Functional Morphology and Diversity (The Natural History of the Crustacea) [1] 0195398033, 9780195398038

Citation preview

Functional Morphology and Diversity

The Natural History of the Crustacea Series SERIES EDITOR: Martin Thiel Editor ia l A dvisory Boar d: Geoff Boxshall, Natural History Museum, London, UK Emmett Duffy, Virginia Institute of Marine Sciences, Gloucester, USA Darryl Felder, University of Louisiana, Lafayette, USA Gary Poore, Victoria Museum, Melbourne, Australia Bernard Sainte-Marie, Fisheries and Oceans Canada, Mont-Joli, Canada Gerhard Scholtz, Humboldt University Berlin, Berlin, Germany Fred Schram, Friday Harbor Marine Laboratory, Seattle, USA Les Watling, University of Hawaii, Hawaii, USA Functional Morphology and Diversity (Volume 1) Edited by Les Watling and Martin Thiel

Functional Morphology and Diversity The Natural History of the Crustacea, Volume 1

EDITED BY LES WATLING AND MARTIN THIEL

1

3 Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trademark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016

© Oxford University Press 2013 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. Library of Congress Cataloging-in-Publication Data The natural history of the Crustacea. p. cm. Includes bibliographical references and index. ISBN 978-0-19-539803-8 (hardcover: alk. paper) Martin, 1962– QL435.N38 2013 595.3—dc23

9 8 7 6 5 4 3 2 1 Printed in the United States of America on acid-free paper

1. Crustacea.

I. Watling, Les.

II. Thiel, 2012012251

PREFACE

Many years ago, when M.T. came to Maine to work with L.W. on shallow subtidal crustaceans, we established some amphipod populations in aquaria at the University of Maine’s Darling Marine Center flowing seawater facility. What ensued was not just a series of remarkable behavioral observations but also a series of long discussions of “how do they do that?” in response to the burrowing and whip making that we observed. The two of us have maintained a lifelong interest in understanding how crustacean morphology “determines” what the animals can do and where they can live. Crustaceans encompass a bewildering array of body forms that they use to occupy almost every habitat type on Earth. In fact, the only habitat not occupied by crustaceans is the open air; that is, they cannot fly. Understanding what modifications of the crustacean body plan have allowed this diversification is the subject of this book. In its basic form, the crustacean body consists of a bilaterally symmetrical, segmented, more or less tubular body, with each segment bearing a pair of appendages. The number of body segments and the design or even presence of appendages vary considerably, with the most extreme cases being found in those groups that have invaded the bodies of other animals. In this book we have asked the chapter authors to explore this diversification of form and to explain how various parts of the crustacean body work. While many authors have examined the functional morphology and anatomy of crustaceans in individual publications or book contributions, this has never been done in an integral way in one volume. We were particularly interested in understanding the design limitations of the crustacean body, for example, how living in a dense fluid medium might restrict the movement capabilities of the animal, and how that would vary depending on the animal’s size. But movement isn’t all that crustaceans do in their daily lives—they have to eat, respire, reproduce, and grow, all of which needs to be controlled so that the animal functions as a successfully coordinated whole. To set the stage, Schram (chapter 1) reviews the range of crustacean body plans, and Haug et al. (chapter 2) review what is known about appendage development in the earliest arthropods as they become what we would recognize now as crustaceans. Then follow four chapters that examine the exterior of the crustacean body. Williams (chapter 3) investigates the genetic control of appendage development as the animal develops. Olesen (chapter 4) surveys the functional constraints of the archetypical crustacean structure, the carapace. Dillaman et al. (chapter 5) take a detailed look at the crustacean cuticle, and Garm and Watling (chapter 6) review the structure and function of setae and other cuticular outgrowths.

