull from, 271, 272, 272 Late Jurassic epoch: basal ceratopsians

from, 387, 389, 390, 391, 429, 430; ceratopsian cladistics and, 400; ceratopsian paleobiogeography during, 388; ceratopsian paleoenvironmental associations and taphonomy in, 439; crustal rifting during, 329; earliest ceratopsians from, 428; paleogeography during, 392. 394, 395, 401; recent basal ceratopsian discoveries from, 221; sympatric theropods from, 293; Yinlong downsi from, 236 Late Paleozoic: dinocephalian behavior during, 356 Laterosphenoids: CPC 278, 102; Diabloceratops eatoni n. gen. & sp., 127 Laurasia: ceratopsians from, 428–429 Laysan albatross: eye size, 312; nocturnal lifestyle of, 320 Lea Park Formation: geologic correlation, 84 Leaellynasaura amicagraphica: in highlatitude ecosystems, 457 Lecaniella: as Kikak-Tegoseak Quarry palynomorph, 462 Lee, Andrew, xxi Leidy, Joseph: Peter Dodson and, 4 Leiolopisma suteri: nocturnal lifestyle of, 320 Lenticular sandstone facies association: at Kikak-Tegoseak Quarry, 461–465 Lenticular sandstone lithofacies: Kaiparowits Formation, 480, 480, 481 Lepisostids: in Rattlesnake Mountain microsites, 525 Leptoceratops, 83; as basal ceratopsian, 222, 222; basioccipital, 223; basisphenoid, 224; bonebeds, 452–453; in ceratopsian cladistics, 392, 400, 409; ceratopsian mandibles versus that of, 247; in ceratopsian paleobiogeography, 398; frontals, 87; functions of cranial ornamentation in, 282; in Montanoceratops cladistics, 78, 80, 81; mandible, 243– 245, 244, 246; Montanoceratops cerorhynchus versus, 68–69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79; Montanoceratops versus, 68–69, 70; Prenoceratops sp. versus, 85, 87, 88; Protoceratops andrewsi versus, 246; provenance, 387, 390, 391, 431; RFTRA systematics of, 9; from Scollard Formation, 83–85; skull and mandible, 234; stratigraphy, paleoenvironmental associations, and taphonomic studies, 441; studied specimens, 223; taxonomy, 429; Udanoceratops versus, 391; Wayne Barlowe painting of, 10

Leptoceratops cerorhynchus: taxonomy, 68– 69. See also Montanoceratops cerorhynchus Leptoceratops gracilis: Auroraceratops rugosus versus, 242; basicranium and palate, 228; basioccipital, 225; coronoid/mandible measurements, 248; from Hell Creek Formation, 203; mandible, 243– 245, 244; mandibular element measurements for, 236; Montanoceratops cerorhynchus versus, 68; Rubeosaurus ovatus n. gen. & comb. versus, 159, 161, 166; skull and cladogram, 295, 296, 297; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 304, 305; stratigraphy, paleoenvironmental associations, and taphonomic studies, 441; studied specimens, 223, 235; in studies of ceratopsian sympatry, 294 Leptoceratops sp.: from Dinosaur Park Formation, 88 Leptoceratopsids: ceratopsian dispersal and, 399; in ceratopsian paleobiogeography, 398; Craspedodon versus, 399; Diabloceratops eatoni n. gen. & sp. and, 118; dispersal events among, 387; Montanoceratops cerorhynchus among, 68, 69, 76; in Montanoceratops cladistics, 77–79, 78, 80, 81; North American, 68; Sweden neoceratopsian versus, 399 Leptolepidites: as Kikak-Tegoseak Quarry palynomorph, 462 Lesothosaurus: psittacosaur phylogeny and, 55 Lesothosaurus diagnosticus: eye size and body mass of, 317, 318 Lethbridge, 85 Lethbridge Coal Zone: ceratopsian paleoenvironmental associations and taphonomy in, 443; pachyrhinosaur from, 408; Pachyrhinosaurus n. sp. and, 141, 143, 143, 152, 153; Prenoceratops sp. from, 85 Lewis Shale: in ceratopsid stratigraphy, 412 Li, Daqing: Peter Dodson and, 11, 12 Liaoceratops: Archaeoceratops yujingziensis n. sp. versus, 65–66; as basal ceratopsian, 222, 222; basioccipital, 223; basisphenoid, 223, 224; in ceratopsian cladistics, 392, 399, 400; choanae, 231; discovery, 11; mandible, 244, 245–246; Montanoceratops cerorhynchus versus, 70, 73, 76; in Montanoceratops cladistics, 78, 80, 81; palatines, 228, 229, 231; Prenoceratops sp. versus, 87; provenance, 387, 389, 390, 430; Psittacosaurus versus, 41, 52, 53–54; pterygoids, 225, 226, 227;

skull and mandible, 234; studied specimens, 223; vomers, 230 Liaoceratops yabzigouensis: anatomical studies, 221; basicranium and palate, 227, 231; basioccipital, 224; coronoid/ mandible measurements, 248; mandible, 244, 245–246; mandibular element measurements for, 236; Psittacosaurus versus, 53; skull, 230; stratigraphy, paleoenvironmental associations, and taphonomic studies, 441; studied specimens, 222, 223, 235 Liaoning Province: basal ceratopsians from, 222; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440, 441; Chaoyangsaurus youngi from, 235; Hongshanosaurus houi from, 238; Liaoceratops yanzigouensis from, 245; psittacosaurs from, 22, 23, 26, 28, 29, 329, 330, 333; psittacosaurs with bristle-like structures from, 335, 337; Psittacosaurus from, 11–12, 329; Psittacosaurus taphonomy in, 432 Life habits. See Behavior Life reconstructions: Diabloceratops eatoni n. gen. & sp., 137; Medusaceratops lokii n. gen. & sp., 186; Rubeosaurus ovatus n. gen. & comb., 165 Liliacidites: as Kikak-Tegoseak Quarry palynomorph, 462 Liliacidites variegatus: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Limb-bone-end ossification: relevance to ontogeny of fossil vertebrates, 252 Line diagrams: in chasmosaurine intraspecific interaction analysis, 284 Linear clay concentrations: at KikakTegoseak Quarry, 469 Lines of arrested growth (LAGs), 259 Lisangou Formation: ceratopsian paleoenvironmental associations and taphonomy in, 439 Liscomb bonebed, 456 Lissamphibians: osteomyelitis in, 348 Literature: ceratopsian, 429, 565 Lithofacies: Kaiparowits Formation fossilbearing, 480–481, 480, 482–484, 491– 492; Triceratops fossil record and, 552 Lithostratigraphy: Big Bend National Park, 523 Little Bow River: Pachyrhinosaurus from, 141 Lizards: from Big Bend National Park, 521; in Big Bend microsites, 533; from Djadokhta Formation, 322; nocturnal, 320; pushing contests among, 290; in

Rattlesnake Mountain microsites, 524, 525; sclerotic rings among, 311; sprawling pose of, 330; studies of sympatric extant, 293 Loads: in finite element modeling, 265, 267; in Triceratops frill, 264–265, 266, 267, 268–269, 268 Localities: Big Bend National Park, 522; of ceratopsid bonebeds, 448, 449, 450; Fox Protoceratops, 509–511, 510, 511; Kaiparowits ceratopsids, 480, 481, 482, 483, 484–491, 485, 486, 489, 490; around Kikak-Tegoseak Quarry, 459; mapping and excavating Kaiparowits, 482, 484; Montana Triceratops, 552, 553, 554, 555, 556, 557, 558, 559, 560, 561, 562 Locomotion: ceratopsian, 421; chasmosaurine, 340–354; digital reconstruction of articulation and movement in studying, 349, 350–351, 351–352; forelimbs in ceratopsid, 341–342; paleopathology diagnoses in understanding, 349; scale model manipulation in studying, 349– 350, 351–352 Locomotor stresses: secondary degenerative changes due to, 341 Loewen, Mark A., xiii, xviii, 99, 405, 478 London: William Cutler’s Eoceratops skeleton in, 542, 546–547 Long-grained bone surface texture, 251, 253, 254–255, 254, 255, 259–260; histology, 256; of immature bone, 258; mottled bone surface texture in, 258 Longrich, Nick, xiii, xviii, 308 Loons: eye sizes, 312 Loranthacites pilatus: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Louisiana State University Museum of Natural Science (LSUMNS): Aguja Formation fossils at, 524 Lourinhanosaurus: gastroliths, 333 Low-latitude localities: dinosaurs in, 456 Lower jaws: neonate psittacosaur, 51; psittacosaur, 43, 45, 46. See also Mandibles; Predentaries Lower Moreno Hill Formation: Zuniceratops christopheri and, 92, 92 Lower Scollard Formation: in CMN 8547 stratigraphy, 191 Lower shale member: in Aguja Formation geology, 523 Lower unit: Kaiparowits Formation, 482 Lucas, Spencer G., xviii, 169 Lujiatun Beds: ceratopsian paleoenvironmental associations and taphonomy in, 439, 440, 441; psittacosaurs from,

index

597

Lujiatun Beds (continued) 22–23, 26, 28; Psittacosaurus taphonomy in, 432 Lujiatun Village: basal ceratopsians from, 222, 223 Lull, Richard Swann, 565; on New Mexico ceratopsians, 169; William Cutler’s Eoceratops and, 547 Lund, Eric K., xviii, 99 Lusk, Wyoming: Triceratops from, 551 Maastrichtian stage: basal ceratopsians from, 390, 391; basal neoceratopsians from, 83, 88, 431; Big Bend microsites and, 533; in Big Bend National Park geology, 521, 522, 523, 524; bonebeds from, 456–477, 458, 459; ceratopsian cladistics and, 400, 409, 410; ceratopsian distribution during, 111; ceratopsian family tree and, 134; ceratopsian paleoenvironmental associations and taphonomy in, 440, 411, 442, 444, 445–446; ceratopsid cladistic analysis and, 110; ceratopsid evolution during, 406; in ceratopsid fossil record, 419, 422; ceratopsid occurrences in, 413; ceratopsid stratigraphy during, 412; chasmosaurines from, 408, 411; in CMN 8547 stratigraphy, 191; dating formations of, 413; geologic correlation of western North American, 84; Kikak-Tegoseak Quarry palynologically dated as, 459; Kikak-Tegoseak Quarry palynomorphs from, 461; Leptoceratops gracilis from, 68, 243; Mexican ceratopsians from, 100; in Mexican stratigraphy, 101; Mongolian dinosaurs from, 322; Montanoceratops cerorhynchus from, 68, 70; Naashoibito Member in, 170–171; neoceratopsian bonebeds from, 433; neoceratopsians from, 431; North American ceratopsids during, 414–415, 416, 417, 422, 423; North American paleogeography during, 415; in North Slope stratigraphy, 460; Ojoceratops fowleri n. gen. & sp. from, 169, 171, 178; pachyrhinosaurs from, 408; Rattlesnake Mountain microsites and, 525; semi-aquatic ceratopsids from, 199; Tatankaceratops sacrisonorum n. gen. & sp. from, 204, 205, 214; Triceratops from, 551; Triceratops skull from, 272, 272 Maastrichtian-Paleocene boundary: Hell Creek Project and, 552, 553, 560; Prince Creek Formation and, 458, 460 MacEwan, Grant: Calgary Public Museum collection and, 544

598 index

Macos Formation: in ceratopsid stratigraphy, 412 Macrofauna: abundance of Kaiparowits, 481 Macrofossil bonebeds: Hilda megabonebed and, 507; Kaiparowits Formation, 480, 482, 483–484 Macronarians: nasal functions in, 232 Macronectes: eye size, 312 Madagascar: insect modification of dinosaur bone from, 509; Mesozoic paleogeography of, 397 Magellanic penguin: as burrow dweller, 323 Magnetic resonance imaging (MRI): in diagnosing paleopathologies, 356, 377 Magnetostratigraphic analysis: ceratopsid cladistic analysis and, 110; dating Mexican Late Cretaceous strata via, 100 Magnirostris, 7; in ceratopsian family tree, 134; Diabloceratops eatoni n. gen. & sp. versus, 117; Diabloceratops n. gen. versus, 132, 133–134; provenance, 387, 389, 431; skull and mandible, 234; Zuniceratops christopheri versus, 96 Magnirostris dodsoni: holotype material, 133; publication of, 11; stratigraphy, paleoenvironmental associations, and taphonomic studies, 441 Magpies: eye sizes, 312 Maiasaura: ceratopsid species diversity and turnover and, 419; occurrence of, 416 Major sandstone lithofacies: Kaiparowits Formation, 480, 480, 481, 482 Majungasaurus: caudal pathologies in, 368–369 Makovicky, Peter J., xviii, xxi, 68; on dinosaur growth rates, 5 Mallon, Jordan C., xviii, 189 Malocclusion: absence in ceratopsids, 365 Mammals: in arid environments, 323; in Big Bend microsites, 533; from Big Bend National Park, 521, 522; ceratopsians versus, 10; ceratopsid evolution versus, 421; ceratopsids versus, 356; dinosaur diseases versus those of, 341; from Djadokhta Formation, 322; Grand Staircase–Escalante localities and distribution, 119; in Late Cretaceous paleogeography, 396, 398; ontogenetic patterns among, 252; osteomyelitis in, 348; in Rattlesnake Mountain microsites, 524, 525; sclerotic rings among, 311; semi-aquatic ceratopsids and, 199; Triceratops versus, 288 Mancicorpus trapeziforme: as KikakTegoseak Quarry palynomorph, 461, 462

Mancil, Ron: Kikak-Tegoseak Quarry discovered by, 457 Mandibles: Archaeoceratops oshimai, 240– 241, 241; Archaeoceratops yujingziensis n. sp., 62–63, 63; Auroraceratops rugosus, 241–243, 241; Chaoyangsaurus youngi, 235–236, 237; evolution of basal ceratopsian, 234–250; Hongshanosaurus houi, 238, 239; Leptoceratops gracilis, 243–245, 244; Liaoceratops yanzigouensis, 244, 245–246; measurements of basal ceratopsian, 236; Montanoceratops cerorhynchus, 68; Pachyrhinosaurus n. sp., 145, 149–150, 151; pathologies in, 364–365, 364; Protoceratops andrewsi, 244, 246; Psittacosaurus, 214, 238–240, 239; Tatankaceratops sacrisonorum n. gen. & sp., 210–211, 211, 212, 213, 214; Yinlong downsi, 236–237, 237; Zuniceratops christopheri, 91 Mandibular fenestrae: neonate psittacosaur, 51; psittacosaur, 54 Mandibular flange: neonate psittacosaur, 48, 51 Manganese concretions: in Hell Creek Formation, 554 Maniraptorans: KBP discoveries of, 479, 481 Manitoba: Calgary Public Museum collection and, 546, 547; letters from Cutler to Woodward from, 548–549 Mansfield Bonebed: Medusaceratops lokii n. gen. & sp. from, 182, 183, 183, 185, 186, 187 Manual phalanges. See Phalanges Manual unguals. See Unguals Manus: available ceratopsid specimens, 343; in Chasmosaurus locomotion, 350– 351, 350; CMN 8547, 190, 195, 197, 198; locomotor behavior from pathological, 340–354; osteomyelitis in theropod and hadrosaur, 347; pathological Chasmosaurus, 342, 374; pathologies in ceratopsid, 372; pathology in Chasmosaurus belli, 344–346, 347; pathology in Chasmosaurus irvinensis, 342, 343–344, 345; psittacosaur, 329, 332, 333, 334; in psittacosaur swimming, 335; Tatankaceratops sacrisonorum n. gen. & sp., 211 Manyberries, Alberta, 183; ceratopsian distribution at, 100, 111; Charles M, Sternberg at, 545 Mapping: Kaiparowits Formation, 482 Maps: paleogeographic, 393–394, 395, 396, 397 Marginal frill ossifications, 102

Marginocephalians: psittacosaurs as, 329 Marine invertebrates: ceratopsid evolution versus, 421 Marine paleoenvironments: ceratopsids and, 414 Marsh, Othniel Charles, 565; Peter Dodson and, 4, 10; on Triceratops horn function, 271; Triceratops named by, 551 Marsupials: Grand Staircase–Escalante localities and distribution, 119 Martin, Larry D., xiii, xviii, 328 Maryland: in ceratopsian paleobiogeography, 401 Maryland neoceratopsian, 391, 393, 395, 399 Mass kills: Hilda mega-bonebed, 497, 505– 506, 506–507; psittacosaur, 432 Massospondylus carinatus: eye size and body mass of, 317 Masticatory system: in ceratopsian systematics, 10 Mating competition: ceratopsian horns in, 271; ceratopsid horns and frills and, 304–305; in Triceratops intraspecific interactions, 288–289 Mating displays: horns and frills as, 282, 377 Mating signals: among ceratopsids, 420– 421 Matrix: surrounding/inside Shenandoah University Triceratops skull, 273–275, 276–278 Matt Lamanna: Peter Dodson and, 3, 10 Matthew, William D.: on semi-aquatic ceratopsids, 199 Mature bone: surface texture, 252, 258 Maturity. See Reproductive maturity; Sexual maturity; Skeletal maturity; Somatic maturity Maxillae: accessory antorbital fenestra and, 132, 133; at Afternoon Delight Triceratops locality, 558, 562; Archaeoceratops yujingziensis n. sp., 61, 61–62, 63; basal ceratopsian, 225, 228, 229; in chasmosaurine intraspecific interaction analysis, 284; Coahuilaceratops magnacuerna n. gen. & sp., 105, 106, 108; Diabloceratops eatoni n. gen. & sp., 122, 123–124, 123, 126; Diabloceratops n. gen., 129, 130; in Diabloceratops n. gen. ontogeny, 135; modeling, 296–298; Montanoceratops cerorhynchus, 69, 70; Pachyrhinosaurus n. sp., 144, 145, 145; pathologies in, 359; Protoceratops, 319; psittacosaur, 41; Shenandoah University Triceratops, 273; Tatankaceratops sacrisonorum n. gen.

& sp., 208; Triceratops, 557; Zuniceratops christopheri, 93, 94 Maxillary (MX) biting. See Bite forces Maxillary fossa: psittacosaur, 41 Maxillary protuberance: psittacosaur, 41 Mazongshan, 60; Archaeoceratops oshimai from, 240; Archaeoceratops yujingziensis n. sp. from, 59–67, 60, 65; Auroraceratops rugosus from, 241; basal ceratopsians from, 223; ceratopsian paleoenvironmental associations and taphonomy in, 440; dinosaurs from, 66, 67; joint dinosaur expeditions in, 59–60 Mazongshan Dinosaur Project, 59 McCarthy, Paul J., xviii, 456 McCarthy, Shawna: Peter Dodson and, 10 McCone County: Hell Creek Formation and, 552; High Triceratops locality in, 558, 560; Sierra Skull Triceratops locality in, 561; Triceratops from, 552 McCrae Formation: Torosaurus latus from, 417 McDonald, Andrew T., xviii, 91, 156 McGuire Creek: High Triceratops locality along, 558 McIntosh, John S.: Peter Dodson and, 7 McKinney Springs tongue: in Aguja Formation geology, 523 Mean annual pressure (MAP): Prince Creek Formation and, 472 Mean annual temperaturwe (MAT): Prince Creek Formation and, 472 Meandering trunk channel facies association: at Kikak-Tegoseak Quarry, 460–461 Measured sections: Aguja Formation, 523, 524 Measured sections: Hilda mega-bonebed, 498, 499, 500 Medial premaxillary foramen: psittacosaur, 42 Medicine Hat, 190 Medullary bone: relevance to ontogeny of fossil vertebrates, 252 Medusaceratops lokii n. gen. & sp., 181–188, 565; description, 183–185; diagnosis, 182, 183; discovery, 181; life reconstruction, 186; locality, 183, 183; other Montana ceratopsids versus, 181–182; stratigraphy/biostratigraphy, 182, 187; systematic paleontology, 182–183; taxonomy, 185–187 Medusaceratops n. gen.: biostratigraphy, 187 Mega-bonebeds: Alberta (Hilda) centrosaurine, 377, 495–508, 503, 504; ceratopsian, 497, 507; in Dinosaur Provin-

cial Park, 495; host bed of Hilda, 500– 502 Megaherbivores: species diversity and turnover among, 422–423 Mei long: eye size and body mass of, 317, 318 Meileyingzi Beds: psittacosaurs from, 29 Memorial Park Library Building: William Cutler’s Eoceratops skeleton and, 542, 543–544 Menefee Formation: centrosaurine squamosals from, 135, 136; neoceratopsians from, 96; in Wahweap Formation stratigraphy/dating, 130 Mergansers: eye sizes, 312 Mergus: eye size, 312 Mesa Southwest Museum: Zuniceratops christopheri at, 92 Mesotarsal joint: psittacosaur, 331 Mesozoic ecosystems, 317 Mesozoic Era: biogeography, 416; ceratopsian paleobiogeography during, 388; dispersal patterns among nonceratopsians during, 396–398; paleogeography during, 392–396, 395, 396– 398, 396, 397, 399–401 Metabolic rates: ceratopsid, 405 Metacarpals: Aguja Formation dromaeosaurid, 529, 532; Aguja Formation microsite, 529; in Chasmosaurus locomotion, 350–351, 350, 351–352; CMN 8547, 197, 198; pathological, 340; pathologies in ceratopsid, 372; pathology in Chasmosaurus irvinensis, 342, 343–344, 345; pathology in Chasmosaurus belli, 344–346, 347, 348; psittacosaur, 332 Metatarsals: Archaeoceratops yujingziensis n. sp., 65, 66; in bipedal locomotion, 331– 332; CMN 8547, 196, 198; Zuniceratops christopheri, 95, 96 Mexican axolotl: caudal fin of, 335 Mexico: basal ceratopsians from, 390; bonebeds in, 448, 450, 452; centrosaurines from, 141, 417; ceratopsian distribution in, 389; ceratopsian paleoenvironmental associations and taphonomy in, 445; ceratopsians from, 99–116, 429, 566; ceratopsid distribution in, 99; in ceratopsid fossil record, 418, 419; ceratopsid stratigraphy in, 412; chasmosaurine from, 408; new dinosaur discoveries in, 12; in pachyrhinosaur biogeoraphy, 417–418; stratigraphy in, 101 Microceratops: as basal ceratopsian, 222 Microceratus. See Microceratops

index

599

Microceratus sulcidens, 565 Microfauna: abundance of Kaiparowits, 481 Microfossil assemblages: in Hilda megabonebed, 502; Kaiparowits Formation, 480, 482, 483, 484; studies of, 507 Microlaminated illuvial clay: at KikakTegoseak Quarry, 466, 467, 468 Micromorphological features: Prince Creek Formation, 464, 465, 466, 467, 468–469, 471–472 Microraptor gui: eye size and body mass of, 317, 318 Microreticulatisporites uniformis: as KikakTegoseak Quarry palynomorph, 461 Microsites: Big Bend National Park, 522– 524, 523, 524–527, 527–528, 528–532, 532–534; formation of Big Bend, 534 Microvenator: discovery, 4 Microvertebrate fossils: from Kaiparowits Plateau, 118, 119 Microvertebrate sites: in Big Bend National Park, 521–524 Microvertebrates: Zuniceratops christopheri and, 92 Mid-latitude localities: dinosaurs in, 456 Middle Cretaceous: paleogeography during, 394, 395, 397 Middle Jurassic epoch: basal ceratopsians from, 387, 389, 390; ceratopsian cladistics and, 400 Middle unit: Kaiparowits Formation, 482 Migration/migrations: ceratopsid, 417, 422, 431; Hilda mega-bonebed and, 495; neoceratopsian bonebeds and, 433 Mihirungs: relative eye sizes, 311 Milk River, 85; Medusaceratops lokii n. gen. & sp. from, 182, 183 Milk River Formation: basal neoceratopsians from, 83; in ceratopsid stratigraphy, 412; geologic correlation, 84; in Wahweap Formation stratigraphy/dating, 130 Minhe Formation: ceratopsian paleoenvironmental associations and taphonomy in, 442 Minimum number of individuals (MNI): in ceratopsid bonebeds, 450; in Hilda mega-bonebed, 495, 505–506 Minor sandstone lithofacies: Kaiparowits Formation, 480, 480, 481, 482 Miocene epoch: Kikak-Tegoseak Quarry palynomorphs from, 461; sympatric hyaenas from, 293 Miocene-Recent mammals: ceratopsid evolution versus, 421