v

vi

Preface Our examination of the integrated functional aspects of the crustacean body starts with Boxshall and Jaume’s overview of crustacean antennae (chapter 7). Watling (chapter 8) integrates food consumption and digestion, showing that mouthpart appendage structure, foregut morphology, and digestive enzyme secretion are all related to diet. Belanger (chapter 9), Faulkes (chapter 10), Yen (chapter 11), and Boudrias (chapter 12) examine modes of locomotion in crustaceans and the functional constraints imposed on locomotory appendages by the physics of the medium in which they live. We end the book with a series of chapters dealing with system-level integrative functional anatomy. Bauer (chapter 13) reviews the structure and function of appendages used for grooming and reproduction. Wirkner and Richter (chapter 14) examine the integration of respiratory structures with internal details of the circulatory system. An overview of the internal anatomy of the reproductive system is provided by L ópez Greco (chapter 15). And lastly, the coordination of the whole body is dealt with in Sullivan and Herberholz’s review of crustacean nervous systems (chapter 16). We have asked the authors of these chapters to deal with their part of the crustacean body in isolation. Of course, we recognize that crustaceans are living creatures, and in order to live they have managed to integrate the functional aspects of their bodies that we have studied so well on their own. What we have not done is try to develop a fully integrated model of the crustacean body that accounts for all the physical aspects of the various habitats in which crustaceans live. We suspect this is possible but may be beyond the reach of any one of us.

ACKNOWLEDGMENTS

Our thanks go foremost to all contributors who attended to all our requests during the preparation of this book. Their expertise and their willingness to invest time into the writing of their contributions make up the value of this book. We especially thank our editorial assistants, Ivan Hinojosa and Lucas Eastman, who expertly revamped many of the figures and who carefully read and edited all the chapters. The generous contribution from Universidad Católica del Norte was essential for this project—we are very grateful for the continuous support which allowed us to focus on the task. The vision and foresight of the university authorities made this project possible and we hope that this and the upcoming volumes fulfil their expectations. We also thank those colleagues who read entire or parts of chapters. The publisher, Oxford University Press, gave us a lot of freedom, and in particular, we express our appreciation to Peter Prescott, Tisse Tagaki, Jeremy Lewis, and Hallie Stebbins for their help during the past few years. L.W. would like to acknowledge the role of his students in stimulating him to think about crustaceans as functionally whole units, and his friend C. Nouvian, who listened with interest to many crustacean natural history stories while we were working on other things. Finally, we thank our families for their patience and interest in this project—without their tolerance, none of this would have been possible. Editing of this book was generously supported by Universidad Católica del Norte, Chile.

vii

CONTRIBUTORS

EDITORS Martin Thiel Facultad Ciencias del Mar Universidad Católica del Norte Larrondo 1281 Coquimbo Chile Les Watling Department of Biology University of Hawaii at Manoa 152 Edmondson Hall Honolulu, HI 96822 USA

viii

Michel A. Boudrias Department of Marine Science and Environmental Studies University of San Diego Science and Technology 267 5998 Alcalá Park San Diego, CA 92110 USA Geoff Boxshall Department of Zoology The Natural History Museum Cromwell Road London SW7 5BD UK

AUTHORS Raymond T. Bauer Department of Biology University of Louisiana PO Box 42451 Lafayette, LA 70504 USA

Richard M. Dillaman Department of Biology and Marine Biology University of North Carolina at Wilmington 601 South College Road Wilmington, NC 28403–5915 USA

Jim Belanger Department of Biology West Virginia University PO Box 6057 Morgantown, WV 26506 USA

Zen Faulkes Department of Biology The University of Texas–Pan American 1201 W. University Drive Edinburg, TX 78539 USA

Contributors Anders Garm Department of Cell and Organism Biology Lund University Helgonavagen 3 222362 Lund Sweden Carolin Haug University of Greifswald Zoological Institute and Museum Department of Cytology and Evolutionary Biology Soldmannstr. 23 17487 Greifswald Germany Joachim T. Haug University of Greifswald Zoological Institute and Museum Department of Cytology and Evolutionary Biology Soldmannstr. 23 17487 Greifswald Germany Jens Herberholz Department of Psychology and Neuroscience and Cognitive Science Program 2123H Bio-Psych Building University of Maryland College Park, MD 20742 USA Damià Jaume Instituto Mediterrá neo de Estudios Avanzados IMEDEA (CSIC-UIB) Miquel Marquès 21 07190 Esporles (Mallorca, Illes Balears) Spain Laura S. L ópez Greco Facultad de Ciencias Exactas y Naturales University of Buenos Aires Pabellón 2, Intendente Gü iraldes 2160 Ciudad Universitaria C1428EGA Buenos Aires Argentina