600 index

Mirror Mesa interval: Zuniceratops christopheri and, 92 Missouri River: Ceratops montanus from near, 182 Mixed bonebeds, 524; Agujaceratops in, 520–537 Mixed feeders: eye sizes of avian, 312, 313, 314, 315 Miyashita, Tetsuto, xviii, 83, 565 Models: of ceratopsian skulls, 294–296, 296–298, 297, 301; of Triceratops frill structural properties, 264–270 Molnar, Ralph, xxi Monclova, 101 Mongolia, 60, 510; basal ceratopsians from, 222, 231, 390, 429–430; basal neoceratopsian taphonomy in, 432–433, 434; basal neoceratopsians from, 431; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440, 441; ceratopsian paleoenvironments in, 428; ceratopsian taphonomy in, 432; ceratopsians from, 132, 429, 566; dinosaurs from, 322; discovery of Psittacosaurus in, 21; insect trace fossils with Protoceratops skeletons from, 509–520, 512, 513, 514, 515, 516, 517; new dinosaur discoveries in, 11, 12; Protoceratops andrewsi from, 246, 309, 321–322, 322–323, 322; Protoceratiops from, 432–433, 434; psittacosaurs from, 22, 22, 32, 329, 330 Mongolian Academy of Sciences: dinosaur expeditions to Mongolia by, 509; ‘‘fighting dinosaurs’’ at, 432; Fox Protoceratops and, 510; Protoceratops at, 319; trace fossils at, 512 Mongoliazizhigou: ceratopsian paleoenvironmental associations and taphonomy in, 442 Monitors: nocturnal, 320 Monoclonius: from Cerro del Pueblo Formation, 100; manus, 343; New Mexico frill material versus, 169; Peter Dodson’s systematics of, 9, 10 Monoclonius crassus: from Montana, 182; Peter Dodson and, 10 Monoclonius lowei: pathological frill of subadult, 356, 364; Peter Dodson’s work on, 8, 9 Monoclonius nasicornus: Rubeosaurus ovatus n. gen. & comb. versus, 158 Monoclonius recurvicornus: from Montana, 182 Monoclonius sphenocerus: from Montana, 182 Monodominant bonebeds. See Bonebeds

Montana, 183; Achelousaurus from, 142, 152; American centrosaurines outside, 117; basal neoceratopsians from, 431; Big Bend dinosaurs versus those of, 521, 533; bonebeds in, 448, 450, 451, 452– 453; centrosaurines from, 130, 141; ceratopsian paleoenvironmental associations and taphonomy in, 441, 442, 443, 445, 446; ceratopsid biostratigraphy in, 187; ceratopsid distribution in, 99, 111; in ceratopsid fossil record, 418, 419; ceratopsid species diversity and turnover in, 419, 421; ceratopsid stratigraphy in, 412; ceratopsids from, 7; chasmosaurine from, 411; dating formations of, 413; dinosaur diversity in, 88; dinosaurs from, 118; geologic correlation in, 84; Hell Creek Project localities, 552, 553, 554, 555, 556, 557, 559, 560, 561, 562; historical bias in Triceratops collecting from, 551–563, 552, 553; index map, 552; insect modification of dinosaur bone from, 509; juvenile and subadult Triceratops from, 556; Kaiparowits Formation and, 478; Late Cretaceous ceratopsids from, 181; Leptoceratops gracilis from, 243; locality map, 85; Medusaceratops lokii n. gen. & sp. from, 181–188; Montanoceratops cerorhynchus from, 68, 70; pachyrhinosaurs from, 408; paleoenvironments of, 416, 417; paleopathologies in bonebed specimens from, 376; Peter Dodson in, 10; Prenoceratops sp. from, 83; Rubeosaurus ovatus n. gen. & comb. from, 156–168; Styracosaurus ovatus from, 408; sympatric ceratopsians from, 302–303; Torosaurus latus from, 417; Triceratops skull from, 271, 272, 272 Montana State University: Triceratops studies by, 552 Montanoceratops: from Belly River Group, 83; in ceratopsian cladistics, 392, 400; in ceratopsian paleobiogeography, 398; Leptoceratops versus, 68–69; from Montana, 181; Prenoceratops sp. versus, 88; provenance, 387, 390, 391, 431; taxonomy, 429; from Willow Creek Formation, 85; Zuniceratops christopheri versus, 94 Montanoceratops cerorhynchus, 68–82; cladistic analysis, 77–79, 79–81; diagnosis, 70; holotype material, 69–70, 70– 76; phylogenetic analysis, 69; referred material, 76–77; stratigraphy, paleoenvironmental associations, and taphonomic studies, 441; systematic paleontology, 69–77; taxonomy, 68–69, 77–79

Monterey, 101 MOR 1125. See ‘‘B. rex’’ site MOR 449, 163; stratigraphy, 161–162, 163 MOR 492, 156, 157–159, 157, 158, 159, 160–162, 160, 162, 163, 165; referral to Rubeosaurus ovatus n. gen. & comb., 160–162, 165; stratigraphy, 161, 162 MOR 542: described, 76; in Montanoceratops cladistics, 78, 80, 81 Moreno Hill Formation: ceratopsian paleoenvironmental associations and taphonomy in, 442; in North American paleogeography, 415; Zuniceratops christopheri from, 91, 92, 92, 418 Morphology: chasmosaurine horn and frill, 282–292; Triceratops, 551, 562 Morphometric analysis: in ceratopsian systematics, 9–10, 9; Peter Dodson on, 7 Morrin, Alberta, 190 Morrison Formation: Peter Dodson’s work on, 7; sauropods from, 555 MORT Triceratops skeleton, 555–556, 555 Mother Log stump: Zuniceratops christopheri and, 92 Moths: bone modification by, 515–516 Mottled bone surface texture, 251, 253, 254–255, 254, 255, 259–260; in centrosaurines, 258–259; histology, 256 Mottles: at Kikak-Tegoseak Quarry, 469 Mudstone facies: of Hilda mega-bonebed, 500–502, 500, 504, 505 Mudstone facies association: at KikakTegoseak Quarry, 465–466 Mudstone lithofacies: for Big Bend microsites, 534; Dinosaur Park Formation, 499; Hell Creek Formation, 554–555, 554, 556, 557, 558, 560, 561; for Hilda mega-bonebed, 495, 496, 498, 499; Kaiparowits Formation, 480, 480, 481, 482, 483, 484; in ‘‘Purple Hill’’ section, 525, 526, 527; Triceratops in, 551, 562; undiagnosed Kaiparowits ceratopsid in, 490, 491 Mule deer: in arid environments, 323 Multiple exostoses: osteomyelitic infections and, 348 Multitaxic bonebeds. See Bonebeds Multituberculates: Grand Staircase– Escalante localities and distribution, 119; in Rattlesnake Mountain microsites, 525 Musculature: in bipedal locomotion, 332; in ceratopsian frill stress analysis, 268– 269; in ceratopsian skull and jaw analysis, 294; ceratopsid frills and, 364; in

psittacosaur swimming, 335; Psittacosaurus, 328 Museo del Desierto: Coahuilaceratops magnacuerna n. gen. & sp. at, 104; CPC 278 at, 102; CPC 279 at, 103; fieldwork in Mexico by, 100 Museum of Nature and Science (Alaska): Kikak-Tegoseak Quarry fieldwork by, 457 Museum of Northern Arizona: Wahweap Formation explorations by, 118 Museum of the Rockies (MOR): Fox Protoceratops and, 510; Hell Creek Project by, 552–553, 553, 554; Rubeosaurus ovatus n. gen. & comb. at, 157; Triceratops studies by, 552 Museums: dinosaur specimens for display in, 555 Myledaphus: Pachyrhinosaurus n. sp. stratigraphy and, 143 Naashoibito Member: ceratopsian paleoenvironmental associations and taphonomy in, 445; in ceratopsid stratigraphy, 412; chasmosaurine from, 411; dinosaurs from, 170; Ojoceratops fowleri n. gen. & sp. from, 169–180; stratigraphy/biostratigraphy, 170, 177, 178 Nanhsiungchelyids: in Rattlesnake Mountain microsites, 525 Nares: basal ceratopsian, 221; Kaiparowits ceratopsid C, 489; psittacosaur, 329, 334; Psittacosaurus, 328; of semi-aquatic ceratopsids, 200. See also Nostrils Nasal air flow: basal ceratopsian, 221 Nasal bosses: Achelousaurus, 150–151; pachyrhinosaur, 141, 142, 408; Pachyrhinosaurus n. sp., 145–147, 145, 146, 150–151; Pachyrhinosaurus n. sp. (Grande Prairie), 150–151; pathologies in, 357, 358 Nasal horns/horncores. See Epinasals; Horns; Nasals Nasal region: function in ceratopsians, 232 Nasals: accessory antorbital fenestra and, 132, 133; chasmosaurine intraspecific interactions and, 289; Coahuilaceratops magnacuerna n. gen. & sp., 103, 105– 106, 105, 107; Diabloceratops eatoni n. gen. & sp., 122, 123, 123; in Diabloceratops n. gen. ontogeny, 134–135; MOR 449, 161–162, 163; mottled bone surface texture of centrosaurine, 258–259; Ojoceratops fowleri n. gen. & sp., 172, 174, 175; in Ojoceratops fowleri n. gen. & sp. taxonomy, 177; Pachyrhinosaurus n. sp., 144, 145–147, 145; pathologies in,

357–359, 358, 359; Protoceratops, 309; psittacosaur, 41; Rubeosaurus ovatus n. gen. & comb., 157–158, 157, 158; Shenandoah University Triceratops, 273, 275, 276; Tatankaceratops sacrisonorum n. gen. & sp., 206–207, 208; Zuniceratops christopheri, 91, 93–94, 93 National Museum of Canada: Calgary Public Museum collection and, 544; in Chasmosaurus locomotion studies, 349; CMN 8547 at, 189, 191–192, 192, 194; Dale Russell and Peter Dodson at, 8; Peter Dodson and, 3, 4; posture of ceratopsian specimens at, 10 National Museum of Natural History. See United States National Museum (USNM) National Museum of Natural Science. See National Museum of Canada Natural History Museum (London). See British Museum of Natural History (BMNH) Natural selection: in ceratopsid evolution, 422 Neck protection: Triceratops frill as, 282– 283 Necrobia rufipes: Protoceratops skeleton insect scavenging versus that by, 510– 511 Necrophagous insect traces: Protoceratops skeletons with, 509–520, 512, 513, 514, 515, 516, 517, 517 Necrophagous insects: in ecology, 517 Necrophila americana: Protoceratops skeleton insect scavenging versus that by, 510– 511 Nedcolbertia justinhoffmani: in Mesozoic paleogeography, 398 Nedoceratops. See Diceratops Negative allometry: among bird eyes, 313– 314, 315, 316; among dinosaur eyes, 313–314, 315, 316, 317, 318; of eye size versus body mass, 312–313, 313, 315, 316, 318 Nei Mongol Autonomous Region: Archaeoceratops localities in, 60; psittacosaurs from, 26, 32 Nelson Creek: Sierra Skull Triceratops locality along, 558 Nemegt Formation: dinosaurs from, 322, 322 Nemegtomaia: from Nemegt Formation, 322 Nemegtosaurus: eye size and body mass of, 317; from Nemegt Formation, 322, 322 Nemegtosaurus mongoliensis: eye size and body mass of, 318

index

601

Neoceratopsians, 8–9; Archaeoceratops yujingziensis n. sp. among, 59, 60; Auroraceratops rugosus versus, 241; basicranial and palatal characters, 231– 232; basioccipitals, 223; in ceratopsian paleobiogeography, 398, 423; in ceratopsid fossil record, 418, 419; Chaoyangsaurus youngi versus, 235–236; chronostratigraphy of, 390; Craspedodon versus, 399; Diabloceratops eatoni n. gen. & sp. among, 120, 123, 124, 127, 131, 135; Diabloceratops n. gen. versus, 132; dispersal events among, 387; distribution of, 389–392; from Europe, 392, 393; eyes, 314; functions of cranial ornamentation among, 282–283, 289–290; Hell Creek, 203; Leptoceratops gracilis versus, 243; Liaoceratops yanzigouensis versus, 245; life habits, 428; mandibular evolution, 234, 246–248; Medusaceratops lokii n. gen. & sp. among, 182; Montanoceratops cerorhynchus among, 68, 69, 70, 73, 74–75, 77–78; North American, 181; from Oldman Formation, 83–90; palates, 225–231, 227, 228, 230, 231; palatines, 228, 229; paleoenvironmental associations of, 430–431; paleoenvironments for, 428, 431, 433–434; phylogenetic analysis, 222; Prenoceratops sp. among, 85; psittacosaur/ceratopsid mandibles versus those of, 247–248; psittacosaurs versus, 221; Psittacosaurus versus, 41, 238; species diversity and turnover among, 421; stratigraphy, paleogeography, and diversity of, 430; studied specimens, 222–223; table of genera and species, stratigraphy, paleoenvironmental associations, and taphonomic studies of basal, 440–442; taphonomy, 433; taphonomy of basal, 432– 433, 434; Tatankaceratops sacrisonorum n. gen. & sp. among, 204; taxonomy, 418, 429; from Xinminpu Group, Mazongshan, 59; Triceratops versus, 264; vomers, 230–231; Zuniceratops christopheri among, 91, 96, 392 Neonate skulls: psittacosaur, 45–52, 47, 48, 49, 50–51 Neotological studies: relevance to ontogeny of fossil vertebrates, 251 Nesting: Big Bend microsites and, 533; by Big Bend National Park dinosaurs, 521, 524–525; neoceratopsian bonebeds and, 433; at Rattlesnake Mountain microsites, 525. See also Egg entries; Eggshells Neural spines: Protoceratops caudal, 323

602 index

New Mexico, 170; centrosaurine squamosals from, 135, 136; ceratopsian paleoenvironmental associations and taphonomy in, 442, 445, 446; ceratopsid and other vertebrate remains from, 169–171; ceratopsid biostratigraphy in, 187; ceratopsid distribution in, 99, 111; ceratopsid stratigraphy in, 412; chasmosaurine from, 411; dating formations of, 413; Late Cretaceous stratigraphy of, 170– 171, 177; neoceratopsians from, 96; Ojoceratops fowleri n. gen. & sp. from, 169–180; paleoenvironments of, 417; Torosaurus latus from, 417; Zuniceratops christopheri from, 91, 92, 92; Zuniceratops from, 187, 418, 419, 431 New Mexico Museum of Natural History (NMMNH): dinosaur faunas surveyed by, 170; Ojoceratops fowleri n. gen. & sp. material at, 171 New World vultures: eye sizes, 312 Nguyen, Peter, xviii, 282 Niche partitioning: neoceratopsian skull shapes as indicating, 293–307; among neoceratopsians, 433 Niemi Ranch: Tatankaceratops sacrisonorum n. gen. & sp. from, 204–205 Night herons: eye size, 312; nocturnal lifestyle of, 320 Night lizards: nocturnal lifestyle of, 320 Night skink: nocturnal lifestyle of, 320 Night snakes: nocturnal lifestyle of, 320 Nighthawks: nocturnal lifestyle of, 320 Nightjars: nocturnal lifestyle of, 320 Niobrara County: Triceratops from, 551, 558 Nipple Butte, 119, 120; Diabloceratops n. gen. discovered at, 118, 120, 128 Nipple Butte skull. See Diabloceratops n. gen.; UMNH VP 16704 Nocturnal birds: eyes of, 308, 310 Nocturnal feeding times: diurnal feeding times versus, 293, 432 Nocturnal lifestyle: arid environments and, 323; avian eye size and, 312, 313, 314, 315, 321; binocular vision in, 321; ceratopsian, 566; in desert environments, 308; diapsids leading, 320; eyes and, 310, 311, 311, 315, 317–323; of owls, 323; Protoceratops, 308–327, 432– 433; Troodon, 311 Nodosaurids/nodosaurines: KBP discoveries of, 479, 481; in Rattlesnake Mountain microsites, 525, 527–528; species diversity and turnover among, 422

Nodules: in Big Bend microsites, 533; pupation chambers mistaken for, 513; in ‘‘Purple Hill’’ section, 525, 527 Nomadic Expeditions: dinosaur expeditions to Mongolia by, 509 Nomina dubia: Ceratops montanus 182, 187; from Montana, 182; psittacosaur, 23, 25–26, 38–40; Turanoceratops, 391 Nonavian dinosaurs: KBP discoveries of, 479, 481 Non-ceratopsians: Mesozoic dispersal patterns among, 396–398 Non-mammalian vertebrates: inflammatory reactions in, 341 Non-striated surface texture; of mature bone, 258 Norell, Mark: on dinosaur growth rates, 5 Norian stage: Kikak-Tegoseak Quarry palynomorphs from, 461 North America, 183; Agujaceratops bonebeds in, 520–537, 522; ankylosaurs from, 528; basal ceratopsians from, 387, 429, 431; basal neoceratopsians from, 69, 83–85, 88; ceratopsian bonebed taphonomy in, 431–432, 434; ceratopsian bonebeds in, 566; in ceratopsian dispersal, 399, 401; ceratopsian distribution in, 389–392, 398–399, 406–407; in ceratopsian paleobiogeography, 388, 389; ceratopsian paleoenvironments in, 428; ceratopsian stratigraphy, paleogeography, and diversity in, 430; ceratopsid bonebeds in, 450; ceratopsid distribution in, 99–100, 111; ceratopsid species diversity and turnover in, 419–422, 422–423; ceratopsids and paleogeography of, 414–416, 415; ceratopsids from, 7, 324; ceratopsids outside, 391; Diabloceratops eatoni n. gen. & sp. from, 117– 140; in dinosaur paleobiogeography, 389; geologic correlation of the Campanian/Maastrichtian of western, 84; inland sea of, 394–395, 397; KBP dinosaur discoveries in, 479; Late Cretaceous biogeography of, 416–418; Late Cretaceous ceratopsids from, 181; Late Cretaceous/early Tertiary, 522; Leptoceratops gracilis from, 243; Mesozoic paleogeography of, 394–396, 395, 396–398, 396, 397; Montanoceratops cerorhynchus from, 68; neoceratopsians from, 96, 431; new dinosaur discoveries in, 12; pachyrhinosaur biogeoraphy in, 417–418; Prince Creek Formation and, 457; recent basal ceratopsian discoveries in, 221–222; Triceratops fossil record in, 551, 555–556;

Triceratops museum collections in, 552; Zuniceratops christopheri from, 91–98 North Dakota: ceratopsian paleoenvironmental associations and taphonomy in, 446; Torosaurus latus from, 417; Triceratops from, 551 North Horn Formation: bonebeds in, 448, 450, 451; ceratopsian paleoenvironmental associations and taphonomy in, 446; in ceratopsid stratigraphy, 412; in North American paleogeography, 415; Torosaurus latus from, 417 North Slope: Kikak-Tegoseak Quarry bonebed on, 456–477, 458, 459 North-West Travellers Building: Calgary Public Museum collection at, 544 Northern biome: Big Bend dinosaurs versus those of, 521; bonebeds in, 452; ceratopsid species diversity and turnover in, 419–420, 422; in Late Cretaceous North American biogeography, 416– 417, 419; Rattlesnake Mountain microsites versus, 524–525 Norton, David W., xviii, 456; KikakTegoseak Quarry discovered by, 457 Nostrils: ceratopsian, 232. See also Nares Nothronychus mckinleyi: Zuniceratops christopheri and, 91, 92, 93 Nuchal crest: psittacosaur, 42 Nuiqsut: Kikak-Tegoseak Quarry near, 457 Null hypothesis: in chasmosaurine intraspecific interaction analysis, 284, 285 Numenius: eye size, 312 Nutritional stress: in dinosaurs, 365 NVision Modelmaker scanner, 266 Nycticorax: eye size, 312; nocturnal lifestyle of, 320 Occipital condyles: basal ceratopsian, 223; Ceratops montanus, 182; Diabloceratops eatoni n. gen. & sp., 127; Diabloceratops n. gen., 129, 130; Ojoceratops fowleri n. gen. & sp., 171; Pachyrhinosaurus n. sp. (Alaska), 456, 470, 472; Tatankaceratops sacrisonorum n. gen. & sp., 209, 209 Occipital fossae: psittacosaur, 42 Occiputs (occipita): basal ceratopsian, 224; Diabloceratops eatoni n. gen. & sp., 125, 127; in Diabloceratops n. gen. ontogeny, 134–135; psittacosaur, 224; Psittacosaurus major, 31, 229; Zuniceratops christopheri, 93 Ocean Point: Kikak-Tegoseak Quarry and, 459; Prince Creek Formation and, 458 Oilbirds: nocturnal lifestyle of, 320, 323 Ojo Alamo ceratopsian (chasmosaurine),

411; biogeography, 111; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 445; Coahuilaceratops magnacuerna n. gen. & sp. versus, 111; stratigraphy, 412. See also Ojoceratops fowleri n. gen. & sp. Ojo Alamo Formation, 170; ceratopsian paleoenvironmental associations and taphonomy in, 445; in ceratopsid stratigraphy, 412; dinosaur faunas from, 170; in North American paleogeography, 415; Ojoceratops fowleri n. gen. & sp. from, 169–180; stratigraphy/biostratigraphy, 170, 177, 178; taxonomy of ceratopsid specimens from, 177–178 Ojo Alamo Sandstone, 169 Ojoceratops fowleri n. gen. & sp., 169–180, 565; description, 172; diagnosis, 171; history and discovery of, 169–171; localities, 170, 171; material referred to, 172–176; postcranial material, 172, 176, 176; systematic paleontology, 171–172; taxonomy, 176–177, 177–178 Old World vultures: eye sizes, 312 Oldman Formation: Aguja Formation versus, 527; Albertaceratops from, 181; Albertaceratops nesmoi from, 407–408; basal neoceratopsian from, 83–90; bonebeds in, 448, 450; Centrosaurus brinkmani from, 408; ceratopsian paleoenvironmental associations and taphonomy in, 443; in ceratopsid stratigraphy, 412; geologic correlation, 84; Hilda mega-bonebed and, 497, 497, 498, 499, 499, 500; Kaiparowits Formation versus, 479; in Medusaceratops lokii n. gen. & sp. stratigraphy, 182; in North American paleogeography, 415; stratigraphy, 303 Olfaction: of semi-aquatic ceratopsids, 200. See also Smell Olfactory bulbs: in nocturnal and crepuscular birds, 323 Olson, Everett C.: Peter Dodson and, 3 Omnigov (Omnogov): Fox Protoceratops locality at, 510 Omnivores: eyes of, 308; Protoceratops as, 315, 317 Onefour, Alberta: shoreline deposits at, 507 Ontogenetic stages: distinguishing, 251– 252; Pachyrhinosaurus n. sp. (Alaska), 470–471; in Triceratops intraspecific interactions, 288 Ontogeny: Brachyceratops, 156;