Andreas Maas Biosystematic Documentation University of Ulm Helmholtzstrasse 20 89081 Ulm Germany Shannon Modla Delaware Biotechnology Institute 15 Innovation Way, Suite 117 Newark, DE 19716 USA Jørgen Olesen Natural History Museum of Denmark (Zoological Museum) University of Copenhagen Universitetsparken 15 DK-2100 Copenhagen Ø Denmark Stefan Richter Universität Rostock Institut f ü r Biowissenschaften Abteilung f ü r Allgemeine und Spezielle Zoologie Universitätsplatz 2 18055 Rostock Germany Robert Roer Department of Biology and Marine Biology University of North Carolina at Wilmington 601 South College Road Wilmington, NC 28403–5915 USA Frederick R. Schram Burke Museum University of Washington at Seattle Post Box 1567 Langley, WA 98260 USA

ix

x

Contributors Thomas Shafer Department of Biology and Marine Biology University of North Carolina at Wilmington 601 South College Road Wilmington, NC 28403–5915 USA Jeremy M. Sullivan Department of Neurology Johns Hopkins University 855 North Wolfe Street 2nd Floor, Rangos Building Baltimore, MD 21205 USA Dieter Waloszek Biosystematic Documentation University of Ulm Helmholtzstrasse 20 89081 Ulm Germany Les Watling Department of Biology University of Hawaii at Manoa 152 Edmondson Hall Honolulu, HI 96822 USA

Terri A. Williams Department of Biology Trinity College 300 Summit Street Hartford, CT 06106 USA Christian S. Wirkner Universität Rostock Abteilung f ü r Allgemeine und Spezielle Zoologie Institut f ü r Biowissenschaften Universitätsplatz 2 18055 Rostock Germany Jeannette Yen School of Biology Georgia Institute of Technology 310 Ferst Drive Atlanta, GA 30332–0230 USA

CONTENTS

1. Comments on Crustacean Biodiversity and Disparity of Body Plans • 1 Frederick R. Schram 2. Evolution of Crustacean Appendages • 34 Joachim T. Haug, Andreas Maas, Carolin Haug, and Dieter Waloszek 3. Mechanisms of Limb Patterning in Crustaceans • 74 Terri A. Williams 4. The Crustacean Carapace: Morphology, Function, Development, and Phylogenetic History • 103 Jørgen Olesen 5. The Crustacean Integument: Structure and Function • 140 Richard M. Dillaman, Robert Roer, Thomas Shafer, and Shannon Modla 6. The Crustacean Integument: Setae, Setules, and Other Ornamentation • 167 Anders Garm and Les Watling 7. Antennules and Antennae in the Crustacea • 199 Geoff Boxshall and Damià Jaume 8. Feeding and Digestive System • 237 Les Watling 9. Appendage Diversity and Modes of Locomotion: Walking • 261 Jim Belanger

xi

xii

Contents 10. Morphological Adaptations for Digging and Burrowing • 276 Zen Faulkes 11. Appendage Diversity and Modes of Locomotion: Swimming at Intermediate Reynolds Numbers • 296 Jeannette Yen 12. Swimming Fast and Furious: Body and Limb Propulsion at Higher Reynolds Numbers • 319 Michel A. Boudrias 13. Adaptive Modification of Appendages for Grooming (Cleaning, Antifouling) and Reproduction in the Crustacea • 337 Raymond T. Bauer 14. Circulatory System and Respiration • 376 Christian S. Wirkner and Stefan Richter 15. Functional Anatomy of the Reproductive System • 413 Laura S. López Greco 16. Structure of the Nervous System: General Design and Gross Anatomy • 451 Jeremy M. Sullivan and Jens Herberholz Index • 485

Functional Morphology and Diversity

This page intentionally left blank

1 COMMENTS ON CRUSTACEAN BIODIVERSITY AND DISPARITY OF BODY PLANS

Frederick R. Schram

Abstract The science of natural history is built on twin pillars: cataloging the species found in nature, and reflecting on the variety and function of body plans into which these species fit. We often use two terms, diversity and disparity, in this connection, but these terms are frequently used interchangeably and thus repeatedly confused in contemporary discourse about issues of function and form. Nevertheless, diversity and disparity are distinct issues and must be treated as such; each influences our views of the evolution and morphology of crustaceans.

CRUSTACEAN DIVERSIT Y Crustaceans exhibit great disparity in basic body plans (I return to this subject below), but disparity of crustacean form is different from crustacean biodiversity, that is, the number of species we have within any particular group. No one knows for certain the exact number of species within any group of organisms, although the situation might improve with the appearance of online catalogs for particular groups. The people who set up these databases and maintain them as new species are added and old species are placed in synonymy provide a much-needed service toward adequately cataloging the tree of life. Nevertheless, as humans we like numbers—they are easily understood. So I have made my own tally (Table 1.1) and present a summary of estimates compiled from various authorities as to the total number of crustacean species. There is clearly no agreement on numbers among the authors listed in Table 1.1, although the estimates have gone up through time. With the exception of Minelli (1993) and Brusca and

Functional Morphology and Diversity. Edited by Les Watling and Martin Thiel. © 2013 Oxford University Press. Published 2013 by Oxford University Press.

1

2

Functional Morphology and Diversity Table 1.1. Various estimates of global numbers of species of crustaceans. Estimated number of species 44,950 32,000 68,171 40,000 39,000 75,000 55,364 38,000 49,658

Source Bouchet (2006) Brusca and Brusca (1990) Brusca and Brusca (2003) Groombridge and Jenkins (2000) May (1988) Meglitsch and Schram (1991) Minelli (1993) Ruppert and Barnes (1994) This chapter

Brusca (2003), who appear to have attempted a real count, the other authors obviously provided rounded off and rough estimates. For example, the number provided by Meglitsch and Schram (1991) was an estimate of what the highest number might be at some point in time when knowledge of the number of species will have reached a plateau. Although we know a great deal about groups of invertebrates, our knowledge is not very good and rather incomplete. I examined the patterns through time in documenting animal taxon diversity (Schram 2003) and noted several periods during which plateaus of relative inaction followed bursts in activity. It seems clear from these charted patterns that we are currently in one of those periods of increased activity, but whether we will soon reach a new plateau, or whether increased use of molecular techniques to identify monophyletic groups might continue to add new taxa at all levels—from phylum down to species—I cannot say. However, increasing application of molecular techniques does seem to indicate that we have underestimated the degree of cryptic speciation in nature. Having stated this, I feel honor bound by the charge given to me by the editors to provide my own numbers, so I tally here the currently known crustacean species. Table 1.2 is based on a census of relevant websites, currently available monographic literature, and the best estimates of authorities active in one or more of these groups. The reader should keep in mind that this is a tally of species numbers at this point in time, and these figures can only increase as our knowledge of these taxa evolves. In fact, the survey made by Martin and Davis (2006) seems to indicate that no asymptotes are yet emerging in the pace at which new species are being described. First, the total number of species obtained by this survey, 49,658, is not too far off from the estimates of Bouchet (2006) and Minelli (1993). Within that number, some things deserve special notice. Of the two largest groups on this list, Maxillopoda and Malacostraca, the numbers are of similar magnitude—almost 19,000 and something more than 29,000, respectively. The number of maxillopodans can only increase. The 9,500 copepods is only an estimate, although it may stand close to the actual numbers of currently described species. Nevertheless, copepod taxonomy is an active discipline, and increasingly sophisticated techniques of study will help isolate cryptic species. The 8,008 species of ostracodes is only an estimate, and if we factor in fossil species, we would more than double that number. Furthermore, the application of molecular methods in Crustacea will likely affect our understanding of species level biodiversity. For example, I am surprised at the relatively low number for the thecostracans, but parasitism is rampant in the group, and underestimates of species diversity would prevail in taxa with such

Crustacean Biodiversity and Disparity of Body Plans Table 1.2. Census of species numbers in various crustacean groups. Taxon Branchiura (Argulida) Branchiura (Pentastomida) Mystacocarida Branchiopoda Anostraca Cyclestherida Laevicaudata Notostraca Spinicaudata Cladocera Maxillopoda Copepoda Ostracoda Myodocopida Podocopida Thecostraca Ascothoracica Cirripedia Acrothoracica Rhizocephala Thoracica Facetotecta Tantulocarida Remipedia Cephalocarida Malacostraca Phyllocarida Stomatopoda Eumalacostraca Syncarida Bathynellacea Anaspidacea Peracarida Amphipoda Cumacea Isopoda Lophogastrida Mictacea Mysida sensu lato Mysida sensu stricto Stygiomysida Spelaeogriphacea Tanaidacea