Brachyceratops montanensis, 161; Centrosaurus apertus, 163; ceratopsian, 205, 207–208; chasmosaurine, 185; Diabloceratops n. gen., 134–135; among Hell Creek neoceratopsians, 203; histology of ceratopsian bone surface texture changes during, 251–263; Ojoceratops fowleri n. gen. & sp., 177; Protoceratops andrewsi, 246; Protoceratops in the study of, 308–309; in psittacosaur taxonomy, 23–24; Triceratops, 551, 558, 561, 562; Zuniceratops christopheri, 95–96 Oogenera: in Rattlesnake Mountain microsites, 524 Opisthotics: Diabloceratops eatoni n. gen. & sp., 127 Opisthotony: of psittacosaur skeletons, 330 Optical scanner, 266 Orbits: eyes within, 311; Kaiparowits ceratopsid B, 488; of high-latitude vertebrates, 457; Protoceratops, 319; psittacosaur, 329, 334; Psittacosaurus, 328; of semi-aquatic ceratopsids, 200; Tatankaceratops sacrisonorum n. gen. & sp., 205. See also Eyes Ordos Basin: ceratopsian paleoenvironmental associations and taphonomy in, 440; psittacosaurs from, 32 Ordos Plateau: psittacosaurs from, 330 Organic matter: at Kikak-Tegoseak Quarry, 466, 468 Orientation. See Cranial orientation Ornithiaschia, 25. See also Ornithischians Ornithischian eggshells: in Rattlesnake Mountain microsites, 524 Ornithischians: Archaeoceratops yujingziensis n. sp.among, 60; in Big Bend microsites, 533; bones displaying insect damage, 515, 516, 517; ceratopsids as, 405; CMN 8547 among, 190; CPC 278 among, 102; Diabloceratops eatoni n. gen. & sp. among, 120; eyes of, 308, 314; gastroliths found with, 333; Grand Staircase–Escalante localities and distribution, 119; Medusaceratops lokii n. gen. & sp. among, 182; Montanoceratops cerorhynchus among, 69; Ojoceratops fowleri n. gen. & sp. among, 171; Prenoceratops sp. among, 85; Rubeosaurus ovatus n. gen. & comb. among, 157; sclerotic rings among, 312; Sweden neoceratopsian versus, 399; Tatankaceratops sacrisonorum n. gen. & sp. among, 204 Ornithoid prismatic eggshells: in Rattlesnake Mountain microsites, 524

index

603

Ornithoid ratite eggshells: in Rattlesnake Mountain microsites, 524 Ornithomimids: Aguja Formation, 525, 529, 531; in Agujaceratops bonebeds, 520; in Kikak-Tegoseak Quarry bonebed, 456, 467; Nemegt and Djadokhta, 322; in Rattlesnake Mountain microsites, 525, 529, 531 Ornithomimus: Aguja Formation ornithomimids versus, 531 Ornithopods: Aguja Formation microsite, 528–530, 529, 530; basicranium and palate, 231; femora, 331, 332; finite element modeling of, 265; KBP discoveries of, 479, 481; long-grained bone surface texture in, 258; psittacosaurs versus, 332; Sweden neoceratopsian versus, 399 Orodromeus: occurrence of, 416 Orycteropus: in arid environments, 323 Oryx: chasmosaurines versus, 283 Osborn, Henry Fairfield, 565; discovery of psittacosaurs and, 21; Peter Dodson and, 4 Oshih: psittacosaurs from, 22, 32 Osmólska, Halszka, 232, 565 Osmundaceae: Prince Creek Formation, 473 Osmundacidites wellmanii: as KikakTegoseak Quarry palynomorph, 462 Ossification: relevance to ontogeny of fossil vertebrates, 252 Osteichthyes: Grand Staircase–Escalante localities and distribution, 119; in KikakTegoseak Quarry bonebed, 456, 467 Osteochondritis dessicans: diagnosing in fossils, 348 Osteochondroma: diagnosing in fossils, 348 Osteoderms: from Rattlesnake Mountain microsites, 527, 528 Osteolysis: in ceratopsid caudals, 369 Osteomas: in Pachyrhinosaurus scapula, 372 Osteomyelitic infections: diagnosing in fossils, 348 Osteomyelitis: in Centrosaurus dentary, 365; in ceratopsid caudals, 369; diagnosing in fossils, 346, 347–348, 352; in Pachyrhinosaurus parietal, 363 Osteonal remodeling, 255, 257, 258, 259– 260 Osteons: primary, 254, 256; secondary, 255, 257, 259–260, 275, 276 Osteopathies: among Albertan ceratopsians, 356; ceratopsian, 355; in ceratopsid hindlimbs, 372–373, 373; as evi-

604 index

dence for agonistic behavior, 374–375, 375–376, 376–377; mandibular, 364– 365, 364; tail, 368–372 Osteophytes, 348 Ostrom, John, 565; Peter Dodson and, 3, 4, 6, 7, 9. See also American Journal of Science John Ostrom issue Ott, Christopher J., xviii, 203 Ottawa, 544, 545. See also National Museum of Canada; University of Ottawa Outer bone layer: of Triceratops horns, 275, 275, 276, 277, 278–279, 280 Overbank deposits: neoceratopsian taphonomy in, 433; in ‘‘Purple Hill’’ section, 526, 527 Overbank sediments: Kaiparowits Formation, 478 Overbank settings: for ceratopsid bonebeds, 448–451 Ovibos moschatus: Triceratops versus, 288 Oviraptor: collected from Bayn Dzak, 320, 322 Ovorkhangai: psittacosaurs from, 32 Owlet-nightjars: nocturnal lifestyle of, 320 Owls: nocturnal lifestyle of, 323; sclerotic rings, 311, 320; studies of sympatric extant, 293 Oxfordian stage: ceratopsian cladistics and, 400; Yinlong downsi from, 236 Oxygen isotope studies: thermoregulation demonstrated via, 272 Pachycephalosaurians/pachycephalosaurids/pachycephalosaurs: basicranium and palate, 231; from Big Bend National Park, 521; Diabloceratops eatoni n. gen. & sp. and, 118; head butting in, 356; in Mesozoic paleogeography, 398; species diversity and turnover among, 422 Pachyrhinosaurs: biogeography, 417–418; in ceratopsid fossil record, 418; ceratopsid species diversity and turnover and, 421–422; from Alberta and Montana, 152; high-latitude bonebed, 456– 477, 458, 459; occurrence of, 413; phylogeny, 405; taxonomy, 408 Pachyrhinosaurus, 12; behavior, 566; biogeography, 417–418; bone histology, 259; bonebeds, 452; caudal pathologies in, 368, 370; in centrosaurine cladistics, 164; in centrosaurine evolution, 163, 165; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; in ceratopsid species diversity and turnover, 422; cervical pathologies in, 365,

365; and CMN 8547 as Anchiceratops, 198; flank butting behavior, 377; frill, 132; frills as sexual display structures in, 6; head butting in, 356; immature versus mature bone surface texture in, 252; jugals/epijugal patholigies in, 359–360, 361; at Kikak-Tegoseak Quarry, 467, 469, 473; maxillary pathologies in, 359; Medusaceratops lokii n. gen. & sp. versus, 183; MOR 449 versus, 161; nasal pathologies in, 357, 358; osteopathies of, 355; in Pachyrhinosaurus n. sp. cladistics, 153; parietal pathologies in, 362– 364, 363; pathological skull of, 356; pectoral girdle pathologies in, 372; phalangeal stress fractures in, 373–374; phylogeny, 411; provenance, 431; quadrate pathologies in, 360, 361; rib pathologies in, 367, 369; rostral pathologies in, 357, 357; Rubeosaurus ovatus n. gen. & comb. versus, 162, 163; severe fractures in, 375; squamosal pathologies in, 359, 360– 361; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216; taxonomy, 429; Wayne Barlowe painting of, 10 Pachyrhinosaurus canadensis, 4; bonebeds, 448, 449, 450; in centrosaurine cladistics, 164; in ceratopsian cladistics, 409; in ceratopsid fossil record, 418; CMN 8547 versus, 196–197; Medusaceratops lokii n. gen. & sp. versus, 186; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153–154; Pachyrhinosaurus n. sp. versus, 141, 147, 150, 152; Peter Dodson’s work on, 8; Rubeosaurus ovatus n. gen. & comb. versus, 159, 161, 166; skull, 406; skull/bite strength, 303; skull roof pathologies in, 359; stratigraphy, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; taxonomy, 408 Pachyrhinosaurus lakustai, 565. See also Pachyrhinosaurus n. sp. (Grande Prairie) Pachyrhinosaurus n. sp., 141–155; bonebeds, 448, 449, 450; circumnarial region, 147; description, 144–150; discovery, 141; frill, 144, 145, 146, 148–149; identification as pachyrhinosaur, 141–142; locality, 142–144, 142; mandible, 145, 149–150, 151; manual pathologies in, 372; occurrence of, 413; phylogenetic analysis, 141, 150, 151, 152, 153–154; postcranial elements, 144; skull, 141, 143, 144–150, 145, 146; skull and cladogram, 295, 296, 297–298; skull strengths

and measurements, 299, 300, 301, 302, 303, 304–305, 304; snout, 144–147, 145, 146; stratigraphy, 142–144, 143, 144, 303; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; taxonomy, 150–153, 152, 153–154, 408; temporal region, 147–149 Pachyrhinosaurus n. sp. (Alaska): bonebed, 456–477, 458, 459; ceratopsian paleoenvironmental associations and taphonomy in, 444; occurrence, 141; postcranial elements, 469, 471; skull, 469, 470 Pachyrhinosaurus n. sp. (Grande Prairie), 141; bonebeds, 356, 377, 378, 379; ceratopsian paleoenvironmental associations and taphonomy in, 444; cervical pathologies in, 365; frill, 164; nasal pathologies in, 357; osteopathy in bonebed specimens, 378, 379; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153–154; Pachyrhinosaurus n. sp. versus, 146, 147, 150–151; paleopathologies in bonebed specimens, 376; parietal pathologies in, 362–364, 363; rib pathologies in, 367; Rubeosaurus ovatus n. gen. & comb. versus, 159, 161, 161, 166; skull, 406; skull roof pathologies in, 359 Pacific Ocean: Mesozoic paleogeography of, 395, 396, 397 Pakowki Formation: in ceratopsid stratigraphy, 412; geologic correlation, 84; Hilda mega-bonebed and, 497 Palatal rami: psittacosaur, 42 Palates: Archaeoceratops yujingziensis n. sp., 61, 62; basal ceratopsian, 221, 222–223, 224, 225–231, 226, 227, 228, 231–232, 231; Diabloceratops eatoni n. gen. & sp., 125–126; Diabloceratops n. gen., 129, 130; neonate psittacosaur, 48–51; Pachyrhinosaurus n. sp., 145; psittacosaur, 42; Psittacosaurus, 52; Psittacosaurus major, 29, 30 Palatines: Archaeoceratops yujingziensis n. sp., 61, 62; basal ceratopsian, 222, 225, 227–229; Diabloceratops eatoni n. gen. & sp., 124, 126; Pachyrhinosaurus n. sp., 145 Paleo-BondJ Penetrant Stabilizer, 253, 254 Paleobiogeography, 387–388; ceratopsian, xiii, 387–404; of dinosaurs, 388–389 Paleobiology: Protoceratops 309; psittacosaur, 329–330, 330–331, 331–335, 335–337 Paleocene epoch: in Big Bend National

Park geology, 521, 523; in CMN 8547 stratigraphy, 191; Hell Creek Project and, 552, 553; Kikak-Tegoseak Quarry palynomorphs from, 461; in Mexican stratigraphy, 101; Naashoibito Member in, 170 Paleochannel deposits: neoceratopsian taphonomy in, 433 Paleochannel settings: for ceratopsid bonebeds, 448–451 Paleocommunities: for Big Bend microsites, 532–534 Paleoecology: of Big Bend National Park, 521; ceratopsian, xiii. See also Paleoenviroments; Paleoenvironmental associations Paleoenvironmental associations: basal ceratopsian, 429–430, 430–431; ceratopsian, 428–446; neoceratopsian, 430– 431; table of ceratopsian, 439–446 Paleoenvironments: for Big Bend microsites, 532–534; for ceratopsids, 414–416, 428, 439–446; fossil insect-modified skeletons and, 518; for Hilda megabonebed, 503–505; for Kaiparowits ceratopsids, 491–492; Kaiparowits Formation, 479–482, 480, 481; Kikak-Tegoseak Quarry bonebed, 456–477; North American ankylosaurs in, 528; paleosols and, 467; for Protoceratops, 309, 321–322, 322–323, 322; for Psittacosaurus, 328– 329 Paleogene: marine transgressions lasting into, 414 Paleogeographic distribution: ceratopsian, 429, 430 Paleogeography: ceratopsian, 430; ceratopsian paleobiogeography and, 388, 392–396, 395, 396, 397; Late Cretaceous/early Tertiary western hemisphere, 522; North America, 414–416, 415 Paleomap Project, 392, 393 Paleontology: of Agujaceratops bonebeds, 520; finite element modeling in, 265 Paleopathologies (paleopathology): ceratopsian, xiii, 355–384; differential diagnosis in, 340–341, 346–349; in dinosaur research, 355–356; found in ceratopsid bonebeds, 375, 376; future directions in ceratopsian, 377–379; inferring chasmosaurine locomotor behavior using, 340–354. See also Pathology Paleopedology: in and around KikakTegoseak Quarry, 459–460, 463, 464, 466, 467

Paleosols: in Agujaceratops bonebeds, 520; in Big Bend National Park geology, 521; in Big Bend microsites, 533; Kaiparowits Formation, 481; in and around KikakTegoseak Quarry, 459–460, 466–467, 468–469, 472; micromorphological features of, 467; in ‘‘Purple Hill’’ section, 525, 526, 527 Paleozoic Era: fossils in Calgary Public Museum collection, 546 Palpebrals: CPC 278, 102; Diabloceratops eatoni n. gen. & sp., 122, 123, 124; Diabloceratops n. gen., 129; Pachyrhinosaurus n. sp., 147; psittacosaur, 42–43, 44; Psittacosaurus, 52; Psittacosaurus xinjiangensis, 38 Paludal settings: basal ceratopsians in, 429, 431 Palynodata 2000 database, 459 Palynology: Prince Creek Formation, 458– 459, 461, 462, 463, 464. See also Pollen Palynomorphs: in Kikak-Tegoseak Quarry bonebed paleoenvironment, 456, 459, 461, 462, 463, 464 Panoplosaurus rugosidens: from Rattlesnake Mountain microsites, 528 Panzarin, Lukas, 165 Papules: at Kikak-Tegoseak Quarry, 466, 467, 469 Parasaurolophus: from De-na-zin Member, 170 Parasphenoid: basal ceratopsian, 223, 230 Parietal-squamosal shelf: psittacosaur, 42 Parietals: at Afternoon Delight Triceratops locality, 558; bone surface texture, 255; centrosaurine, 164; Coahuilaceratops magnacuerna n. gen. & sp., 108, 109; Diabloceratops eatoni n. gen. & sp., 117, 122, 126–127, 127–128, 131; Diabloceratops n. gen., 128–129, 129; in Diabloceratops n. gen. ontogeny, 135; in Diabloceratops taxonomy, 136; at High Triceratops locality, 560; horns and frills in ceratopsian sparring and, 283–290; Kaiparowits ceratopsid B, 488; Kaiparowits ceratopsid C, 489, 489; from Mansfield Bonebed, 185–186; Medusaceratops lokii n. gen. & sp., 182; Medusaceratops n. gen., 182, 183–185, 184; Montanoceratops cerorhynchus, 77; morphology of chasmosaurine, 282; mottled bone surface texture of centrosaurine, 258–259; Ojoceratops fowleri n. gen. & sp., 171– 172, 173–174, 173, 174; in Ojoceratops fowleri n. gen. & sp. taxonomy, 177; ornamentation on, 282; Pachyrhino-

index

605

Parietals (continued) saurus n. sp., 141, 142, 149, 149; pathological Centrosaurus, 355–356; pathological Pachyrhinosaurus, 375; pathologies in, 362–364, 362, 363, 377; Rubeosaurus ovatus n. gen. & comb., 156, 158–159, 160–161, 160, 162; Shenandoah University Triceratops, 273; Tatankaceratops sacrisonorum n. gen. & sp., 208–209; Triceratops, 551; in Triceratops frill finite element modeling, 267, 268, 269; Zuniceratops christopheri, 91, 93, 94, 96 Parieto-squamosal marginal ossification: defined, 102 Parks, William A.: Arrhinoceratops described by, 198 Parksosaurus warrenae: eye size and body mass of, 317, 318 Parras, 101 Parras Basin, 101; centrosaurines from, 417; Coahuilaceratops magnacuerna n. gen. & sp. from, 104; CPC 278 from, 102; CPC 279 from, 103 Parras Basin Dinosaur Project, 100–102 Parras Shale: in ceratopsid stratigraphy, 412; stratigraphy, 100, 101 Parrot-beaked dinosaurs. See Psittacosaurs Parrotfish: psittacosaurs versus, 335 Parrots: as burrow dwellers, 323; eye sizes, 312 PAST software: in analysis of eye size versus body mass, 313 Pathogens: hosts and, 341; osteomyelitis and, 348 Pathology: bone surface texture and, 258; ceratopsian, 566; ceratopsid manual, 340–354; in Chasmosaurus belli specimen, 344–346, 347; in Chasmosaurus irvinensis holotype, 343, 344, 345, 345, 346, 347; Chasmosaurus manual, 342; Diabloceratops eatoni n. gen. & sp., 127– 128; modeling locomotion studies and, 351–352; of Pentaceratops postorbital horns, 286; Pachyrhinosaurus n. sp., 141, 144. See also Joint damage; Paleopathologies (paleopathology) Patrick, Omer H.: Calgary Public Museum collection and, 544, 545, 546 Patterns of deformation: of stressed Triceratops frill, 267, 268 Paul, Gregory S.: on ceratopsian posture, 10 Pebble conglomerate lithofacies: Kaiparowits Formation, 480–481, 480, 482, 484 Peccaries: in arid environments, 323 Pectoral girdles: in Chasmosaurus locomo-

606 index

tion studies, 349; CMN 8547, 193, 193, 194, 195, 198; Montanoceratops cerorhynchus, 69, 75; pathologies in ceratopsid, 372; of semi-aquatic ceratopsids, 200 Pedal phalanges: Archaeoceratops yujingziensis n. sp., 65, 66; Montanoceratops cerorhynchus, 76. See also Phalanges Pedal unguals. See Unguals Pedes: in Chasmosaurus locomotion, 350; CMN 8547, 196, 198; osteomyelitis in theropod, 347; pathology in Chasmosaurus irvinensis, 344; phalangeal stress fractures in, 373–374, 374; Protoceratops, with burrows, 512; Protoceratops andrewsi, 324; psittacosaur, 331–332, 333; in psittacosaur swimming, 335; Psittacosaurus meileyingensis, 330 Pedogenic carbonate nodules: in ‘‘Purple Hill’’ section, 525, 527 Pedogenic processes: at Kikak-Tegoseak Quarry, 467, 472 Pedorelicts: at Kikak-Tegoseak Quarry, 466, 467, 469 Pedoturbation: at Kikak-Tegoseak Quarry, 467, 472 Peds: at Kikak-Tegoseak Quarry, 468 Pelorosaurus: in Mesozoic paleogeography, 398 Pelvic girdles: CMN 8547, 193, 193, 194, 195–196, 198; Montanoceratops cerorhynchus, 69, 75, 76; pathologies in Centrosaurus, 372; Protoceratops displaying insect damage, 514 Pen Formation: in Aguja Formation geology, 523; in ceratopsid stratigraphy, 412 Penguins: as burrow dwellers, 323; eye sizes, 312; psittacosaurs versus, 332, 334 Penkalski, Paul G., Jr.: on Avaceratops lammersi, 7–8; Peter Dodson and, 6–7, 7 Pentaceratops: biostratigraphy, 187; bonebeds, 452; in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 284, 286, 287; Coahuilaceratops magnacuerna n. gen. & sp. versus, 99, 104, 106, 108, 109, 111; Diabloceratops n. gen. versus, 133; distribution, 100; Medusaceratops lokii n. gen. & sp. versus, 185; phylogeny, 405, 411; provenance, 431; RFTRA systematics of, 9, 10; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216 Pentaceratops sternbergii: biogeography, 111; bonebeds, 449; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 445; CMN 8547 versus, 193,

194, 195, 195, 196, 198; humerus, 195; occurrence of, 414; skull, 407; skull and cladogram, 295, 296; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412; ulna, 197 Periosteal bone: in Diabloceratops n. gen. ontogeny, 135 Periosteal inflammation: chronic, 346 Periosteal vasculature, 258 Petrels: eye sizes, 312 Pezon, Antoine, xviii, 282 Phalacrocoracids: eye sizes, 312 Phalanges: Aguja Formation dromaeosaurid, 529, 532; Aguja Formation microsite, 529, 531; Archaeoceratops yujingziensis n. sp., 65, 66; in bipedal locomotion, 332; Centrosaurus apertus, 343; Chasmosaurus, 343; in Chasmosaurus locomotion, 350, 351–352; CMN 8547, 195, 196, 197, 198; diaphyseal microfractures in, 349; Montanoceratops cerorhynchus, 76; pathological, 340; pathologies in ceratopsid, 372; pathology in Chasmosaurus belli, 344–346, 347; pathology in Chasmosaurus irvinensis, 342, 343–344, 345; psittacosaur, 329, 332; stress fractures in, 349, 352, 356, 373–374, 374; Tatankaceratops sacrisonorum n. gen. & sp., 211 Phasianus colchicus: sclerotic rings, 311 Pheasant: sclerotic rings, 311 Phenetics: in dinosaur paleobiogeography, 389 Phi-Rho-Z program, 273 Phosphate studies: thermoregulation demonstrated via, 272 Photography: in chasmosaurine intraspecific interaction analysis, 284 Photomosaic-controlled cross sections: Hilda mega-bonebed, 497, 500 Photoreceptors: eyesight resolution and, 309, 310; visual sensitivity and, 310 Phylogenetic analyses: Avaceratops lammersi in, 8; of basal ceratopsians, 221, 222–223, 222; in ceratopsian paleobiogeography, 398–399, 400, 409, 410; ceratopsians, psittacosaurs, and neoceratopsians, 222; Coahuilaceratops magnacuerna n. gen. & sp., 108–109; Diabloceratops eatoni n. gen. & sp. 131, 133–134; Montanoceratops cerorhynchus, 69; Pachyrhinosaurus n. sp., 141, 150, 151, 152, 153–154; Rubeosaurus ovatus n. gen. & comb., 156, 159–160, 161, 162– 163, 164, 166–167. See also Cladistics