Number of species 175 100 13 509 307 1 36 15 ≈150 450 18,911 9,500 ≈8,008 1,608 (+500 fossils) 6,400 (+9,500 fossils) 1,403 >99 1,304 >61 >255 948 12 28 19 10 29,471 39 456 28,976 >187 >170 17 15,686 6,950 1,342 5,270 56 5 1,085 1,075 10 4 940

3

4

Functional Morphology and Diversity Table 1.2. (Continued) Taxon Thermosbaenacea Eucarida Amphionidacea Decapoda Dendrobranchiata Caridea Stenopodidea Reptantia Euphausiacea Total

Number of species 34 13,103 1 13,016 522 2730 57 9,707 86 49,658

The higher taxonomic grouping of this table accords with the conclusions derived from the discussion of disparity of form given later in this chapter. Classes are shown in boldface.

highly reduced body forms. For example, rhizocephalans seem poised on the edge of a renaissance in interest, and the number of species anticipated will increase. Malacostraca constitutes a large number of species, but the species distribution is uneven because some subgroups are very large (amphipods, isopods, reptant decapods), while others are small (mictaceans, spelaeogriphaceans, and the amphionidacean). In fact, any group associated with cave or groundwater habitats appears likely at the lower end of species number estimates, but these habitats are difficult to study, and every attempt to sample these communities turns up new and interesting species, which can only continue into the future. (In this connection, one need only consider the work on crayfish in North America to see what happens when intensive systematic interest is focused on a group.) Some major class- and order-level taxa presently have low species numbers (remipedes and cephalocarids), but here, too, we have animals living in habitats that are difficult to sample (anchialine caves and the deep sea). Other groups contain very cryptic creatures living in places that, although well studied, nevertheless are often overlooked (mystacocarids in interstitial beaches). Because of the great disparity of body plans exhibited by crustaceans, we have a problem in comparing the species numbers in one group with another. The taxa in Table 1.2 are organized around the currently recognized class and order levels, but how does one compare ordinal differences seen in malacostracans with what are called orders within the maxillopodans? Recognizing a decapod from an amphipod is quite easy (both are orders of Eumalacostraca), but not many people could easily distinguish a cyclopoid from a calanoid (they are both orders of Copepoda) without being carefully schooled in the differences. Hence, trying to compare numbers of species within groups across the major taxonomic (class-level) units of crustaceans is truly like comparing apples to oranges or, in this case, lobsters to zooplankton. Nevertheless, strange patterns arise when we look within groups. Consider the peracaridans, for example. Why are there so many species of amphipods (6,900) compared to thermosbaenaceans (34) or mictaceans (5)—approximately two and three orders of magnitude difference? Are amphipods truly that much better adapted to their environments, an explanation often assumed to be true? If so, how and why? Or, are some other factors at play that might augment or possibly even ignore issues of adaptation? Some of these factors might be

Crustacean Biodiversity and Disparity of Body Plans 40 A Number of genera

30

20

10

0

5

10 15 20 25 Number of species in genus

100

Number of genera

B

10

1 1

10 Number of species in genus

100

Fig. 1.1. Arithmetic hollow curve (A) and log-log (B) plots of size distributions of genera of Stomatopoda (as number of included species) roughly conforming to power law N(x) = ax–b. For details about the method, see Minelli et al. (1991).

difficulty of habitat access for study (mentioned above), age of a clade, habitat heterogeneity, and expressions of chance in nature. The various authors of other volumes in this series will explore many of these issues. The element of chance plays an important role in classification. Willis and Yule (1922) and Minelli et al. (1991) observed that the size of supraspecific taxa as related to the included subtaxa (species in genera, genera in families, etc.) follows a power law. They concluded that the structure of biological classification is naturally fractal. This structure can be expressed as a hollow curve that, if plotted on a log-log scale, would conform to N(x) = ax–b. We can illustrate this with one example from Malacostraca, the unipeltate stomatopods (mantis shrimp). As of this writing, we recognize 456 species in 112 genera of mantis shrimp, with an additional 123 nominal species currently in synonymy. If we consider only the 456 recognized species, distribution numbers range from one of the largest genera, Nannosquilla, with some 26 species, down to 36 genera with but a single species each. Graphing this diversity, we can see that on an arithmetic scale it forms a hollow curve (Fig. 1.1A), and on a log-log plot a