Phylogenetic bracketing: in diagnosing paleopathology, 341 Phylogenetic definitions: Psittacosauridae, 25 Phylogeny: ceratopsid, 405, 409, 410, 411; in Late Cretaceous North American biogeography, 417 Physical features: of Kikak-Tegoseak Quarry floodplain paleosols, 468 Pierre Shale: in Tatankaceratops sacrisonorum n. gen. & sp. biostratigraphy, 205 Pigeons: eye sizes, 312 Pinaceae: Prince Creek Formation, 471 Pinacosaurus: collected from Bayn Dzak, 320, 322; trace fossils with, 512 Pipestone Creek Pachyrhinosaurus. See Pachyrhinosaurus n. sp. (Grande Prairie) Plant fragments: in Big Bend microsites, 533; in ‘‘Purple Hill’’ section, 525, 527 Plants: in Calgary Public Museum collection, 546; Hell Creek Formation, 554; at High Triceratops locality, 558; Hilda mega-bonebed, 504; in Kikak-Tegoseak Quarry bonebed, 456; Prince Creek Formation, 471. See also Vegetation Plastomeninae: in Rattlesnake Mountain microsites, 525 Plateosaurus: eye size and body mass of, 317 Platyceratops: Diabloceratops n. gen. versus, 132; provenance, 431; skull and mandible, 234 Platyceratops tatarinovi: stratigraphy, paleoenvironmental associations, and taphonomic studies, 441 Play behavior: injuries from, 375 Pleistocene epoch: caudal pathologies in crocodylians from, 368 Pliocene epoch: Prince Creek Formation and, 457 Podocipediforms: eye sizes, 312 Point bar deposits: at Kikak-Tegoseak Quarry, 460–461 Point digitizer, 266 Point Lookout Sandstone: in ceratopsid stratigraphy, 412 Polacanthus: in Mesozoic paleogeography, 398 Polar ecosystems, 456–457 Polarity chrons: ceratopsian cladistics and, 409, 410; in ceratopsid stratigraphy, 412 Pollen: in Kikak-Tegoseak Quarry bonebed, 456, 472–473. See also Palynology Pollex: pathological, 340, 345, 346, 347, 352. See also Thumb

Polycingulatisporites triangularis: as KikakTegoseak Quarry palynomorph, 461, 462 Polypodiaceae: Prince Creek Formation, 473 Polypodiisporites amplus: as Kikak-Tegoseak Quarry palynomorph, 462 Pompilidae: bone modification by, 517 Pond sediments: Kaiparowits Formation, 478 Poole, Karen, xviii, 91 Popcorn bentonite, 554, 557 ‘‘Popcorn’’ microstructure: of Triceratops matrix, 273–275 Population density: ceratopsid, 418 Population isolates: among ceratopsids, 420–421 Populations: of sympatric ceratopsians, 293–294, 300–305; in Triceratops intraspecific interactions, 287–288 Porituberoolithus: in Rattlesnake Mountain microsites, 524 Porous bone surface texture, 252 Porvenir de Jalpa: Cerro del Pueblo Formation at, 100, 101; Coahuilaceratops magnacuerna n. gen. & sp. from, 104; Mexican ceratopsians from, 100 Postcranial elements: abundance of Triceratops, 555; Aguja Formation microsite, 527, 528, 529, 529, 530, 531, 532; in Agujaceratops bonebeds, 520; British Museum Chasmosaurus, 546, 549; Centrosaurus apertus, 502; in ceratopsid phylogeny, 411; CMN 8547, 189–202, 192, 193, 194, 195, 197; Coahuilaceratops magnacuerna n. gen. & sp., 104; Horseshoe Canyon ceratopsid, 198; Kaiparowits ceratopsid B, 485–487, 486, 487– 488; Kaiparowits ceratopsid C, 489–490, 489; in neoceratopsian taphonomy, 433; Ojoceratops fowleri n. gen. & sp., 172, 176, 176; Pachyrhinosaurus n. sp., 144; Pachyrhinosaurus n. sp. (Alaska), 469, 471; pathologies in ceratopsid, 355, 365–374, 378, 379; pathologies in Chasmosaurus irvinensis, 343, 344, 345, 345, 346, 347; relevance to ontogeny of fossil vertebrates, 252; Shenandoah University Triceratops, 273; Tatankaceratops sacrisonorum n. gen. & sp., 211–213; Triceratops, 558; Udanoceratops tschizhovi, 391; undiagnosed Kaiparowits ceratopsid, 490, 491; Zuniceratops christopheri, 91, 95, 96 Postorbital bosses: pachyrhinosaur, 142; Pachyrhinosaurus n. sp., 145, 147 Postorbital crest: psittacosaur, 42

Postorbital horn: psittacosaur, 42 Postorbital-jugal crest/horn: psittacosaur, 42 Postorbital-jugal fossa: psittacosaur, 42 Postorbitals: at Afternoon Delight Triceratops locality, 558, 562; from Cerro del Pueblo Formation, 102, 103; Coahuilaceratops magnacuerna n. gen. & sp., 107– 108; CPC 278, 102, 103; Diabloceratops eatoni n. gen. & sp., 124; Diabloceratops n. gen., 130; at High Triceratops locality, 558; Montanoceratops cerorhynchus, 69, 72, 72, 77; mottled bone surface texture of centrosaurine, 258–259; Pachyrhinosaurus n. sp., 147; pathologies in, 360; psittacosaur, 41–42; Psittacosaurus sp., 26–28, 27; Rubeosaurus ovatus n. gen. & comb., 158, 159; Styracosaurus ovatus, 156; Tatankaceratops sacrisonorum n. gen. & sp., 205, 206, 207; at Trike II Triceratops locality, 559; Zuniceratops christopheri, 93, 94, 94 Posture: of ceratopsians, 10 Potassium/argon dating: Prince Creek Formation, 458 Prearticulars: Archaeoceratops oshimai, 240; Auroraceratops rugosus, 241, 242–243; Chaoyangsaurus youngi, 235; Hongshanosaurus houi, 238; Leptoceratops gracilis, 244, 245; Liaoceratops yanzigouensis, 245–246; measurements of basal ceratopsian, 236; Montanoceratops cerorhynchus, 74; Protoceratops andrewsi, 246; Psittacosaurus, 239, 240 Predation: on ceratopsids, 420 Predators: eye sizes of avian, 312, 313, 314, 315; eyes of, 310, 311, 311, 315–317; frills as defenses against, 269; Protoceratops as, 308, 315–317; in Triceratops intraspecific interactions, 288; Tyrannosaurus rex on Triceratops, 273. See also Carnivores Predentaries: Archaeoceratops oshimai, 240, 241; Archaeoceratops yujingziensis n. sp., 62, 63; Auroraceratops rugosus, 241–242, 241; basal versus advanced ceratopsian, 247; Chaoyangsaurus youngi, 235; Coahuilaceratops magnacuerna n. gen. & sp., 104, 105, 106, 108; Hongshanosaurus houi, 238; Leptoceratops gracilis, 243, 244; Liaoceratops yanzigouensis, 244, 245; measurements of basal ceratopsian, 236; neonate psittacosaur, 51; Ojoceratops fowleri n. gen. & sp., 172, 175–176, 176; Pachyrhinosaurus n. sp., 149, 150; Protoceratops andrewsi, 246; Psittacosaurus,

index

607

Predentaries (continued) 52, 238–239, 239; Tatankaceratops sacrisonorum n. gen. & sp., 210, 211; Yinlong downsi, 237, 237; Zuniceratops christopheri, 93 Prefrontals: Montanoceratops cerorhynchus, 70; Pachyrhinosaurus n. sp., 147; Tatankaceratops sacrisonorum n. gen. & sp., 206, 206, 207, 210 Prehistoric Park: at Calgary Zoo, 545–546 Preiss, Byron: Peter Dodson and, 10 Premaxilla-maxilla ridge: psittacosaur, 41 Premaxillae: accessory antorbital fenestra and, 132, 133; Archaeoceratops yujingziensis n. sp., 60–61, 61, 62; basal ceratopsian, 225; Coahuilaceratops magnacuerna n. gen. & sp., 103, 104–105, 105, 106; Diabloceratops eatoni n. gen. & sp., 117, 122, 123, 123; Montanoceratops cerorhynchus, 68, 69, 70, 71; neonate psittacosaur, 48; Ojoceratops fowleri n. gen. & sp., 171, 174; in Ojoceratops fowleri n. gen. & sp. taxonomy, 177; Pachyrhinosaurus n. sp., 144–145, 145; Protoceratops, 319; psittacosaur, 41; Psittacosaurus, 52; Rubeosaurus ovatus n. gen. & comb., 157, 157; Shenandoah University Triceratops, 273; Tatankaceratops sacrisonorum n. gen. & sp., 207–208; Triceratops, 557; Zuniceratops christopheri, 93 Premaxillary-rostral (PMX) biting. See Bite forces Prenocephale: from Nemegt Formation, 322 Prenoceratops: bonebeds, 452–453; in ceratopsian cladistics, 392, 400; in ceratopsian paleobiogeography, 398; from Montana, 181; in Montanoceratops cladistics, 78, 80, 81; Montanoceratops cerorhynchus versus, 70, 71, 72, 73, 76, 77, 78, 79; Montanoceratops versus, 70; Prenoceratops sp. versus, 85–87; provenance, 387, 390, 391, 431; skull and mandible, 234; taxonomy, 429 Prenoceratops pieganensis: Prenoceratops sp. versus, 88; stratigraphy, paleoenvironmental associations, and taphonomic studies, 442 Prenoceratops sp.: from Alberta, 83–90; specimen referred to, 85–88, 86, 87; systematic paleontology, 85; taxonomy, 88 Preorbital length: psittacosaur, 40, 54 Presa San Antonia, 101; CPC 278 from, 102; Mexican ceratopsians from, 100 Presant, Frederick S.: William Cutler’s friendship with, 541–542

608 index

Primary bone disease: in non-mammalian vertebrates, 341 Primary osteons, 254, 256 Primates: binocular vision among, 321; relative eye sizes of, 311 Prince Creek Formation, 457–458, 458, 459, 460; bonebeds in, 448, 450, 456– 477, 458, 459; ceratopsian paleoenvironmental associations and taphonomy in, 443, 444; in ceratopsid stratigraphy, 412; measured sections of, 463, 464; in North American paleogeography, 415; pachyrhinosaur from, 408; palynology, 458–459, 461, 462, 463, 464; as representing anastomosed fluvial system, 471 Prince Creek pachyrhinosaur, 408; in ceratopsian cladistics, 409; stratigraphy, 412. See also Alaska pachyrhinosaur; Pachyrhinosaurus n. sp. (Alaska) Princeton Gamma Tech Analyzer, 273 Princeton University Press: Peter Dodson and, 10 Prismatic eggshells: in Rattlesnake Mountain microsites, 524 Probers: eye sizes of avian, 312, 314, 315 Procellariforms: eye sizes, 312 Profile line diagrams: in chasmosaurine intraspecific interaction analysis, 284 Progressive degenerative joint disease: in chasmosaurine manus, 340, 342 Prootics: Diabloceratops eatoni n. gen. & sp., 127 Prosaurolophus maximus: eye size and body mass of, 317, 318 Prospecting: of Hell Creek Formation, 553–555; in Late Cretaceous North American biogeography, 417 Protiguanodon: systematic paleontology, 25; taxonomy, 23 Protiguanodon mongoliense. See Protiguanodon mongoliensis Protiguanodon mongoliensis: skeleton, 330, 331, 336; systematic paleontology, 25, 32; taxonomy, 23 Proto-Caribbean Sea: Mesozoic paleogeography of, 397 Protoceratops: abundance of, 315, 316–317; as basal ceratopsian, 222, 222; basioccipital, 223, 225; basisphenoid, 225; behavior, xiii, 308–327; biology of, 323– 324; bone histology, 259; bonebeds, 566; burrowing by, 323–324, 432–433, 434; in ceratopsian cladistics, 392, 400, 409; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; Coahuilaceratops magnacuerna n. gen. & sp.

versus, 108–109; collected specimens of, 308–309; Fox skeleton, 509–511, 510, 511; functions of cranial ornamentation in, 282; insect trace fossils with, 509– 520, 512, 513, 514, 515, 516, 517; large eyes, 308–327, 309, 310, 319; life habits, 428; mandible, 244, 246; manus, 343; Montanoceratops cerorhynchus versus, 68, 70, 71, 72, 73, 74, 75, 76; in Montanoceratops cladistics, 78, 80, 81; nocturnal lifestyle for, 308–327, 432–433; olfaction by, 323; Prenoceratops sp. versus, 87; provenance, 387, 389, 390, 431; psittacosaurs versus, 42; Psittacosaurus versus, 52, 53–54; RFTRA systematics of, 9; sexual dimorphism in, 9; skull and mandible, 234; studied specimens, 223; taphonomy, 432–433, 434; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216; taxonomy, 429; Triceratops versus, 264; Zuniceratops christopheri versus, 94, 95, 96 Protoceratops andrewsi: basioccipital, 225; in ceratopsian paleobiogeography, 398; collected from Bayn Dzak, 320, 322; coronoid/mandible measurements, 248; large eyes, 308–327, 309, 310, 319; mandible, 244, 246; mandibular element measurements for, 236; pes, 324; Peter Dodson’s work on, 5–6, 6; skulls, 6, 309, 319; stratigraphy, paleoenvironmental associations, and taphonomic studies, 442; studied specimens, 223, 235 Protoceratops hellenikorhinus, 5; Diabloceratops n. gen. versus, 132; Protoceratops andrewsi versus, 246; stratigraphy, paleoenvironmental associations, and taphonomic studies, 442 Protoceratopsians/protoceratopsids, 8–9; basicranium and palate, 231; in ceratopsian family tree, 134; in ceratopsian paleobiogeography, 398; Diabloceratops n. gen. versus, 132–134; monophyly of Asian, 133–134; Montanoceratops cerorhynchus among, 68, 69; paleoenvironments including, 322; RFTRA systematics of, 9; semi-aquatic, 329 Provincial Museum of Alberta: ceratopsian paleopathology studies at, 356; ceratopsian research by, 566; William Cutler’s Eoceratops skeleton and, 542. See also Royal Alberta Museum Provincialism: in Late Cretaceous North American biogeography, 417 Prudhoe Bay: Kikak-Tegoseak Quarry and, 459

Pseudoarthrosis: in Centrosaurus fibula, 373; in Pachyrhinosaurus parietals, 363, 375 Pseudoarticulation: in Pachyrhinosaurus parietal, 363 Pseudosutural divisions: in Ojoceratops fowleri n. gen. & sp. taxonomy, 177 Psittacids: eye sizes, 312 Psittacosaur biochron, 22–23, 329 Psittacosauridae. See Psittacosaurids; Psittacosaurs Psittacosaurids: in ceratopsian cladistics, 400; phylogenetic definition, 25. See also Psittacosaurs Psittacosaurs, 21–58; arguments supporting semi-aquatic lifestyle for, 329–330; basicranial and palatal anatomy, 221– 233; basicranial and palatal characters, 231–232; biochron, 22–23, 329; body proportions, 333, 335; bonebeds, 452, 453; in ceratopsian cladistics, 392; ceratopsid/neoceratopsian mandibles versus those of, 247–248; Chaoyangsaurus youngi versus, 235–236; choanae, 231; clades, 54–55, 55; cranial morphology, 40–52; Craspedodon versus, 399; dentition, 27, 30, 31, 33, 38–39, 38–40, 44, 45, 46; discovery, 21–22; distribution of, 389; forelimb movement, 332–333, 334; forelimbs, 332–333, 334, 335; functional anatomy, 329–330, 331; functions of cranial ornamentation among, 282; gastroliths, 32, 328, 328, 333, 432; genera and species, 23, 52–56, 439–440; hindlimbs, 330, 331, 332, 333, 334, 335; juvenile/ neonate, 45–52, 47, 48, 49, 50–51; life habits, 428; localities, 22; mandibular evolution, 234, 247, 248; manus, 329, 332, 333, 334; morphology, 21; nasal functions in, 232; neoceratopsians versus, 221; nomina dubia among, 23, 25– 26, 38–40; palates, 225–231, 226, 229, 230, 231–232; palatines, 228, 229; paleoenvironmental associations of, 429–430; pedes, 331–332, 333; phylogenetic analysis, 221, 222–223, 222; phylogenies, 55; provenance, 387, 389, 390, 391, 430– 431; semi-aquatic behavior for, 328–339; species diversity and turnover among, 421; stratigraphy, paleoenvironmental associations, and taphonomic studies, 439–440; studied specimens, 222–223; Sweden neoceratopsian versus, 399; systematic paleontology, 25–40, 329; systematics, 8–9; taphonomy, 429–430; taphonomy of Yixian, 432, 434;

Tatankaceratops sacrisonorum n. gen. & sp. versus, 211, 214; taxonomy, 23–24, 52– 56; vomers, 230–231 Psittacosaurus: autapomorphies within, 54; with baby skeletons, 330; as basal ceratopsian, 222, 222; basioccipital, 223; basisphenoid, 224; behavior, xiii; bone histology, 259; bonebeds, 452, 566; Chaoyangsaurus youngi versus, 235; coronoid, 214; discovery, 21; eye size, 311; femora, 331, 332; forelimb movement of, 332–333; functions of cranial ornamentation in, 282; gastroliths, 328, 333; Hongshanosaurus houi versus, 238; from Liaoning, 11–12, 22, 23; mandible, 214, 238–240, 239; Montanoceratops cerorhynchus versus, 74; ornithopods versus, 333; osteomyelitis in, 347; palatines, 229; paleobiology, 336–337; palpebrals, 42–43, 44; phylogenetic definition, 25; Prenoceratops sp. versus, 87; pterygoids, 227; radius and ulna of, 332; RFTRA systematics of, 9; sclerotic ring, 43, 44; semiaquatic behavior for, 328–339; skin impressions, 334; skull and mandible, 234; species taxonomy, 12, 23–24; studied specimens, 223; systematic paleontology, 25–40; tail, 333–334; taphonomy, 432; taxonomy, 23, 24, 52–56, 429; teeth grinding in hatchling, 528; Wayne Barlowe painting of, 10 Psittacosaurus fauna, 59 Psittacosaurus gobiensis: stratigraphy, paleoenvironmental associations, and taphonomic studies, 439. See also Psittacosaurus sp. Psittacosaurus guyangensis: stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; systematic paleontology, 32; taxonomy, 23 Psittacosaurus houi n. comb.: systematic paleontology, 25–26. See also Hongshanosaurus houi Psittacosaurus lujiatunensis: adult skull, 40; anatomical studies, 221; as basal ceratopsian, 222; basicranium and palate, 226; basisphenoid, 225; choanae, 231; coronoid/mandible measurements, 248; diagnosis, 28; discovery, 12, 21; lower jaw, 43; mandible, 238–239; palate, 225; palatines, 228–229; provenance, 22, 430; in psittacosaur phylogenies, 55; Psittacosaurus major versus, 28, 29; Psittacosaurus meileyingensis versus, 32; Psittacosaurus xinjiangensis versus, 38; pterygoids, 225, 226, 227; skull, 230;

stapes, 43–44; stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; studied specimens, 222, 223, 235; systematic paleontology, 25–26, 28; taxonomy, 23, 52, 53, 54; vomers, 230–231 Psittacosaurus major, 565; adult skull, 40, 41; as basal ceratopsian, 222; basicranium and palate, 226; basioccipital, 224; basisphenoid, 224–225; braincase, 42; coronoid/mandible measurements, 248; diagnosis, 28–29; discovery, 12, 21; forelimbs versus hindlimbs, 335; lower jaw, 43; mandible, 238–239, 239; occiput, 229; palate, 42; postorbital horn, 42; provenance, 22; in psittacosaur phylogenies, 55; psittacosaur subclade including, 54–55, 55; Psittacosaurus lujiatunensis versus, 28, 29; Psittacosaurus meileyingensis versus, 32; Psittacosaurus xinjiangensis versus, 38; pterygoids, 226, 227; skeleton, 336; skull, 229; stapes, 43–44; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; studied specimens, 222, 223, 235; systematic paleontology, 28–29, 30, 31; taxonomy, 23, 52, 53, 54; vomers, 230 Psittacosaurus mazongshanensis, 66; as basal ceratopsian, 222; discovery, 12; gastroliths, 333; in psittacosaur phylogenies, 55; Psittacosaurus mongoliensis versus, 32; stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; systematic paleontology, 38–39; taxonomy, 23, 52, 53 Psittacosaurus meileyingensis: adult skull, 40; as basal ceratopsian, 222; diagnosis, 32; discovery, 21; lower jaw, 43; mandible, 238–239; palatines, 229; provenance, 22, 23; in psittacosaur phylogenies, 55; quadratojugal protuberance, 42; stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; systematic paleontology, 29–32; tarsus and pes, 330; taxonomy, 23, 52, 53, 54 Psittacosaurus mongoliensis: adult skull, 40– 41, 43; caudal neural spines, 334; ceratohyals, 44; coronoid/mandible measurements, 248; Developmental Mass Extrapolation of, 5; diagnosis, 26, 32; discovery, 21; eye size and body mass of, 317, 318; eyes, 314; forelimb movement of, 332–333; forelimbs versus hindlimbs, 333, 335; gastroliths, 32, 333; hatchling

index

609

Psittacosaurus mongoliensis (continued) skeleton, 47; hatchling skull, 48, 49, 50– 51; juvenile tarsus, 330; lower jaw, 43, 45; mandible, 238–239; in Montanoceratops cladistics, 78, 80, 81; neonate material, 45–46, 47, 48, 49, 50–51; palate, 42; in phylogenetic definition of Psittacosauridae, 25; postorbital horn, 42; provenance, 22, 23; in psittacosaur phylogenies, 55; Psittacosaurus major versus, 28; Psittacosaurus mazongshanensis versus, 39; Psittacosaurus meileyingensis versus, 32; Psittacosaurus neimongoliensis versus, 34; Psittacosaurus ordosensis versus, 39; Psittacosaurus sinensis versus, 37; Psittacosaurus xinjiangensis versus, 38; sclerotic ring, 43, 44; skeletons, 330, 330, 331, 336; skull roof, 43; stapes, 43– 44; stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; studied specimens, 235; systematic paleontology, 25–26, 32, 33; taxonomy, 23, 24, 52, 53, 54, 55 Psittacosaurus neimongoliensis: adult skull, 40, 41; as basal ceratopsian, 222; diagnosis, 32–34; discovery, 12; femur, 334; forelimb/hindlimb movements of, 332– 333, 334; forelimbs versus hindlimbs, 335; mandible, 238–239; provenance, 22; in psittacosaur phylogenies, 55; Psittacosaurus mongoliensis versus, 32; skeleton, 336; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; studied specimens, 235; systematic paleontology, 32–34; taxonomy, 23, 52, 54–55 Psittacosaurus ordosensis: as basal ceratopsian, 222; mandible, 238–239; in psittacosaur phylogenies, 55; Psittacosaurus mongoliensis versus, 32; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; systematic paleontology, 39; taxonomy, 23, 52 Psittacosaurus osborni: stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; systematic paleontology, 32; taxonomy, 23 Psittacosaurus protiguanodonensis: systematic paleontology, 25; taxonomy, 23 Psittacosaurus sattayaraki: stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; systematic paleontology, 39–40; taxonomy, 23 Psittacosaurus sibiricus: adult skull, 41; as basal ceratopsian, 222; ceratohyals, 44; diagnosis, 34; discovery, 12, 21; femur,

610 index

332; lower jaw, 43; mandible, 238–239; postorbital horn, 42; provenance, 22; in psittacosaur phylogenies, 55; Psittacosaurus major versus, 29; Psittacosaurus ordosensis versus, 39; Psittacosaurus xinjiangensis versus, 38; skeleton, 331, 336; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; systematic paleontology, 34; taxonomy, 23, 53, 54 Psittacosaurus sinensis: adult skull, 40, 41; caudal neural spines, 334; ceratohyals, 44; ceratopsian mandibles versus that of, 247; coronoid/mandible measurements, 248; diagnosis, 37; forelimbs versus hindlimbs, 333; jugal horn, 41; mandible, 238–239; in Montanoceratops cladistics, 78, 80, 81; postorbital horn, 42; provenance, 22; in psittacosaur phylogenies, 55; psittacosaur subclade including, 54–55, 55; Psittacosaurus major versus, 28; Psittacosaurus neimongoliensis versus, 34; Psittacosaurus ordosensis versus, 39; Psittacosaurus sp. versus, 26; Psittacosaurus xinjiangensis versus, 38; skeletons, 330, 331, 336; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; studied specimens, 235; systematic paleontology, 34– 38; taxonomy, 23, 52, 53, 54, 55 Psittacosaurus sp.: adult skull, 40, 41; diagnosis, 26; discovery, 22; forelimbs versus hindlimbs, 335; palate, 42; postorbital horn, 42; provenance, 22; in psittacosaur phylogenies, 55; Psittacosaurus mazongshanensis versus, 39; skeleton, 336; systematic paleontology, 26–28; taxonomy, 23, 53, 54.. See also Psittacosaurus gobiensis Psittacosaurus tingi: systematic paleontology, 32; taxonomy, 23 Psittacosaurus xinjiangensis: as basal ceratopsian, 222; diagnosis, 38; discovery, 12, 21; femora, 331, 332; hindlimb, 330; provenance, 22; in psittacosaur phylogenies, 55; Psittacosaurus major versus, 28; skeletal posture, 330; skeletons, 330, 331, 336; stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; systematic paleontology, 38; taxonomy, 23, 52, 53, 54 Psittacosaurus youngi: stratigraphy, paleoenvironmental associations, and taphonomic studies, 440; systematic paleontology, 34; taxonomy, 23 Pterosaurs: from Big Bend National Park,

521; immature versus mature bone texture among, 252; in Rattlesnake Mountain microsites, 524, 525; sclerotic rings among, 312 Pterygoid fossa: psittacosaur, 42 Pterygoids: Archaeoceratops yujingziensis n. sp., 61, 62; basal ceratopsian, 221, 222, 223–224, 225–227; Diabloceratops eatoni n. gen. & sp., 125–126; Diabloceratops n. gen., 129, 130; Pachyrhinosaurus n. sp., 145; psittacosaur, 42 Ptinidae: Protoceratops skeleton insect scavenging versus that by, 510–511 Pubes: CMN 8547, 193, 193, 194, 195; Montanoceratops cerorhynchus, 75. See also Plvic girdles Puffins: as burrow dwellers, 323 Pugnacity: ceratopsian, 355 Pulse patterns: in ceratopsid species diversity and turnover, 421, 422 Pupation chambers: Ceratophaga, 516; dermestid, 516, 517; in Protoceratops skeletons, 509, 510, 513–515, 513, 514, 515, 516–517, 517–518 ‘‘Purple Hill’’ section: in Rattlesnake Mountain microsite stratigraphy, 525– 527, 526, 527 Pyroclastics: in ceratopsian paleoenvironmental associations, 428, 429–430, 430– 431, 439–446 Pythons: nocturnal, 320 Qiaowan, 60 Qinghai Province: Archaeoceratops localities in, 60 Qingshan Formation: ceratopsian paleoenvironmental associations and taphonomy in, 440 Qingshan Group: psittacosaurs from, 37 Qingyang Lake: psittacosaurs from, 330 Quadrate condyle: neonate psittacosaur, 48 Quadrates: at Afternoon Delight Triceratops locality, 558, 562; Archaeoceratops yujingziensis n. sp., 61, 62; basal ceratopsian, 226–227; basal versus advanced ceratopsian, 247–248; Diabloceratops eatoni n. gen. & sp., 125, 126; at High Triceratops locality, 558; modeling, 294, 296; Montanoceratops cerorhynchus, 77; Pachyrhinosaurus n. sp., 147, 148, 148; pathologies in, 360, 361; Tatankaceratops sacrisonorum n. gen. & sp., 209, 210, 214; Zuniceratops christopheri, 93 Quadratojugal protuberance: psittacosaur, 42

Quadratojugals: Diabloceratops eatoni n. gen. & sp., 125; Montanoceratops cerorhynchus, 69, 71–72, 71; Pachyrhinosaurus n. sp., 147, 148, 148; Tatankaceratops sacrisonorum n. gen. & sp., 209 Quaternary period: in North Slope stratigraphy, 460; pupation chambers from, 517 Quercoidites genustriatus: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Rabbits: as burrow dwellers, 323 Radiation: ceratopsid, 405–427, 409, 410 Radii: ceratopsid, 196; in Chasmosaurus locomotion, 350, 350; CMN 8547, 193, 193, 194, 195, 196; osteomyelitis in, 347; psittacosaur, 332; in psittacosaur swimming, 335 Radiometric dating: of ceratopsian-bearing formations, 411, 413, 417; Kaiparowits Formation, 411–413, 479; Prince Creek Formation, 458; Wahweap Formation, 413 Rancho Nuevo Formation: stratigraphy, 101 Ranges. See Geographic ranges; Temporal ranges Rapid replacement: among ceratopsids, 420–421 Ratite eggshells: in Rattlesnake Mountain microsites, 524 Ratites: osteomyelitic infections in, 348 Rattlesnake Mountain, 522, 526; Agujaceratops bonebeds at, 520, 534; Agujaceratops skull from, 524; in Big Bend National Park geology, 521, 523; microsite taphonomy, 532–534 Rattlesnake Mountain microsites, 523, 524–527; Agujaceratops skull and, 524 Rattlesnake Mountain sandstone member: in Aguja Formation geology, 523, 524 Ravens: eye sizes, 312; sclerotic rings, 311 Rays: in Big Bend microsites, 533; from Big Bend National Park, 521, 522, 524 Recent epoch: Kikak-Tegoseak Quarry palynomorphs from, 461 Red Deer badlands: hadrosaur excavated from, 543–544 Red Deer River, 142, 190; CMN 8547 from, 189, 190; Pachyrhinosaurus from, 141, 142 Red-legged Ham beetles: Protoceratops skeleton insect scavenging versus that by, 510–511 Red-tailed hawk: sclerotic rings, 311 Redbeds: table of ceratopsian, 439–446 Redoxymorphic processes: at KikakTegoseak Quarry, 467, 471–472

Reduced Major Axis (RMA) regression: in analysis of eye size versus body mass, 313, 314 Rega, Elizabeth, xviii, 340 Regressions: ceratopsids and, 414, 420 Relative aperture size (RAS): Protoceratops nocturnality and, 320, 321 Remodeled bone, 255, 257, 258, 259– 260 Repcheck, Jack: Peter Dodson and, 10 Reproductive maturity: relevance to ontogeny of fossil vertebrates, 252 Reproductive traits: in ceratopsids, 420 Reptiles: Grand Staircase–Escalante localities and distribution, 119; ontogenetic patterns among, 252; osteomyelitis in, 348; replacement teeth in, 365 Resistant-Fit Theta-Rho Analysis (RFTRA): in ceratopsian systematics, 9–10, 9 Resolution: eye size and, 309–310 Resorbed bone, 255, 258–259 Resorption: of horns and frills, 375 Resting pose: of articulated psittacosaur specimens, 330, 330 Retina: eyesight resolution and, 309, 310; Tyrannosaurus rex, 311; visual sensitivity and, 310 ‘‘Review of the Ceratopsia’’ (Lull), 565 Reworking: in Big Bend microsites, 533, 534; of bonebeds, 449; of Hell Creek Formation fossils, 555; of Hilda megabonebed fossils, 495 Reynolds Point: Diabloceratops eatoni n. gen. & sp. from, 118, 121 Rhinoceros iguana: pushing contests among, 290 Ribs: CMN 8547, 189, 192, 193, 194–195, 194; Diabloceratops eatoni n. gen. & sp., 122; injuries to ceratopsian, 355, 376– 377; Montanoceratops cerorhynchus, 69, 75; Ojoceratops fowleri n. gen. & sp., 172, 176; Pachyrhinosaurus n. sp. (Alaska), 471; pathologies in ceratopsid, 367–368, 369, 376–377; pathology in Chasmosaurus belli, 344, 348; of semi-aquatic ceratopsids, 200; Tatankaceratops sacrisonorum n. gen. & sp., 211, 215 Ricardoestesia: from Big Bend National Park, 521; in Rattlesnake Mountain microsites, 525 Richardson, Scott: Kaiparowits ceratopsid A discovered by, 484 Rincon Colorado: CPC 279 from, 103 Rivers: semi-aquatic animals in, 199 Roadrunner: eye size, 312 Roberts, Eric, xviii, 478

Rocky Mountains: Mesozoic paleogeography of, 397 Rodríguez-de la Rosa, Rubén A., xviii, 99 Rolling compression: of manus in ceratopsid walking, 351–352 ROM 843: pathologies in, 340–354, 347 ROM 1493, 198 Romer, Alfred Sherwood: Peter Dodson and, 3 Root marks: on Kikak-Tegoseak Quarry fossils, 469–470, 471 Root traces: bone modification by, 516; Djadokhta Formation, 511; Hilda megabonebed, 504; at Kikak-Tegoseak Quarry, 466, 468, 469–470, 471, 471 Rostrals: basal ceratopsian, 225; CMN 8547, 190; Coahuilaceratops magnacuerna n. gen. & sp., 103, 104, 105, 106; Hongshanosaurus houi, 238; neonate psittacosaur, 48; Ojoceratops fowleri n. gen. & sp., 175, 175; Pachyrhinosaurus n. sp., 144, 145, 146; pathologies in, 357, 357; Protoceratops, 319; psittacosaur, 41; Psittacosaurus, 238; Shenandoah University Triceratops, 273; Tatankaceratops sacrisonorum n. gen. & sp., 207–208, 208; Zuniceratops christopheri, 93 Rothschild, Bruce M., xviii, 355 Royal Alberta Museum: Hilda megabonebed studies by, 496, 502, 503. See also Provincial Museum of Alberta Royal Tyrrell Museum of Paleontology (RTMP). See Tyrrell Museum of Palaeontology (TMP) Rubeosaurus n. gen.: in centrosaurine cladistics, 164; stratigraphy, 163 Rubeosaurus ovatus n. gen. & comb., 156– 168, 565; in centrosaurine evolution, 163–165; differential and specific diagnoses, 157; Einiosaurus procurvicornis versus, 159, 160, 161, 161, 162–163; life reconstruction, 165; phylogenetic analysis, 156, 159–160, 161, 162–163, 164, 166–167; skull, 406; skull material, 156, 157–159, 162, 165, 165; stratigraphy, 161–162; systematic paleontology, 157 Rugose bone surface texture, 251, 253, 254, 255–258, 255, 259–260; histology, 257, 258; as indicating pathology, 346 Rumsey, Alberta, 190; CMN 8547 from, 189, 190 Russell, Anthony P., xviii, xxi, 181 Russell, Dale A.: on dinosaur paleogeography, 393–394; Peter Dodson and, 4, 7, 8 Russell, Loris S.: William Cutler and, 542, 543, 545

index

611

Russell Basin Triceratops locality, 558, 560 Russia: basal ceratopsians from, 390; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440; ceratopsians from, 429; psittacosaurs from, 21, 22, 34. See also Siberia Ryan, Michael J., xiii, xiv, xviii, xxi, 141, 181 Rywkin, Shanti B., xviii, 282 Sabinas Basin, 101 Sacrals: Archaeoceratops yujingziensis n. sp., 64, 65; CMN 8547, 194, 198; Montanoceratops cerorhynchus, 75; pathologies in ceratopsid, 367; Psittacosaurus major, 29; Zuniceratops christopheri, 95, 96 Sacre du Printemps (Stravinsky), 3 Sacrison, Stan and Steven: Tatankaceratops sacrisonorum n. gen. & sp. discovered by, 205 Sadlier, Rud: on dinosaur growth rates, 5 Sagavanirktok River: Kikak-Tegoseak Quarry and, 459 Sahara Desert: challenges to faunas in, 323 Salamanders: in Late Jurassic paleogeography, 394; psittacosaurs versus, 335; in Rattlesnake Mountain microsites, 524, 534; sprawling pose of, 330 Salame, Issa, xviii, 282 Salmonella: osteomyelitis and, 348 Salt Lake City: Diabloceratops eatoni n. gen. & sp. at, 118; Zuniceratops christopheri and, 92 Saltillo: Cerro del Pueblo Formation at, 100, 101; Coahuilaceratops magnacuerna n. gen. & sp. from, 104; CPC 278 at, 102; CPC 279 at, 103 Sampson, Scott D., xiii, xviii, xxi, 99, 405, 478 San Juan Basin, 170; ceratopsian paleoenvironmental associations and taphonomy in, 445; Ojoceratops fowleri n. gen. & sp. from, 169–180; stratigraphy/biostratigraphy, 170, 177 Sandpipers: eye sizes, 312 Sandstone facies associations: at KikakTegoseak Quarry, 460–461, 461–465 Sandstone lithofacies: at ‘‘B. rex’’ site, 553, 554; for Big Bend microsites, 533; for Hilda mega-bonebed, 496, 498; Kaiparowits Formation, 479, 480, 480, 481, 482, 483; Montana dinosaur skulls and skeletons from, 555, 558; in ‘‘Purple Hill’’ section, 525, 526, 527, 527; Triceratops in, 551, 555, 555 Sandy Point: Hilda mega-bonebed and, 497

612 index

Sankey, Julia T., xviii, 520 Santonian stage: basal ceratopsians from, 390, 429; basal neoceratopsians from, 83, 431; centrosaurine squamosals from, 135; ceratopsian cladistics and, 400; ceratopsian family tree and, 134; ceratopsian paleoenvironmental associations and taphonomy in, 441, 442; ceratopsid cladistic analysis and, 110; KikakTegoseak Quarry palynomorphs from, 461; in Wahweap Formation stratigraphy/dating, 130 Sarcophagous insects: in decomposition, 515 Saskatchewan, 190, 497; bonebeds studied in, 507; Calgary Public Museum collection in, 547; ceratopsian paleoenvironmental associations and taphonomy in, 446; paleopathologies in bonebed specimens from, 376; Torosaurus latus from, 417; Triceratops from, 551; Triceratops horridus from, 417 Saurischians: Grand Staircase–Escalante localities and distribution, 119; sclerotic rings among, 312 Saurolophus: eye size, 311; from Nemegt Formation, 322, 322 Saurolophus osborni: eye size and body mass of, 317, 318 Sauropelta: discovery, 4 Sauropodomorphs: eyes of, 308, 314; gastroliths found with, 333 Sauropods: discovery of North American Jurassic, 555; eye sizes among, 311; in Mesozoic paleogeography, 398; nasal functions in, 232; in Protoceratops fauna, 322; semi-aquatic, 329; supposed aquatic lifestyles of, 328; teeth grinding in hatchling, 528; from Xinminpu Group, Mazongshan, 59 Sauropsids: sclerotic ring function among, 311, 312 Sauropterygians: gastroliths found with, 333 Saurornithoides: from Big Bend National Park, 521; collected from Bayn Dzak, 320, 322 Saurornitholestes: in Big Bend microsites, 533; in Rattlesnake Mountain microsites, 525, 531–532, 532 Scabby Butte: ceratopsid bonebeds at, 495; Pachyrhinosaurus from, 141 Scale models: ceratopsid locomotion and, 341, 349–350, 351–352 Scanning electron microscopy: of Shenandoah University Triceratops horns, 273

Scapulae: Archaeoceratops yujingziensis n. sp., 64–65, 66; CMN 8547, 193, 193, 194, 195; Ojoceratops fowleri n. gen. & sp., 172, 176, 176; osteomyelitis in ceratopsid, 347; pathologies in Pachyrhinosaurus, 372; Tatankaceratops sacrisonorum n. gen. & sp., 211; Zuniceratops christopheri, 95, 96, 96 Scarab beetles: bone modification by, 515, 516 Scarabaeidae: pupation chambers, 513 Scarff, Sonya: baby Triceratops skull found by, 556, 562 Scavengers: in Big Bend microsite formation, 534; eye sizes of avian, 312, 313, 314, 315; eyes of, 308, 311, 311; Tyrannosaurus rex among, 311 Sceloporus: Peter Dodson’s work on, 4 Schrader Bluff Formation: Prince Creek Formation versus, 457, 460 Schweitzer, Mary, 557 Sclerotic rings: in amniotes, 311–312; Centrosaurus apertus, 309; in measuring dinosaur eye size allometry, 313, 314, 315, 316, 317, 318; in measuring/estimating eye size, 311–312, 311, 313, 314, 315; Protoceratops andrewsi, 308, 309; psittacosaur, 43, 44, 329; various dinosaurian, 317 Scollard Formation: basal neoceratopsians from, 83–85; ceratopsian paleoenvironmental associations and taphonomy in, 441, 446; in ceratopsid stratigraphy, 412; in CMN 8547 stratigraphy, 191; geologic correlation, 84; Leptoceratops gracilis from, 68, 243; in North American paleogeography, 415; Triceratops from, 551 Scolosaurus cutleri: collected by William Cutler, 542 Scotese, Christopher R.: on Paleomap Project, 392, 393 Scott, Craig, xxi Sea lions: psittacosaurs versus, 334 Sea turtles: psittacosaurs versus, 334 Seals: psittacosaurs versus, 332, 334 Seasonal climates: bonebed formation in, 452; ceratopsian paleoenvironments in, 428, 429, 431; Late Crteaceous Big Bend, 521; Prince Creek Formation and, 472 Secondary degenerative changes: due to locomotor stresses, 341; in understanding locomotion, 349 Secondary depression: psittacosaur, 41 Secondary osteons, 255, 257, 259–260; in Triceratops horn outer bone layer, 275, 276

Secondary palate: basal ceratopsian, 225 Secondary sexual characteristics: of psittacosaur skulls, 46 Secretaría Educación y Cultura: fieldwork in Mexico under, 100 Sedimentary microlamination: at KikakTegoseak Quarry, 468 Sedimentology: of Agujaceratops bonebeds, 520; ceratopsian paleoenvironments and, 429; of Hilda mega-bonebed, 495, 496–497, 497; of Kaiparowits Formation, 478–479; of Kikak-Tegoseak Quarry bonebed, 456, 459–471, 463, 464, 466; Rattlesnake Mountain microsites, 525– 527, 526, 527 Semi-aquatic behavior: among dinosaurs, 328–329; for basal ceratopsians, 430, 434; for ceratopsians, 566; for ceratopsids, 189, 199, 200; for psittacosaurs, 328–339, 430, 432, 434. See also Aquatic behavior Semi-aquatic taxa: from Kaiparowits Formation, 481 Semi-arid environments. See Arid environments Semi-erect stance: of ceratopsid forelimbs, 341–342, 349–350, 350–351, 350 Sensitivity: eye size and, 310; eye size allometry and, 313–314 Senter, Philip: on ceratopsian forelimb movement, 332–333 Serendipaceratops: ceratopsian dispersal and, 399; provenance, 390, 428, 431; taxonomy, 429; ulna, 393 Serendipaceratops arthurcclarkei: stratigraphy, paleoenvironmental associations, and taphonomic studies, 442 Sereno, Paul C., xiii, xviii, 21: on dinosaur paleobiogeography, 388–389; on psittacosaurids, 8–9; on Psittacosaurus species, 12 Seriema: eye size, 312 7–Up Sandbar individual (Ojoceratops n. gen.), 174, 175, 176 Sevier foreland basin: Kaiparowits Formation and, 479 Sevier thrust belt: ceratopsids and, 414 Sexual dimorphism: ceratopsian systematics and, 9; chasmosaurine horns and frills and, 282, 283, 323; CMN 8547 and, 198; Fox Protoceratops, 513; Pachyrhinosaurus n. sp., 150–151; Peter Dodson’s work on, 5–6, 6; Protoceratops andrewsi, 246; Protoceratops in the study of, 308– 309; Triceratops, 562; in Triceratops intraspecific interactions, 288

Sexual maturity: bone surface texture and centrosaurine, 259; relevance to ontogeny of fossil vertebrates, 252 Sexual selection: among centrosaurines, 165; among ceratopsids, 420–421; in origin and evolution of horns and frills, 290 Shabarakh Usu: Protoceratops andrewsi collection from, 5 Shallow lacustrine deposits: at KikakTegoseak Quarry, 465 Shandong (Shantung): ceratopsian paleoenvironmental associations and taphonomy in, 440; psittacosaurs from, 26; Psittacosaurus sinensis from, 330, 331 Sharks: in Big Bend microsites, 533; from Big Bend National Park, 521, 522, 524 Sheath matrix layer: chemical analysis, 273–275, 274; Shenandoah University Triceratops skull in, 273–275, 275, 278, 280 Sheathbills: eye sizes, 312 Shelducks: eye sizes, 312 Shelter: in arid envirnments, 323 Shenandoah University: Triceratops skull at, 271, 272–273, 272 Shestakovo: psittacosaurs from, 21, 34 Shestakovskaya Svita: ceratopsian paleoenvironmental associations and taphonomy in, 439 Shindler, Fred: Calgary Public Museum collection and, 545 Shireegiin Gashuun: ceratopsian paleoenvironmental associations and taphonomy in, 441 Shishugou Formation: ceratopsian paleoenvironmental associations and taphonomy in, 439 Shoreline deposits: in Big Bend National Park geology, 521; Hilda mega-bonebed and, 507 Shoulder girdles. See Pectoral girdles Shoulders: modeling in locomotion studies, 351 Shoving matches. See Flank-butting behavior; Head butting; Mating competition; Sparring Shrikes: eye sizes, 312 Shrubs: Prince Creek Formation, 471 Shushugou Formation: Yinlong downsi from, 236 Shuvosaurids: sclerotic rings among, 312 Siberia: basal ceratopsians from, 222, 387; in ceratopsian dispersal, 399; ceratopsian paleoenvironmental associations and taphonomy in, 440; psit-

tacosaurs from, 12, 26, 329; Psittacosaurus species from, 238 Sierra College crew: at Sierra Skull Triceratops locality, 561 Sierra Madre Oriental belt: stratigraphy, 100, 101 Sierra Skull Triceratops locality, 558, 561 Sigmopollis psilatus: as Kikak-Tegoseak Quarry palynomorph, 461 Signaling structures: ceratopsid, 406 Signaling traits: in ceratopsids, 420, 421, 422 Silk Road Dinosaur Expedition, 59 Silphidae: Protoceratops skeleton insect scavenging versus that by, 510–511 Siltstone facies associations: at KikakTegoseak Quarry, 465–466 Siltstone lithofacies: of Hilda megabonebed, 500, 500; Kaiparowits Formation, 480, 480, 481, 482, 483; Triceratops in, 551, 555, 558 Siltstone matrix layer: chemical analysis, 273, 274; Shenandoah University Triceratops skull in, 273, 275, 278 Siltstone/sandstone lithofacies: in ‘‘Purple Hill’’ section, 527 Silurian period: Peter Dodson’s fieldwork in, 4 Simpson, George Gaylord: Peter Dodson and, 3 Sino-American Mazongshan Dinosaur Project, 59 Sino-Japanese Silk Road Dinosaur Expedition, 59 Sinornithosaurus: eye size and body mass, 317 Sinornithosaurus millenii: eye size and body mass, 318 Sirenodon: caudal fin, 335 Sirens: psittacosaurs versus, 335 Sisley, Euston: William Cutler and, 543, 544 Size. See Body size; Skeletal element size Size-independent criteria: in ontogeny of fossil vertebrates, 252 Skeletal age classes: recognizing, 252 Skeletal element size: relevance to ontogeny of fossil vertebrates, 252 Skeletal hyperostosis: in ceratopsian vertebrae, 366 Skeletal maturity: determining, 251 Skeletons: Archaeoceratops yujingziensis n. sp., 60, 64–65, 65, 66; Avaceratops lammersi, 8; Big Bend dinosaur, 521; in Big Bend microsites, 533, 534; bone surface texture of postcranial, 259; British Mu-

index

613

Skeletal maturity (continued) seum Chasmosaurus, 546, 549; Calgary Zoo Corythosaurus, 546; Centrosaurus apertus, 413; Chasmosaurus belli, 342– 343; Chasmosaurus irvinensis, 342; CMN 8547, 189–202, 192, 193, 194; Edmontosaurus, 543; Eoceratops, 542, 546–547; Fox Protoceratops, 513–515, 513, 514, 515; hatchling Psittacosaurus mongoliensis, 47; in Hilda mega-bonebed, 495; indeterminate neoceratopsian, 393; insect traces in Protoceratops, 509–520, 512, 513, 514, 515, 516, 517, 517; insectmodified, 518; Kaiparowits ceratopsid A, 484, 485; Kaiparowits ceratopsid B, 485– 487, 486; Kaiparowits ceratopsid C, 489– 490, 489; among KBP dinosaur discoveries, 479; Montanoceratops cerorhynchus, 68, 69–70, 74–76, 76–77; oldest known Tyrannosaurus rex, 553–554, 553; Pachyrhinosaurus n. sp., 142, 144; prospecting Hell Creek Formation for, 554–555, 554, 555; Protiguanodon mongoliensis, 330, 331; Psittacosaurus major, 29; Psittacosaurus mongoliensis, 32, 330, 330, 331; Psittacosaurus neimongoliensis, 32; Psittacosaurus sibiricus, 34, 331; Psittacosaurus sinensis, 36–37, 37–38, 330, 331; Psittacosaurus sp., 26–28; Psittacosaurus species, 336; Psittacosaurus xinjiangensis, 38, 330, 331; Psittacosaurus youngi, 34; relevance to ontogeny of fossil vertebrates, 252; Tatankaceratops sacrisonorum n. gen. & sp., 203; Tyrannosaurus rex, 555; undiagnosed Kaiparowits ceratopsid, 490, 491; Zuniceratops christopheri, 91, 92–93, 95, 96 Skin impressions: Kaiparowits ceratopsid C, 490; psittacosaur, 334 Skinks: nocturnal, 320 Skull cross-sections: in studies of ceratopsian sympatry, 294–296, 296–298, 297, 301 Skull length: psittacosaur, 40–41, 54 Skull roof: neonate psittacosaur, 46–48; pathologies in, 359; psittacosaur, 41–42, 43 Skull strength: modeling and calculating, 294–296, 296–298, 297, 298–305, 298, 299, 301, 304 Skulls: abundance of Triceratops, 555; in Achelousaurus horneri bonebeds, 448; at Afternoon Delight Triceratops locality, 558, 562; Agujaceratops mariscalensis, 520, 524; Albertosaurus, 542; Anchiceratops longirostris, 189; Anchiceratops long-

614 index

irostris cast with CMN 8547, 191–192, 192; Archaeoceratops, 14; Archaeoceratops oshimai, 224, 240; Archaeoceratops yujingziensis n. sp., 60–64, 61; Auroraceratops rugosus, 13, 224; baby Triceratops, 556; basal ceratopsian, 221–233, 224, 225, 226, 227, 228, 229, 230, 231–232, 231, 234–235; Brachyceratops, 156; British Museum Chasmosaurus, 546, 549; centrosaurine, 406; Centrosaurus apertus, 309, 413, 502; ceratopsian, 134, 295, 296, 296–298, 297, 298–305, 298, 299, 301, 304; from ceratopsian bonebeds, 185–187; in ceratopsian systematics, 9– 10, 9; ceratopsid, 406, 406, 407; in ceratopsid phylogeny, 411; chasmosaurine, 407; CMN 8547, 189, 190–191, 191; Coahuilaceratops magnacuerna n. gen. & sp., 104–108, 105, 106, 107, 108, 109; Diabloceratops eatoni n. gen. & sp., 117– 140, 122, 123, 124, 125, 126, 132, 134, 135, 136, 137; Diabloceratops n. gen., 128–129, 130; Edmontonia rugosidens 528; eye sizes and, 311; functions of ornamentation on, 282–283, 289–290; at Getaway Trike Triceratops locality, 558, 561; from Hell Creek Formation, 555, 557; at High Triceratops locality, 558, 560; Hongshanosaurus houi, 224; horns and frills in ceratopsian sparring and, 283–290, 285, 286, 287, 288, 289; Horseshoe Canyon ceratopsid, 198; immature versus mature bone surface texture in ceratopsian, 253; Kaiparowits ceratopsid A, 484, 485; Kaiparowits ceratopsid B, 484–488, 486; Kaiparowits ceratopsid C, 489, 489; Leptoceratops gracilis, 225, 243; Liaoceratops yanzigouensis, 224, 245; Magnirostris dodsoni, 133; Mexican ceratopsian, 99; Montanoceratops cerorhynchus, 68, 69–74, 76–77, 77; in neoceratopsian taphonomy, 433; niche partitioning and shapes of neoceratopsian, 293–307; Ojoceratops fowleri n. gen. & sp., 171–172; in Ojoceratops fowleri n. gen. & sp. taxonomy, 176–177; Pachyrhinosaurus n. sp., 141, 143, 144– 150, 145, 146; Pachyrhinosaurus n. sp. (Alaska), 456, 469, 470; pathological centrosaurine, 356; pathological Pachyrhinosaurus, 356, 358, 378, 379; pathologies in ceratopsid, 355, 357– 365; pathology in Chasmosaurus belli, 344; prospecting Hell Creek Formation for, 554–555, 554, 555; Protoceratops andrewsi, 6, 225, 246, 319; Protoceratops

displaying insect damage, 514; psittacosaur, 40–52, 43, 47, 48, 49, 50–51; in psittacosaur taxonomy, 23–24; Psittacosaurus guyangensis, 32; Psittacosaurus lujiatunensis, 28; Psittacosaurus major, 28–29, 29, 30, 31, 224; Psittacosaurus mazongshanensis, 38–39; Psittacosaurus meileyingensis, 29–32; Psittacosaurus mongoliensis, 32, 33; Psittacosaurus neimongoliensis, 32–34; Psittacosaurus ordosensis, 39; Psittacosaurus sibiricus, 34; Psittacosaurus sinensis, 35, 36, 37; Psittacosaurus sp., 26–28, 27; Psittacosaurus xinjiangensis, 38; Psittacosaurus youngi, 34; Rubeosaurus ovatus n. gen. & comb., 156, 157–159, 162, 165, 165; at Russell Basin Triceratops locality, 558, 560; at Sierra Skull Triceratops locality, 558, 561; Styracosaurus ovatus, 156, 157, 162, 165; supernumerary holes in, 375; among sympatric ceratopsians, 294, 300–305; Tatankaceratops sacrisonorum n. gen. & sp., 203, 204, 204, 205–211; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 214; trace fossils with Pinacosaurus, 512; Triceratops, 264, 265–266, 271, 272–273, 272, 551, 552, 554, 557, 558; Triceratops horridus, 204; at Trike II Triceratops locality, 558; Tyrannosaurus rex, 555; Udanoceratops tschizhovi, 391; undiagnosed Kaiparowits ceratopsid, 491; William Cutler;s Centrosaurus, 547; Zuniceratops christopheri, 91, 92, 93–95, 93, 94, 95, 96. See also Adult skulls Sloan, Donna, 186 Smell: in nocturnal and crepuscular birds, 323. See also Olfaction Smith, David, xix, 91 Smith, Joshua A.: Diabloceratops n. gen. discovered by, 118, 128; Peter Dodson and, 10 Smithsonian Institution, 544 Smithsonian Museum of Natural History. See United States National Museum (USNM) Smooth bone surface texture, 251, 253, 254, 255–258, 255, 259–260; histology, 257, 258; of mature bone, 258 Snails: in Big Bend microsites, 533; in Rattlesnake Mountain microsites, 524, 526 Snakebirds: eye sizes, 312 Snakes: psittacosaurs versus, 335; sclerotic rings among, 311 Snively, Eric, xxi Snouts: psittacosaur, 40; Psittacosaurus, 52;

Tatankaceratops sacrisonorum n. gen. & sp., 207–208, 208 Social behavior: inferred from bonebed studies, 447–448, 451–452, 452–453 Society of Vertebrate Paleontology: Peter Dodson and, 3 Socorro: Zuniceratops christopheri and, 92 Soft tissues: psittacosaur, 334 Soil nodules: in Big Bend microsites, 533 Solid casts: as cornual sinus vascular structure, 276 SolidWorks software, 266 Solitary wasps: bone modification by, 517 Somatic age classes: recognizing, 252 Somatic maturity: in chasmosaurine intraspecific interactions, 285; relevance to ontogeny of fossil vertebrates, 252 Somerset Island: Peter Dodson’s fieldwork on, 4 Sonar: of nocturnal birds, 323 Sonoran Desert: challenges to faunas in, 323 Sorrentino, Gina, 559 South America: Late Cretaceous/early Tertiary, 522; Mesozoic paleogeography of, 395, 396, 397 South Dakota: ceratopsian paleoenvironmental associations and taphonomy in, 445, 446; ceratopsid stratigraphy in, 412; ceratopsids from, 7; chasmosaurine from, 411; pathological Triceratops squamosal from, 361; Tatankaceratops sacrisonorum n. gen. & sp. from, 203– 218; Torosaurus latus from, 417; Triceratops from, 551 South Korea: insect modification of dinosaur bone from, 509 South Saskatchewan River, 496; Hilda mega-bonebed along, 495, 496, 496, 497, 500, 500, 501, 504 Southern biome: Big Bend dinosaurs as in, 521; bonebeds in, 452; ceratopsid species diversity and turnover in, 419– 420, 422; in Late Cretaceous North American biogeography, 416–417, 419; Rattlesnake Mountain microsites versus, 524–525 Sparring: ceratopsid horns and frills and, 304–305; chasmosaurine, 283–290, 285, 286, 287, 288, 289; evidence for ceratopsid, 374–379; in Pachyrhinosaurus, 356; in Triceratops intraspecific interactions, 288–289 Species: table of ceratopsian, 439–446 Species diversity: ceratopsian, 565; ceratopsid, 405, 406, 407–411, 407, 409,

410, 418–419, 428–429, 430; of Prince Creek Formation, 457; psittacosaur, 329 Species taxonomy: psittacosaur, 12, 23–24 Species turnover: among ceratopsians, 413, 414, 419–422, 422–423 Specific diagnoses: Rubeosaurus ovatus n. gen. & comb., 157. See also Diagnoses (taxonomic) Sphagnaceae: Prince Creek Formation, 473 Sphecidae: bone modification by, 517 Spheniscus magellanicus: as burrow dweller, 323 Sphenisicforms: eye sizes, 312 Sphenodon: nocturnal lifestyle of, 320 Spherulitic eggshells: in Rattlesnake Mountain microsites, 524 Spider beetles: Protoceratops skeleton insect scavenging versus that by, 510–511 Spinosaurs: as semi-aquatic, 329 Splay channel deposits: at Kikak-Tegoseak Quarry, 461–465 Splenials: Archaeoceratops oshimai, 241; Archaeoceratops yujingziensis n. sp., 63, 63; Auroraceratops rugosus, 241, 242; Chaoyangsaurus youngi, 236; Diabloceratops eatoni n. gen. & sp., 122; Hongshanosaurus houi, 238; Leptoceratops gracilis, 243, 244; Liaoceratops yanzigouensis, 245; measurements of basal ceratopsian, 236; Montanoceratops cerorhynchus, 69, 73, 73; Pachyrhinosaurus n. sp., 149; Protoceratops andrewsi, 246; Psittacosaurus, 239, 240; Tatankaceratops sacrisonorum n. gen. & sp., 210, 211, 215; Yinlong downsi, 237 Spondyloarthropathy: in ceratopsian vertebrae, 366, 367, 370 Spoonbills: eye sizes, 312 Spores: in Kikak-Tegoseak Quarry bonebed, 456 Sprawling pose: of articulated psittacosaur specimens, 330, 331; of ceratopsid forelimbs, 341–342, 349–350, 350–351 Squamates: nocturnal lifestyles among, 317, 320 Squamosals: in Achelousaurus horneri bonebeds, 448; at Afternoon Delight Triceratops locality, 558; bone surface texture, 255; from Cerro del Pueblo Formation, 102, 103, 103; CMN 8547, 190–191, 191, 196; Coahuilaceratops magnacuerna n. gen. & sp., 108, 109; CPC 279, 103, 103; Diabloceratops eatoni n. gen. & sp., 117, 120, 122, 123, 125, 127–128; Diabloceratops n. gen., 128–129, 129; in Diabloceratops n. gen. ontogeny, 135, 135; in Diabloceratops taxonomy, 135,

135, 136; horns and frills in ceratopsian sparring and, 283–290; Mexican ceratopsian, 99; Montanoceratops cerorhynchus, 69, 72–73, 72, 77, 77; morphology of chasmosaurine, 282; mottled bone surface texture of centrosaurine, 258– 259; Ojoceratops fowleri n. gen. & sp., 169, 171–172, 171, 172, 172–173; in Ojoceratops fowleri n. gen. & sp. taxonomy, 176–177, 177–178; ornamentation on, 282; Pachyrhinosaurus n. sp., 144, 145, 146, 147, 148–149; pathological Centrosaurus, 356; pathologies in, 359, 360–362; pathology in Chasmosaurus belli, 344; pathology in Chasmosaurus irvinensis, 344; Shenandoah University Triceratops, 273; Tatankaceratops sacrisonorum n. gen. & sp., 208, 209; Torosaurus utahensis, 171; Triceratops, 551, 554; in Triceratops frill finite element modeling, 264, 267, 268, 269; Zuniceratops christopheri, 91, 96 St. George’s Island: Calgary Public Museum collection and, 544, 545 St. Mary River Formation: basal neoceratopsians from, 83; bonebeds in, 448; ceratopsian paleoenvironmental associations and taphonomy in, 441, 443; in ceratopsid stratigraphy, 412; geologic correlation, 84; Montanoceratops cerorhynchus from, 68, 70, 76; Pachyrhinosaurus canadensis from, 141 Stance: in ceratopsid locomotion, 341– 342, 349–350, 350–351, 350; in tetrapod locomotion, 341 Standardless assay, 273 Stapes: neonate psittacosaur, 51; psittacosaur, 43–44, 45 Staphylococcus aureus: osteomyelitis and, 348 State Museum of Pennsylvania: New Mexico dinosaur faunas surveyed by, 170; Ojoceratops fowleri n. gen. & sp. material at, 171 Statistics: in chasmosaurine intraspecific interaction analysis. 284, 289 Steep Draw Sandstone: Zuniceratops christopheri and, 92 Stegoceras: in ceratopsian cladistics, 392, 400; in Montanoceratops cladistics, 78, 80, 81; RFTRA systematics of, 9 Steinbichler Optotechnik Optotrak scanner, 266 Steinkerns: in Big Bend microsites, 533 Stenopelix valdensis: in Mesozoic paleogeography, 398

index

615

Step cycle: ceratopsid, 342, 351–352; in tetrapod locomotion, 341 Step-wise directional evolution: among ceratopsids, 421 Stereisporites antiquasporites: as KikakTegoseak Quarry palynomorph, 462 Stereisporites regium: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Sternal plates: CMN 8547, 195 Sternberg, Charles H.: Corythosaurus collected by, 546; Triceratops collected by, 551; William Cutler’s Eoceratops and, 547 Sternberg, Charles M., 565; Calgary Public Museum collection and, 544, 545, 547; ceratopsian research by, 566; CMN 8547 discovered by, 189, 190, 197–198; Pachyrhinosaurus canadensis described by, 141; Peter Dodson and, 3–4 Sternberg family: William Cutler and, 542 Sternberg, George F.: Styracosaurus ovatus discovered by, 156 Sternberg, Raymond N.: Calgary Public Museum collection and, 545 Sternines: eye sizes, 312 Stevens, Kent, xxii Steveville, Alberta: Calgary Public Museum collection and, 546 Stone curlew: eye size, 312; nocturnal lifestyle of, 320 Straight Cliffs Formation: ceratopsian family tree and, 134; in ceratopsid stratigraphy, 412; Diabloceratops n. gen. & sp. and, 122, 128; stratigraphy/dating, 129, 130, 131 Strain energy density: in Triceratops frill, 264, 266, 267, 268 Strains: in finite element modeling, 265, 267 Stratigraphic distribution: ceratopsian, 429, 430; ceratopsid, 405–427, 409, 410, 428 Stratigraphy: Agujaceratops bonebed, 520; Big Bend National Park, 523; ceratopsian, 430, 566; ceratopsian family tree and, 134, 400, 409, 410; of ceratopsid bonebeds, 375, 448, 449–450; in ceratopsid cladistic analysis, 110; ceratopsid evolution and, 411–414; ceratopsid species diversity and turnover and, 420, 421; CMN 8547, 191, 196–197; for Diabloceratops eatoni n. gen. & sp., 118, 119, 120, 121, 122, 129–131; for Diabloceratops n. gen., 128; Hell Creek Formation, 552, 553; Hilda mega-bonebed, 495, 496–497, 497, 498, 499, 501; Kaiparowits Formation, 478–479, 479–482,

616 index

480; Kikak-Tegoseak Quarry, 457–458, 458, 460, 463, 464; Late Cretaceous New Mexico, 170–171, 177; in Late Cretaceous North American biogeography, 417; Medusaceratops lokii n. gen. & sp., 182, 187; Prince Creek Formation, 457– 458, 458, 460, 463, 464; Rattlesnake Mountain microsites, 525–527, 526, 527; Rubeosaurus ovatus n. gen. & comb. 161–162; Shenandoah University Triceratops skull, 272; of sympatric ceratopsians, 293, 300–303, 303; table of ceratopsian, 439–446 Stream channel deposits: Rattlesnake Mountain microsites in, 524 Streptococcus: osteomyelitis and, 348 Stress deformation: of Triceratops frill, 267, 268 Stress fractures: in ceratopsian phalanx, 356; in ceratopsid phalanges, 373–374, 374; diagnosing in fossils, 349, 352 Stresses: in finite element modeling, 265, 267; locomotor, 341; in Triceratops frill, 264, 267, 268 Striated bone surface texture, 252 Strigiforms: eye sizes among, 311, 312, 320, 321; studies of sympatric extant, 293 Strigops habroptilus: nocturnal lifestyle of, 320, 323 Strix: sclerotic rings, 311 Structural properties: modeling Triceratops frill, 264–270; of Triceratops horns, 271– 281 Struthiomimus: Aguja Formation ornithomimids versus, 531 Sturgeons: psittacosaurs versus, 335 Styracosaurus: from Alberta and Montana, 152; behavior, 566; bonebeds, 356, 377, 452, 505; caudal pathologies in, 370, 371; in centrosaurine cladistics, 164; in centrosaurine evolution, 163; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; in ceratopsid species diversity and turnover, 422; Diabloceratops eatoni n. gen. & sp. versus, 128, 135; from Dinosaur Park Formation, 156; dorsal pathologies in, 366, 367; frill, 132, 132; immature versus mature bone surface texture in, 252; Medusaceratops lokii n. gen. & sp. versus, 185; osteopathies of, 355; in Pachyrhinosaurus n. sp. phylogeny, 151, 152, 153–154; paleopathologies in bonebed specimens, 376; parietal pathologies in, 362; pathological dorsals of, 356; Peter Dodson’s systematics of, 9;

phalangeal stress fractures in, 373–374; phylogeny, 405, 411; Prenoceratops sp. versus, 85; provenance, 431; RFTRA systematics of, 9; rib pathologies in, 367; Rubeosaurus ovatus n. gen. & comb. versus, 157, 158, 160, 162, 163; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216; taxonomy, 408; Wayne Barlowe painting of, 10 Styracosaurus albertensis: bonebeds, 448, 449, 450, 566; in ceratopsian biostratigraphy, 152; in ceratopsian cladistics, 409; ceratopsian paleoenvironmental associations and taphonomy in, 444; CMN 8547 versus, 193, 196; frill, 164; Medusaceratops lokii n. gen. & sp. verssu, 182; occurrence of, 413, 414, 416; Pachyrhinosaurus versus, 142; Peter Dodson’s systematics of, 9; Peter Dodson’s work on, 8; Rubeosaurus ovatus n. gen. & comb. versus, 156, 159, 161, 161, 165, 166; skull, 406; skull and cladogram, 295, 296; skull strengths and measurements, 299, 300, 301, 301, 302, 303, 304, 305; stratigraphy, 303, 412; taxonomy, 408 Styracosaurus ovatus, 156, 165; bonebeds, 449; from Montana, 181; occurrence of, 413, 416; Pachyrhinosaurus versus, 142; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; taxonomy, 408; from Two Medicine Formation, 156. See also Rubeosaurus ovatus n. gen. & comb. Suanjingzi Basin: dinosaurs from, 66 Subadults: in bonebeds, 456; caudal pathologies in, 369; distinguishing via bone surface texture, 251, 252, 253, 259; at Getaway Trike Triceratops locality, 558, 561; Hell Creek Triceratops, 556, 557; at High Triceratops locality, 558, 560; Kaiparowits ceratopsid A, 484, 485; Kaiparowits ceratopsid B, 488; Pachyrhinosaurus n. sp. (Alaska), 470–471; pathological Monoclonius lowei, 356, 364; pathological Pachyrhinosaurus parietals of, 375; pelvic pathologies in, 372; at Russell Basin Triceratops locality, 558; Triceratops, 551, 558; in Triceratops intraspecific interactions, 288; at Trike II Triceratops locality, 558; undiagnosed Kaiparowits ceratopsid, 490, 491 Subaqueous burial: of Kaiparowits ceratopsid B bones, 487 Subchondral fractures: diagnosing in fossils, 348

Suhongtu: psittacosaurs from, 26 Sullivan, Robert M., xix, 169 Supracranial cavity: Pachyrhinosaurus n. sp., 147 Supraorbital horns/horncores. See Horns; Postorbitals Surangulars: Archaeoceratops oshimai, 240, 241; Archaeoceratops yujingziensis n. sp., 62–63, 63; Auroraceratops rugosus, 241, 242; basal versus advanced ceratopsian, 247–248; Chaoyangsaurus youngi, 236; Hongshanosaurus houi, 238; Leptoceratops gracilis, 243, 244; Liaoceratops yanzigouensis, 244, 245; measurements of basal ceratopsian, 236; Montanoceratops cerorhynchus, 69, 73–74; Pachyrhinosaurus n. sp., 150, 151; Protoceratops andrewsi, 246; Psittacosaurus, 239–240, 239; Tatankaceratops sacrisonorum n. gen. & sp., 210, 211, 214; Yinlong downsi, 237, 237 Surfacer software, 266 Suzhousaurus megatherioide: from Gongpoquan Basin, 59 Swallows: eye sizes, 312 Swallowtail gull: nocturnal lifestyle of, 320 Sweden neoceratopsian, 390, 391, 392, 393, 399; in ceratopsian cladistics, 400 Swiftlets: nocturnal lifestyle of, 323 Swifts: eye sizes, 312 Swimming: among psittacosaurs, 328, 332, 334, 335 Symmetrodonts: Grand Staircase– Escalante localities and distribution, 119 Sympatry: neoceratopsian skull shapes, niche partitioning, and, 293–307 Synapomorphies: centrosaurine, 160; long postorbital horns as ceratopsid, 283; Psittacosaurus, 52, 53; Rubeosaurus n. gen. and Einiosaurus, 160, 162–163 Synapsids: immature versus mature bone texture among, 252; sclerotic rings among, 312 Syncerus caffer: chasmosaurines versus, 283 Syncervicals: CMN 8547, 192, 193, 193, 194 Syndesmophytes, 348 Synonymies: psittacosaur, 23–24, 25–26, 32 Synovitis: diagnosing in fossils, 346, 347– 348 Synsacrum: CMN 8547, 194, 198 Systematic paleontology: Archaeoceratops yujingziensis n. sp., 60; CMN 8547, 190; Coahuiaceratops magnacuerva n. gen. & sp., 104–108; Diabloceratops eatoni n.

gen. & sp., 120–122; of indeterminate Mexican ceratopsids, 102–104; Medusaceratops lokii n. gen. & sp., 182–183; Montanoceratops cerorhynchus, 69–77; Ojoceratops fowleri n. gen. & sp., 171– 172; Prenoceratops sp., 85; psittacosaur, 25–40, 329; Rattlesnake Mountain microsites, 527–532, 529, 530, 532; Rubeosaurus ovatus n. gen. & comb., 157; Tatankaceratops sacrisonorum n. gen. & sp., 204–205 Systematics: ceratopsian, xiii, 566; ceratopsid, 9–10, 9; psittacosaur, 8–9; Triceratops, 551, 552, 562. See also Cladistics; Phylogenetic analyses; Phylogeny; Systematic paleontology; Taxonomy Tabular sandstone and siltstone facies association: at Kikak-Tegoseak Quarry, 465 Tabular sandstone lithofacies: Kaiparowits Formation, 480, 480, 481, 482 Tachengzi Formation: ceratopsian paleoenvironmental associations and taphonomy in, 439 Tadorna: eye size, 312 Tail amputation: ceratopsid, 369 Tail standing: by Protoceratops skeletons, 511 Tails: pathologies in ceratopsid, 368–372, 370, 371; Protoceratops displaying insect damage, 514; psittacosaur, 333–334, 337; in psittacosaur swimming, 335; Psittacosaurus, 328. See also Caudal entries Talisman Energy Fossil Gallery: CMN 8547 displayed in, 191–192, 192, 194 Talley Mountain bonebeds: dinosaurs from, 521 Talley Mountain Field Area, 522 Talley Mountain microsites: in Aguja Formation geology, 523, 524; theropod teeth from, 531 Tanke, Darren H., xiv, xix, 141, 355, 541; historical research by, 565–566 Tanoue, Kyo, xix, 14, 59, 221, 234; Peter Dodson and, 7, 7, 11 Tanzania: William Cutler’s death in, 542 Taphocoenoses: in bonebed studies, 447 Taphonomic analysis: Kaiparowits Formation, 482 Taphonomic biases: in bonebed studies, 447 Taphonomic modes: Big Bend versus Dinosaur Park, 533–534; Kaiparowits Formation, 481, 482–484, 491–492 Taphonomic patterns: ceratopsian, 431– 433, 433–434

Taphonomic studies: table of ceratopsian, 439–446 Taphonomy: Avaceratops lammersi, 7–8; centrosaurine bonebed, 495–496; centrosaurine mega-bonebed, 495–508; ceratopsian, xiii, 428–446, 566; ceratopsian bonebed, 448–451; in ceratopsian skull and jaw analysis, 294, 296; ceratopsid bonebed, 375, 376; of Hilda megabonebed, 502–503, 503, 503–506, 504; Kaiparowits ceratopsids, 478–494, 481; at Kikak-Tegoseak Quarry, 467–471; of Mongolian Protoceratops skeletons, 511; pathology versus, 357; Peter Dodson on, 7; Protoceratops, 317, 324; of Protoceratops fauna, 322–323; of psittacosaur skeletons, 330–331, 330, 331; Rattlesnake Mountain microsites, 532–534; Shenandoah University Triceratops skull, 272– 273, 272, 278–279, 280; Triceratops fossil record and, 552 Tarbosaurus: eye size, 313; from Nemegt Formation, 322, 322 Tarbosaurus bataar: eye size and body mass of, 317 Tarchia: from Nemegt Formation, 322, 322 Tarsals: Montanoceratops cerorhynchus, 76 Tarsi: psittacosaur, 331; Psittacosaurus meileyingensis, 330; Psittacosaurus mongoliensis, 330 Tatankaceratops sacrisonorum n. gen. & sp., 203–218, 565; biostratigraphy, 205; cladistic analysis, 213, 216–217, 216; description, 203, 205–213; diagnosis, 204; discovery, 203, 205; locality, 204– 205; size, 203; systematic paleontology, 204–205; taxonomy, 213–215, 216–217, 216 Taung child, 8 Taxa: paleobiogeography of ceratopsian, 388–389 Taxodiaceae: Prince Creek Formation, 471, 473 Taxodiaceaepollenites hiatus: as KikakTegoseak Quarry palynomorph, 462 Taxonomic diversity: of Prince Creek Formation, 457 Taxonomy: Archaeoceratops yujingziensis n. sp., 65–66, 66–67; ceratopsian, 429, 565, 566; ceratopsid, 405, 406, 407–411, 429; within ceratopsid bonebeds, 448, 449; ceratopsid fossil record and, 418–419; Chasmosaurus mariscalensis, 100; CMN 8547, 196–198; Coahuilaceratops magnacuerna n. gen. & sp., 99, 109–111; Diabloceratops eatoni n. gen. & sp., 135–136;

index

617

Taxonomy (continued) Eoceratops versus Chasmosaurus, 542; of Hilda mega-bonebed fossils, 502–503; in Late Cretaceous North American biogeography, 417; of Mansfield Bonebed ceratopsians, 185–186; Medusaceratops lokii n. gen. & sp., 185–187; Montanoceratops cerorhynchus, 68–69, 77–79; neoceratopsian, 418, 429; Ojoceratops fowleri n. gen. & sp., 176–177, 177–178; Pachyrhinosaurus n. sp., 150–153, 152, 153– 154; Pachyrhinosaurus n. sp. (Alaska), 470–471; Prenoceratops sp., 88; psittacosaur, 12, 23–24, 25–26, 28, 52–56; Tatankaceratops sacrisonorum n. gen. & sp., 213–215, 216–217, 216; Torosaurus, 417; Zunicerartops, 418; Zuniceratops christopheri, 91 Taylor, C. R.: on thermoregulatory function of goat horns, 280 Taylor, Don: Hilda mega-bonebed studies by, 496 Tectonic Highlands: Zuniceratops christopheri and, 92 Teeth: Aguja ankylosaur, 527–528, 530; Aguja ceratopsian, 530; Aguja hadrosaurid, 528–530, 530; Aguja theropod, 530–532, 530, 532; in Big Bend microsites, 533, 534; from Big Bend National Park, 521, 522, 524, 525, 527–528, 528– 530, 530–531, 530, 532, 532; in Hilda mega-bonebed, 505. See also Dentition; Tooth entries Temperate climates: ceratopsian paleoenvironments in, 431 Temperature: in arid environments, 323 Temporal ranges: basal ceratopsian, 430; basal neoceratopsian, 430–431, 430; centrosaurine, 430; ceratopsid, 400, 405, 406, 409, 410, 413, 416–418, 430; chasmosaurine, 430 Tendons: pathologies of, 368, 372 Tenebrionid beetles: bone modification by, 516–517 Tenebruionidae: pupation chambers, 513 Tenontosaurus: discovery, 4 Tensile strength: in ceratopsian skulls, 294–296, 296–298, 297, 298–305, 298, 299, 301, 304 Terlingua Creek sandstone member: in Aguja Formation geology, 523 Terlingua Field Area, 522 Terlingua local fauna: in Aguja Formation geology, 523; ankylosaurs in, 528 Termites: bone modification by, 515, 516– 517

618 index

Terns: eye sizes, 312 Terrestrial deposits: in Big Bend National Park geology, 521 Terrestrial habitats: as ceratopsid paleoenvironments, 414–416, 420, 430, 434 Terrestrial paleofaunas: from Kaiparowits Plateau, 118, 119 Terrestrial vertebrates: in Big Bend microsites, 533; in-life resting postures of fossil, 330 Terrestriality: arboreality versus, 293; among dinosaurs, 328; of Psittacosaurus, 329; psittacosaur, 332 Territorial competition: ceratopsian horns in, 271 Tertiary period: in Big Bend National Park geology, 522, 523; in Mexican stratigraphy, 101; in North Slope stratigraphy, 457, 460 Teshekpuk Lake: Kikak-Tegoseak Quarry and, 459 Tethys Ocean: Mesozoic paleogeography of, 397 Tetrapods: ceratopsids versus, 342; feet and locomotion of, 341 Texas: Agujaceratops bonebeds in, 520–537; Big Bend National Park in, 522; bonebeds in, 448, 450, 451; ceratopsian paleoenvironmental associations and taphonomy in, 444, 446; ceratopsid distribution in, 99; ceratopsid stratigraphy in, 412; Ojoceratops fowleri n. gen. & sp. and, 178; paleoenvironments of, 416, 417; Torosaurus latus from, 417 Texas Memorial Museum: Agujaceratops skull at, 524 Textural features: of Kikak-Tegoseak Quarry floodplain paleosols, 468–469 Thailand: ceratopsian paleoenvironmental associations and taphonomy in, 440; ceratopsians from, 429; psittacosaurs from, 26, 329; Psittacosaurus species from, 238 Thanatocoenoses: of neoceratopsian bonebeds, 433 The Horned DinosaursCA Natural History (Dodson), 565; publication of, 10 Therizinosaurians/therizinosaurids/therizinosaurs: from Gongpoquan Basin, 59; in Mesozoic paleogeography, 398, 401; Nemegt, 322; Zuniceratops christopheri and, 91, 92 Therizinosaurus: from Nemegt Formation, 322, 322 Thermoregulation: ceratopsian horns and, 271–272, 280; Protoceratops, 323; Tri-

ceratops frill in, 323; Triceratops horns and, 271 Theropod eggshells: in Rattlesnake Mountain microsites, 524 Theropods: in Big Bend microsites, 533; from Big Bend National Park, 521, 524, 525; dentition, 316; eyes, 308, 314; finite element modeling of, 265; in highlatitude ecosystems, 457; in Hilda megabonebed, 505; insect borings in bones of, 512; KBP discoveries of, 479, 481; at Kikak-Tegoseak Quarry, 467, 469, 473; Late Jurassic and Late Cretaceous sympatric, 293; long-grained bone surface texture in, 258; in Mesozoic paleogeography, 398; orbit shapes, 311; osteomyelitis in, 347; Protoceratops versus, 319; rib damage due to scavenging, 367; from Xinminpu Group, Mazongshan, 59 Thierrien, François, xxii Thin lenticular sandstone facies association: at Kikak-Tegoseak Quarry, 461–465 Thin sectioning: of ceratopsian phalanges, 343 Thin sections of bone: preparation, 253– 254 Third Geology and Mineral Resources Exploration Academy, 59 Thoracic vertebrae. See Dorsals Three-dimensional digital scans: ceratopsid locomotion and, 341, 349, 350– 351, 351–352 Three-dimensional models: of ceratopsian skulls, 294–296, 296–298, 297, 301; of Triceratops frill, 264–270 Threehills Creek: William Cutler at, 541– 542, 543 Threskiornithids: eye sizes, 312 Thumb: in Chasmosaurus locomotion, 350–351, 350, 351–352; locomotion and paleopathology of, 349–351; pathology in Chasmosaurus belli, 344– 346, 347, 349; pathology in Chasmosaurus irvinensis, 343–344, 345, 345. See also Pollex Thurston, James E.: Dinosaur Park fieldwork by, 543 Thurston, James Robert: William Cutler and, 542–543 Tibiae: CMN 8547, 193, 193, 194, 196; Montanoceratops cerorhynchus, 76; osteomyelitis in, 347; pathological hypsilophodontid and hadrosaur, 347; Zuniceratops christopheri, 95, 96 Time bias: in dinosaur paleobiogeography, 389

Tineid moths; bone modification by, 515– 516 Tirabasso, Alex, xix, 340 Titanosaurids: in Mesozoic paleogeography, 398 Titanosauriforms: from Gongpoquan Basin, 59 Titanosaurs: Nemegt, 322; occurrence of, 416 Tithonian stage: ceratopsian cladistics and, 400 Titus, Alan L., xix, 478 TMP 2002.76.1, 141–155 TMP 82.11.1: described, 76–77, 77 TMP 87.89.8, 83–90; locality map, 85 Tokaryk, Tim: juvenile chasmosaurine horns found by, 556 Tooth decay: absence in ceratopsids, 365 Tooth grinding: in baby dinosaurs, 528 Tornillo Basin: Big Bend National Park in, 521 Tornillo Flat, 522 Torosaurus: bonebeds, 566; in ceratopsian frill stress analysis, 268; in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 284, 286, 288; CMN 8547 versus, 198; Coahuilaceratops magnacuerna n. gen. & sp. versus, 104, 107, 109, 111; distribution, 100; frill function, 283; horns, 356; juvenile horns, 556; Medusaceratops lokii n. gen. & sp. versus, 185; occurrence of, 414–415, 416, 417; Ojoceratops fowleri n. gen. & sp. versus, 177, 178; phylogeny, 405; provenance, 431; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 216, 216; Tatankaceratops sacrisonorum n. gen. & sp. versus, 205, 207; taxonomy, 408, 417 Torosaurus latus: biogeography, 111; bonebeds, 449; in ceratopsian cladistics, 410; ceratopsian paleoenvironmental associations and taphonomy in, 446; in ceratopsid fossil record, 419; Coahuilaceratops magnacuerna n. gen. & sp. versus, 109; from Hell Creek Formation, 203; occurrence of, 413, 417; Ojoceratops fowleri n. gen. & sp. versus, 169, 177, 178; skull, 407; skull and cladogram, 295, 296; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412 Torosaurus utahensis: biogeography, 111; bonebeds, 448, 449, 450, 451; Coahuilaceratops magnacuerna n. gen. & sp. ver-

sus, 109; occurrence of, 417; Ojoceratops fowleri n. gen. & sp. versus, 169, 173, 177–178; squamosal, 171; stratigraphy, 412 Torreon, 101 Torsional forces: in ceratopsian skulls, 294–296, 296–298, 297, 298–305, 298, 299, 301, 304 Tortoises: from Djadokhta Formation, 323 Trace fossils: Djadokhta Formation, 511– 513, 512, 513. See also Bite marks; Burrows; Footprints; Insect entries; Root entries; Tracksites; Trackways; Trampling Trachodon cantabrigiensis: in Mesozoic paleogeography, 398 Tracksites: in Big Bend microsites, 533; North American ankylosaur, 528 Trackways: ceratopsian, 10; ceratopsid locomotion and, 341, 349–350; Zuniceratops christopheri and, 92. See also Footprints Trampling: in Big Bend microsite formation, 534; of Hilda mega-bonebed fossils, 495; of Kaiparowits ceratopsid B bones, 487; in neoceratopsian taphonomy, 433; rib/vertebral damage due to, 367, 368. See also Bioturbation Trans Canada Highway, 190 Transgressions: ceratopsids and, 414, 420, 421–422. See also Western Inland Sea Transgressive-regressive cycles: ceratopsids and, 414, 420, 422 Transverse palatine wing: basal ceratopsian, 228 Trauma: osteomyelitis and, 347–348 Tree Reconciliation Analysis (TRA): in dinosaur paleobiogeography, 389 Tree snakes: nocturnal lifestyle of, 320 Triceratops, 6; bonebeds, 433, 448, 449, 450–451, 450, 452, 566; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; in chasmosaurine intraspecific interaction analysis, 284, 285, 287–289, 289; in Chasmosaurus locomotion studies, 349; CMN 8547 versus, 198; Coahuilaceratops magnacuerna n. gen. & sp. versus, 104, 106, 107, 109, 111; Diabloceratops eatoni n. gen. & sp. versus, 125, 131; Diabloceratops n. gen. versus, 129; distinctiveness of, 293–294; field misidentification of, 558; fossil record, 555–556; frill, 265, 268–269; frill function, 282–283; frill thermoregulation, 323; frills as sexual display structures in, 6; goat horns

versus, 280; growth series in, 259; Hell Creek Project and, 552; historical bias in collecting from Montana, 551–563; horn structure and function, 271–281; horns, 356; immature versus mature bone surface texture in, 253; marginal frill ossifications, 102; Medusaceratops lokii n. gen. & sp. versus, 185; Montana localities, 552, 553, 554, 555, 556, 557, 558, 559, 560, 561, 562; in Montanoceratops cladistics, 78, 80, 81; nasal pathologies in, 357–359; New Mexico frill material versus, 169; occurrence of, 414–415; Ojoceratops fowleri n. gen. & sp. versus, 177, 178; paleoenvironments of, 415, 417; pathology in, 347; Peter Dodson and, 10; phylogeny, 405; posture, 10; prospecting Hell Creek Formation for, 554–555, 554, 555; provenance, 431; RFTRA systematics of, 9, 10; skull, 555, 557; species diversity of, 417; species taxonomy, 551–552, 558–562; squamosal pathologies in, 361, 362; structural frill models for, 264–270; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 213–214, 215, 216, 216; Tatankaceratops sacrisonorum n. gen. & sp. versus, 205, 206, 207, 208, 209, 211; taxonomy, 408; Tyrannosaurus versus, 555; Wayne Barlowe painting of, 10; Zuniceratops christopheri versus, 94, 96 Triceratops flabellatus: fossilized horn sheaths found with, 271 Triceratops horridus: in ceratopsian cladistics, 410; in ceratopsian paleobiogeography, 398; ceratopsian paleoenvironmental associations and taphonomy in, 446; in ceratopsid fossil record, 419; ceratopsid species diversity and turnover and, 419–420; CMN 8547 versus, 194, 195, 195, 196; Coahuilaceratops magnacuerna n. gen. & sp. versus, 109; frill, 265–266, 265, 266, 267, 268–269, 268; from Hell Creek Formation, 203; humerus, 195; occurrence of, 413, 414, 417; Ojoceratops fowleri n. gen. & sp. versus, 169, 178; in phylogenetic definition of Psittacosauridae, 25; range/ duration of, 413, 414; skull, 407; skull and cladogram, 295, 296, 297; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412; Tatankaceratops sacrisonorum n. gen. & sp. versus, 205, 207, 208; taxonomy, 552; ulna, 197; as youngest known ceratopsid, 413, 414, 428

index

619

Triceratops prorsus: in ceratopsian cladistics, 410; Coahuilaceratops magnacuerna n. gen. & sp. versus, 109; from Hell Creek Formation, 203; stratigraphy, 412; Tatankaceratops sacrisonorum n. gen. & sp. versus, 208; taxonomy, 552 Trichomonas: osteomyelitis and, 348 Tricolpites microreticulatus: as KikakTegoseak Quarry palynomorph, 462 Tricolpopollenites parvulus: as KikakTegoseak Quarry palynomorph, 461, 462 Tricolporate pollen: as Kikak-Tegoseak Quarry palynomorph, 462 Trike II Triceratops locality, 557, 558, 559 Trilobosporites crassus: as Kikak-Tegoseak Quarry palynomorph, 462 Tringa: eye size, 312 Trionychids: in Big Bend microsites, 533; in Rattlesnake Mountain microsites, 524, 525 Triprojectates: Prince Creek Formation, 473 Troodon: Big Bend microsites and, 533; crepuscular lifestyle, 311; occurrence of, 416 Troodon formosus: in high-latitude ecosystems, 457; at Kikak-Tegoseak Quarry, 467 Troodontids: in Hilda mega-bonebed, 505 Trophic overlap: among ceratopsids, 420 Tropic, 119 Tropic Shale Formation: ceratopsian family tree and, 134 Trudopollis meekeri: as Kikak-Tegoseak Quarry palynomorph, 461, 462 Trunk channel facies association: at KikakTegoseak Quarry, 460–461 Tuatara: nocturnal lifestyle of, 320; sclerotic rings, 311 Tuberculosis: osteomyelitis in, 348 Tübingen: David Wesihampel at, 8 Tuchengzi Formation: Chaoyangsaurus youngi from, 235 Tugriken Shireh: ceratopsian paleoenvironmental associations and taphonomy in, 442; insect trace fossils from, 509–510, 510, 511, 512, 516, 516, 517; Protoceratops from, 308, 316–317, 321– 322, 322, 324, 432, 433 Tugrugiin Shireh. See Tugriken Shireh Tugrugyin Member: Protoceratops from, 511 Tugulu Formation: ceratopsian paleoenvironmental associations and taphonomy in, 440 Tugulu Group: psittacosaurs from, 38

620 index

Tullock Formation: Hell Creek Formation versus, 553; Hell Creek Project and, 552; Triceratops skull and, 272 Tuluwak Tongue: of Prince Creek Formation, 457 Tumanova, Tatiyana: on Developmental Mass Extrapolation, 5 Tumarkin-Deratzian, Allison R., xix, xxii, 251; Peter Dodson and, 7, 7, 10 Turanoceratops: in ceratopsian cladistics, 400; ceratopsian dispersal and, 399; in ceratopsid paleobiogeography, 389; Diabloceratops n. gen. versus, 134; Late Cretaceous North American ceratopsids and, 181; provenance, 387, 389, 390, 391, 431 Turanoceratops tardabilis: distribution, 406; stratigraphy, paleoenvironmental associations, and taphonomic studies, 442 Turbinals: ceratopsian, 232 Turgai Sea: Mesozoic paleogeography of, 394, 395, 396, 397, 401 Turkeys: osteomyelitis in, 348 Turnover. See Species turnover Turonian stage: basal ceratopsians from, 390, 392, 418, 431; ceratopsian cladistics and, 400; ceratopsian family tree and, 134; ceratopsian paleoenvironmental associations and taphonomy in, 442; ceratopsid cladistic analysis and, 110; in ceratopsid fossil record, 419; Turanoceratops from, 181; Zuniceratops christopheri from, 91, 92; Zuniceratops from, 187 Turtles: in Big Bend microsites, 533; from Big Bend National Park, 521; in Hilda mega-bonebed, 502; psittacosaurs versus, 334; in Rattlesnake Mountain microsites, 524, 525 Two Medicine centrosaurine: in ceratopsian cladistics, 409; stratigraphy, 412. See also Rubeosaurus entries Two Medicine Formation: Achelousaurus from, 142, 152; basal neoceratopsians from, 83, 88; bonebeds in, 448, 450, 450, 451, 452; centrosaurines from, 141; Cerasinops from, 88; ceratopsian paleoenvironmental associations and taphonomy in, 441, 442, 443; ceratopsid species diversity and turnover in, 421–422; in ceratopsid stratigraphy, 412; ceratopsids from, 7, 181–182, 413; dating, 413; dinosaurs from, 118; geologic correlation, 84; Kaiparowits Formation versus, 478, 479, 491; in North American paleogeography, 415; pachyrhinosaurs from, 408; paleoenvironments of, 415, 416,

417; Rubeosaurus ovatus n. gen. & comb. from, 156–168; stratigraphy, 163; Styracosaurus ovatus from, 408; sympatric ceratopsians from, 302–303; in Wahweap Formation stratigraphy/ dating, 130 Two Rocks Balanced hadrosauromorph: Zuniceratops christopheri and, 92 Two Rocks Balanced outcrop area: Zuniceratops christopheri and, 92, 92 Two-dimensional animation: in chasmosaurine horn and frill morphology analysis, 282, 284, 285–290, 285, 286, 287, 288, 289 Tyrannosaurians. See Tyrannosaurids Tyrannosaurids: from Aguja Formation, 521; from Aguja Formation microsite, 529, 530–531, 530; in Agujaceratops bonebeds, 520; KBP discoveries of, 479, 481; in Kikak-Tegoseak Quarry bonebed, 456; nutritional stress in, 365; in Rattlesnake Mountain microsites, 525. See also Tyrannosauroids; Tyrannosaurs Tyrannosauroids: Zuniceratops christopheri and, 92 Tyrannosaurs: caudal injuries in, 368; collected from Bayn Dzak, 320, 322; dentition, 316; orbit shapes, 311; in Protoceratops fauna, 322; species diversity and turnover among, 422 Tyrannosaurus: Triceratops frill as protection from, 282; Triceratops versus, 555 Tyrannosaurus rex: bite marks, 273; oldest known, 553–554, 553; osteomyelitis in, 347; relative eye size, 311, 312; skeleton and skull, 555 Tyrrell, J. B.: Albertosaurus skull collected by, 542 Tyrrell Museum of Palaeontology (TMP), xiii; Calgary Public Museum collection and, 546, 547; ceratopsian paleopathology studies at, 356; ceratopsid bonebeds excavated by, 377; Dinosaur Systematics Symposium at, 9; fieldwork in Mexico by, 100; Hilda mega-bonebed studies by, 496–497, 502; Medusaceratops lokii n. gen. & sp. material in, 182, 183; Pachyrhinosaurus collected by, 141, 153; Pachyrhinosaurus n. sp. at, 142 Tyson, Helen: on ceratopsian frill mechanics, 264, 269 Udanoceratops: in ceratopsian cladistics, 392, 400; ceratopsian dispersal and, 399; in ceratopsian paleobiogeography, 398, 399; mandible,246; Montanoceratops

cerorhynchus versus, 70, 71, 73, 74, 75, 76, 78, 79; in Montanoceratops cladistics, 78, 80, 81; Montanoceratops versus, 70; provenance, 387, 389, 390, 431 Udanoceratops tschizhovi: Montanoceratops cerorhynchus versus, 74; provenance, 391; stratigraphy, paleoenvironmental associations, and taphonomic studies, 442 Udan-Sayr: ceratopsian paleoenvironmental associations and taphonomy in, 442 Ukhaa Tolgod: ceratopsian paleoenvironmental associations and taphonomy in, 440; Protoceratops from, 308, 316–317, 321–322, 322, 432, 511; taphonomy, 432 Ulan Bator: dinosaur expeditions from, 509, 510; Protoceratops at, 319; psittacosaurs from, 22 Ulnae: of Australian ceratopsian, 391, 393; ceratopsid, 196, 197; in Chasmosaurus locomotion, 350, 350; CMN 8547, 193, 193, 194, 195, 196, 197; pathologies in Centrosaurus, 372; psittacosaur, 332; in psittacosaur swimming, 335; Tatankaceratops sacrisonorum n. gen. & sp., 211 Umiat: Kikak-Tegoseak Quarry near, 457, 459 UMNH VP 12198: taphonomy, 485–487, 486, 487. See also Utah Museum of Natural History (UMNH) UMNH VP 16699, 120–122, 122, 123–128, 123, 124, 125, 126; UMNH VP 16704 versus, 128–129, 134–136 UMNH VP 16704, 128–129, 129; UMNH VP 16699 versus, 128–129, 134–136 UMNH VP 16800, 489–490, 489. See also Kaiparowits ceratopsid C UMNH VP Locality 145: Kaiparowits ceratopsid B from, 485–487, 486, 491 UMNH VP Locality 277: Kaiparowits ceratopsid B from, 488 UMNH VP Locality 480: Kaiparowits ceratopsid B from, 488 UMNH VP Locality 512: undiagnosed ceratopsids from, 491 UMNH VP Locality 662: Kaiparowits ceratopsid B from, 488 UMNH VP Locality 684: Kaiparowits ceratopsid B from, 488 UMNH VP Locality 890: Kaiparowits ceratopsid A from, 484, 485 UMNH VP Locality 925: undiagnosed ceratopsids from, 491 UMNH VP Locality 940: Kaiparowits ceratopsid C from, 489–490, 489

UMNH VP Locality 942: Kaiparowits ceratopsid B bonebed at, 486, 487–488, 491 UMNH VP Locality 945: undiagnosed ceratopsids from, 490, 491 UMNH VP Locality 951: Kaiparowits ceratopsid A from, 484, 485 UMNH VP Locality 960: Kaiparowits ceratopsid B from, 488 Undiagnosed ceratopsids: Kaiparowits Formation, 490, 491, 492 Unfilled grooves: as cornual sinus vascular structure, 276 Unguals: Aguja Formation microsite, 529; Centrosaurus apertus, 343; Chasmosaurus, 343; CMN 8547, 190, 195, 196, 197; Montanoceratops cerorhynchus, 76; Protoceratops andrewsi, 324; psittacosaur, 329, 332; Tatankaceratops sacrisonorum n. gen. & sp., 211; Zuniceratops christopheri, 95, 96 Ungulate horns: in thermoregulation, 323 United States, 183; Agujaceratops bonebeds in, 520; basal ceratopsians from, 390; basal neoceratopsians from, 83–85; centrosaurine squamosals from, 135; ceratopsian bonebed taphonomy in, 432; ceratopsians from, 429, 566; dinosaur discoveries in, 555; sympatric ceratopsians from, 302–303; Tatankaceratops sacrisonorum n. gen. & sp. from, 203– 218 United States Army: in Kikak-Tegoseak Quarry fieldwork, 457 United States National Museum (USNM): New Mexico ceratopsian material in, 169; studied specimens at, 312; Triceratops skull at, 265–266 Units: Kaiparowits Formation, 481–482 Unity, Saskatchewan: shoreline deposits at, 507 University of Alaska: Kikak-Tegoseak Quarry discovered by, 457 University of Alberta, xiii; Peter Dodson and, 3 University of Calgary, 544 University of California Museum of Paleontology (UCMP): Triceratops studies by, 552, 554 University of Kansas Natural History Museum and Biodiversity Research Center: insect scavenging research at, 510– 511, 516 University of Manitoba: Calgary Public Museum collection and, 546, 547

University of Oklahoma: Wahweap Formation explorations by, 118 University of Ottawa: Peter Dodson and, 3, 4 University of Pennsylvania: Peter Dodson at, 6 University of Texas at Austin, xiii University of Toronto: ceratopsid correspondence at, 198 University of Washington Burke Museum: studied specimens at, 312 Unweighted Pair-Group Method with Arithmetic Averages (UPGMA): in dinosaur paleobiogeography, 389 Upchurch, Paul: on dinosaur paleobiogeography, 389 Upland facies: as ceratopsid paleoenvironments, 416 Upper Jurassic: Peter Dodson’s work on, 7. See also Late Jurassic epoch Upper Scollard Formation: in CMN 8547 stratigraphy, 191 Upper Shale Member (Aguja Formation), 523; dating, 413 Upper unit: Kaiparowits Formation, 482 Upright stance: in ceratopsid locomotion, 341–342, 349–350, 350–351 Urho: psittacosaurs from, 38 Urodeles: Grand Staircase–Escalante localities and distribution, 119 Uromastix: dentition, 316 USNM 11869, 156, 157, 157, 162, 165, 165; stratigraphy, 161 Utah, 119, 480; bonebeds in, 448, 450, 451, 452; ceratopsian paleoenvironmental associations and taphonomy in, 443, 445, 446; ceratopsid distribution in, 99, 111; in ceratopsid fossil record, 418, 422; ceratopsid stratigraphy in, 412; ceratopsids from, 413–414; chasmosaurines from, 408–411; dating fossils in, 394; dating Kaiparowits Formation in, 411– 413; Diabloceratops eatoni n. gen. & sp. from, 117–140; insect modification of dinosaur bone from, 509; Kaiparowits ceratopsid A from, 408; in Mesozoic paleogeography, 396–397, 399; paleoenvironments of, 416, 417, 418; taphonomy of Kaiparowits ceratopsids from, 478–494; Torosaurus latus from, 417; Wahweap centrosaurine from, 408; Zuniceratops christopheri and, 92 Utah Friends of Paleontology: Wahweap Formation explorations and, 118 Utah Geological Survey (UGS): Wahweap Formation explorations and, 118

index

621

Utah Museum of Natural History (UMNH): Diabloceratops eatoni n. gen. & sp. at, 118; fieldwork in Mexico by, 100; Kaiparowits Formation joint survey with BLM, 478–479, 482, 483. See also UMNH entries Utah neoceratopsian, 391, 393, 399 Uzbekistan: ceratopsian paleoenvironmental associations and taphonomy in, 440, 441, 442; ceratopsians from, 406, 429, 431; Turanoceratops from, 181 Valanginian stage: basal ceratopsians from, 390; ceratopsian cladistics and, 400; psittacosaurs from, 329 Variability. See Individual variation Varriale, Frank: Peter Dodson and, 6, 7 Vascular bone, 258; on Triceratops braincase, 279; in Triceratops horns, 271, 275– 276, 276–278, 277, 278, 279–280 Vegetation: ceratopsids and, 414; in KikakTegoseak Quarry bonebed paleoenvironment, 456, 472–473. See also Plants Velafrons coahuilensis: in Mexican stratigraphy, 100 Velociraptor: collected from Bayn Dzak, 320, 322; Protoceratiops with, 432 Venenosaurus: in Mesozoic paleogeography, 398 Ventrolateral flange: psittacosaur, 43 Vertebrae: Aguja Formation microsite, 529; ankylosaur, 528; Archaeoceratops yujingziensis n. sp., 64, 65; ceratopsid, 196, 199; CMN 8547, 189, 192–194, 192, 193, 194. 196, 198, 199; Montanoceratops cerorhynchus, 69, 70, 74–75, 75, 76, 77; osteomyelitis in, 347; pathologies in ceratopsid, 365–366, 365, 366–367, 366, 367, 367, 368–372, 368, 370, 371; Protoceratops displaying insect damage, 514; protoceratopsian, 329; of semi-aquatic ceratopsids, 199; Tatankaceratops sacrisonorum n. gen. & sp., 211; Zuniceratops christopheri, 95, 96 Vertebrate fossils: in ‘‘Purple Hill’’ section, 525–527, 526, 527 Vertebrate microfossil assemblages: Kaiparowits Formation, 480, 482, 483, 484 Vertebrate paleofaunas: from Kaiparowits Plateau, 118, 119 Vertebrate paleontology: finite element modeling in, 265 Vertebrate remains: from New Mexico, 169–171 Vertebrates: in Agujaceratops bonebeds, 520; in arid environments, 323; in Big

622 index

Bend microsites, 533; in Big Bend National Park geology, 521; biogeography of, 418; in Calgary Public Museum collection, 546; ceratopsid evolution versus, 421; Hell Creek Formation, 553; in high-latitude ecosystems, 456–457; in Hilda mega-bonebed, 495, 502, 505; inlife resting postures of fossil, 330; Kaiparowits Formation, 479, 481, 481; in Mexican stratigraphy, 100; Prince Creek Formation and, 458; in Rattlesnake Mountain microsites, 524–527, 526, 527; sclerotic rings among, 329 Veterinary gross anatomy: paleontology and, 6 Vicariance patterns: in ceratopsian paleobiogeography, 388–389 Vickaryous, Matt, xxii Vipers: nocturnal, 320 Virgelle Shale: in ceratopsid stratigraphy, 412 Virtual Triceratops Project, 266 Visceral gout, 349 Vision. See Eyes Visual acuity: eye size allometry and, 313– 314; of predators, 313–314, 315; Tyrannosaurus rex, 311 Visually hunting predators: eyes of, 308 Voids: at Kikak-Tegoseak Quarry, 468, 469 Volcanic mudflows: Psittacosaurus burial in, 432 Volcano-lacustrine deposits: psittacosaurs from, 330 Vomers: basal ceratopsian, 221, 222, 225, 230–231; Pachyrhinosaurus n. sp., 145; psittacosaur, 42 von Mises stress: in Triceratops frill, 267, 268 Vulpes macrotis: in arid environments, 323 Vultures: eye sizes, 312 Wahweap centrosaurine: in ceratopsian cladistics, 409; in ceratopsid fossil record, 418, 419, 422; as earliest known ceratopsid, 413; long postorbital horns, 283; Medusaceratops lokii n. gen. & sp. versus, 183; Rubeosaurus ovatus n. gen. & comb. versus, 162; skull and cladogram, 295, 296; skull strengths and measurements, 299, 300, 301, 302, 303, 304, 305; stratigraphy, 412; stratigraphy, paleoenvironmental associations, and taphonomic studies, 443; taxonomy, 418; taxonomy and distribution, 408; Zuniceratops christopheri versus, 96. See also Diabloceratops entries

Wahweap Formation, 119; ceratopsian family tree and, 134; ceratopsian paleoenvironmental associations and taphonomy in, 443; in ceratopsid stratigraphy, 412; dating, 413; Diabloceratops eatoni n. gen. & sp. from, 117–140, 119, 120, 121; exploration of, 118; Kaiparowits Basin Project and, 479; in North American paleogeography, 415; stratigraphy/dating, 129–131 Wahweap Project, 118 Wahweapian Land Vertebrate Age (LVA): in Wahweap Formation stratigraphy/ dating, 131 Walk cycle: modeling ceratopsid, 351–352 Wallace, W. A., 198 Walters, Robert: Peter Dodson and, 10 Wankel Tyrannosaurus rex, 555 Wanthaggi Formation: ceratopsian paleoenvironmental associations and taphonomy in, 442 Wapiti Formation: bonebeds in, 448, 450; ceratopsian paleoenvironmental associations and taphonomy in, 443, 444; in North American paleogeography, 415; pachyrhinosaurs from, 408; Pachyrhinosaurus n. sp. (Grande Prairie) from, 141, 152, 378, 379 Wapiti pachyrhinosaur A: in ceratopsian cladistics, 409; stratigraphy, 412. See also Pachyrhinosaurus n. sp. (Grande Prairie) Wapiti pachyrhinosaur B: in ceratopsian cladistics, 409; stratigraphy, 412 Warping: in ceratopsian skull and jaw analysis, 294, 296 Washington, D.C., 544 Wasps: bone modification by, 517 Water: arid environments and, 323 Water resources: bonebed formation and, 451 Wayan Formation: basal ceratopsians from, 387 Wear facets: on dentition, 316, 319 Weathering: in Hell Creek Formation, 554, 555, 556, 557, 558, 560; of Hilda mega-bonebed fossils, 495; of Kaiparowits ceratopsid B bones, 487; of Kikak-Tegoseak Quarry fossils, 469, 471 Webbed manus: psittacosaur, 332, 335 Weber State University: Wahweap Formation explorations by, 118 Weevils: pupation chambers, 517 Weishampel, David B.: Peter Dodson and, 6, 7, 8, 9 Welsh, Ed: at Rattlesnake Mountain, 524

Western Interior Basin (WIB): biogeography of, 416–417, 418; ceratopsians from, 566; ceratopsid distribution throughout, 99–100, 111; in ceratopsid fossil record, 418, 419; ceratopsid species diversity and turnover and, 419–422, 422–423; ceratopsids from, 405, 406, 407, 414–415, 416; fossils from, 100; Hilda mega-bonebed and, 497; Kaiparowits Formation and, 478, 491; Torosaurus latus from, 417; Triceratops from, 551, 556 Western Interior Seaway: in Big Bend National Park geology, 521; in ceratopsian paleoenvironments, 428, 431; ceratopsid distribution and, 99; ceratopsids from, 181; Hell Creek Formation and, 552, 554; Hilda mega-bonebed and, 499; Late Cretaceous ceratopsids and, 414–416, 415; in Tatankaceratops sacrisonorum n. gen. & sp. biostratigraphy, 205; Zuniceratops christopheri and, 92. See also Cretaceous Western Interior Seaway (KWIS) Whitemud Formation: in CMN 8547 stratigraphy, 191; geologic correlation, 84 Willow Creek Formation: basal neoceratopsians from, 83, 85, 88; geologic correlation, 84 Willow Wash, 170, 170; Ojoceratops fowleri n. gen. & sp. from, 171 Winnipeg: Calgary Public Museum collection and, 546, 547; letters from Cutler to Woodward from, 548–549 Witmer, Larry M., xxii; Peter Dodson and, 6, 7 Wolfe, Douglas G., xix, 91 Wolverton, Brad, 137 Woodward, A. Smith: correspondence from William Cutler to, 547–549 World War I: William Cutler during and after, 542 World War II: Calgary Public Museum collection during, 544 Works Progress Administration (WPA): Aguja fossils collected by, 528 Wrasses: psittacosaurs versus, 335 Wrists: modeling in locomotion studies, 351 Wu, Xiao-chun, xxii Wuerho: psittacosaurs from, 38 Wyoming: Big Bend dinosaurs versus those of, 521; ceratopsian distribution in, 111; ceratopsian paleoenvironmental associations and taphonomy in, 445, 446; insect modification of dinosaur bone

from, 509; Leptoceratops gracilis from, 243; Triceratops from, 551, 555; Triceratops horridus from, 417 Wyoming Dinosaur Center (WDC): Medusaceratops lokii n. gen. & sp. at, 182, 183 Xantusiids: nocturnal, 320 Xinjiang Province: ceratopsian paleoenvironmental associations and taphonomy in, 439; psittacosaurs from, 26, 329; Yinlong downsi from, 236 Xinjiang Uyghur (Uygur) Autonomous Region: Archaeoceratops localities in, 60; psittacosaurs from, 38; Psittacosaurus speciesfrom, 238 Xinmingpu Formation: ceratopsian paleoenvironmental associations and taphonomy in, 439 Xinmingpu (Xinminpu) Group: Archaeoceratops yujingziensis n. sp. from, 59, 60; Auroraceratops rugosus from, 241; basal ceratopsians from, 222, 223; ceratopsian paleoenvironmental associations and taphonomy in, 440; dinosaurs from, 59 Xuanhuaceratops: as basal ceratopsian, 221, 222; discovery, 11; Montanoceratops cerorhynchus versus, 70; in Montanoceratops cladistics, 78, 80, 81; skull and mandible, 234; taxonomy, 429 Xuanhuaceratops niei: stratigraphy, paleoenvironmental associations, and taphonomic studies, 439 Yale Computer Center, 5 Yale University: Peter Dodson at, 4–5 Yamaceratops: discovery, 11; in ceratopsian cladistics, 392, 400; Montanoceratops cerorhynchus versus, 70, 71, 73, 76; in Montanoceratops cladistics, 78, 80, 81; Prenoceratops sp. versus, 87; provenance, 387, 389, 390, 431; skull and mandible, 234 Yamaceratops dorngobiensis: stratigraphy, paleoenvironmental associations, and taphonomic studies, 442 Yaverlandia bitholus: in Mesozoic paleogeography, 398 Yinlong: as basal ceratopsian, 221, 222; in ceratopsian cladistics, 400; in ceratopsian paleogeography, 394; ceratopsian mandibles versus that of, 247; Chaoyangsaurus youngi versus, 235, 236; discovery, 11; mandible, 236–237; in Montanoceratops cladistics, 78, 80, 81; prove-

nance, 387, 389, 390, 428, 430; Psittacosaurus versus, 41, 52, 53–54, 56; skull and mandible, 234; taphonomy, 430; taxonomy, 429 Yinlong downsi: coronoid/mandible measurements, 248; mandible, 236–237; mandibular element measurements for, 236; provenance, 391; Psittacosaurus versus, 54; stratigraphy, paleoenvironmental associations, and taphonomic studies, 439; studied specimens, 235 Yixian Formation: basal ceratopsians from, 222, 223; ceratopsian paleoenvironmental associations and taphonomy in, 439, 440, 441; Hongshanosaurus houi from, 238; Liaoceratops yanzigouensis from, 245; psittacosaurs from, 22–23, 26, 28, 32, 53; Psittacosaurus taphonomy in, 432 You, Hai-Lu, xix, 11, 59, 221, 234; Peter Dodson and, 7, 7, 10–11 Ypresian stage: Kikak-Tegoseak Quarry palynomorphs from, 461 Yujingzi Basin: Archaeoceratops yujingziensis n. sp. from, 59, 60; dinosaurs from, 66, 67 Yumenzhen, 60 Z coal complex: Hell Creek Formation and, 552, 556, 558 Zangerlia: from Djadokhta Formation, 323 Zhangye, 60 Zhao, Xijin: on psittacosaurs, 21 Zippi, Pierre, xix, 456 Zircon dating: in Wahweap Formation stratigraphy/dating, 129–130 Zizhuqu: ceratopsian paleoenvironmental associations and taphonomy in, 442 Zuni Basin: Zuniceratops christopheri from, 92 Zuniceratops: biostratigraphy, 187; Coahuilaceratops magnacuerna n. gen. & sp. versus, 108–109; Diabloceratops eatoni n. gen. & sp. versus, 117, 127, 131, 132, 133, 135, 136; in ceratopsian cladistics, 392, 400; in ceratopsian family tree, 134; in ceratopsid cladistic analysis, 110, 112; long postorbital horns, 283; Medusaceratops lokii n. gen. & sp. versus, 186; in Montanoceratops cladistics, 78, 80, 81; occurrence of, 414, 418; Prenoceratops sp. versus, 88; provenance, 387, 390, 391, 431; in Tatankaceratops sacrisonorum n. gen. & sp. cladistics, 213, 216, 216; taxonomy, 418, 429; Turanoceratops versus, 391

index

623

Zuniceratops christopheri, 91–98; in ceratopsian cladistics, 409; in ceratopsid fossil record, 419; description, 93–95, 95– 96; discovery, 91; holotype material, 91, 92–93, 92, 93, 93–95, 95, 95–96, 96;

624 index

ontogeny, 95–96; phylogeny, 96; postcranial elements, 91, 95, 96; provenance, 392; skeleton, 96; skull elements, 91, 92, 93–95, 93, 94, 95, 96; stratigraphic context/occurrence, 92–93, 92, 96; stratigra-

phy, paleoenvironmental associations, and taphonomic studies, 442; taxonomy, 91 Zygogonium tunetanum: as Kikak-Tegoseak Quarry palynomorph, 462