5

6

Functional Morphology and Diversity straight line emerges (Fig. 1.1B). The fractal pattern becomes apparent when examining genera within families (data not shown), where we would again see a log-log plot that roughly matches that of species in genera. Whether this pattern appears in other groups of crustaceans remains to be tested, but I have no doubt that it will hold as it has in other groups of animals and plants. As humans, we are naturally inclined to seek causative explanations for patterns of biodiversity. However, I believe we do not necessarily need to explain why one particular genus, such as Nannosquilla with 26 species, is somehow better adapted than its confamilial sister genera, in this case Mexisquilla and Keppelius, each with only a single species. As we chart species biodiversity, we should be open to the possibility that the relative number of taxa within any particular group may represent nothing other than the manifestation of the operations of a stochastic, fractal universe, to say nothing of the vagaries of individual taxonomic decisions. Many authorities might reject my pessimism here, but at the very least, a stochastic, fractal biodiversity has to be one of several alternative hypotheses to consider.

CRUSTACEAN DISPARIT Y OF BODY PLANS The crustaceans are the most variable of all the arthropod groups; that is, there is a great disparity of body plans throughout their ranks (Fig. 1.2). If we are to assume that Crustacea is a monophyletic group, then they are not like any other arthropods. This high degree of variability is a very real problem with some serious implications, because if we take this disparity of form at face value, then we should seriously question whether all these various groups can constitute a single monophylum. When one looks at other major arthropod groups outside of Crustacea, there appears to be no great disparity of plan within these taxa (see Meglitsch and Schram 1991); members of each group fit a concise definition. For example, members of Insecta (Hexapoda) have a body divided into a five-segment head with the first postantennal segment bearing appendages modified as a labrum, a three-segment thorax with two sets of wings in the pterygote insects borne on the second and third segments, and an abdomen of 10–12 somites. All insects conform to this definition with some exceptions, for example, allowing for fusion of segments at the terminus of the abdomen or modification of wing arrangements. Insects have a unified body plan. Myriapoda as a whole do vary in some features such as body length but have in common that their trunk is not divided into a thorax and abdomen and that their gonopores are generally located on the anterior aspect of the trunk. The individual groups of myriapods conform to common plans: Symphyla have 12 trunk segments with the gonopores on the fourth somite; Pauropoda bear 12 trunk segments with the gonopore on the second somite; and Diplopoda, with several very distinct orders, all exhibit well-developed diplosomites, that is, pairs of segments fused dorsally but distinct ventrally, and their gonopores are located on the second trunk segment. The individual orders of diplopods vary only regarding the total number of trunk somites: pselaphognaths have at least 10–12, but colobognaths can exceed 30. Chilopoda have variable trunk segment numbers, extending from 15 to more than 180 pairs of legs, depending on group, but all chilopods without exception have long antennae and modify the first trunk limb as a fang equipped with a poison gland to facilitate their carnivory. Centipedes also uniquely bear gonopores on the posterior aspect of the trunk. The subphylum Cheliceriformes exhibits only a few “head” segments, essentially two, and these are fused with the anterior, or locomotory, part of the trunk to form a prosoma. The anteriormost somite (the one just posterior to the asegmental acron), the homolog of the antennal segment in other arthropods, does not carry antennae but rather is equipped with a pair of chelicerae. The second segment, what in other arthropods is referred to as the first postantennal

Crustacean Biodiversity and Disparity of Body Plans

Fig. 1.2. Disparate body types among crustaceans.

segment, typically bears a well-developed set of limbs, albeit variously developed. There are no exceptions to this basic format. Within the cheliceriforms, the highly distinctive Pycnogonida appear to be all legs, their prosoma reduced to a thin cylinder. The mouth is located terminally on a long proboscis. The small turreted “head” bears chelicerae, a second set of limbs called palps, and a third set of limbs modified as ovigers in the males. Posterior to these limbs, most pycnogonids utilize four pairs of legs for locomotion (a few forms with five or six pairs are known). All sea spiders conform to this body plan. Arachnida have a six-segment prosoma, with chelicerae, pedipalps, and four pairs of walking legs. The trunk bears an additional opisthosoma of diverse form but composed of some 13 somites, with the first segment greatly reduced as a narrow pedicel and the second bearing the gonopores. Opisthosomal limbs are missing or greatly reduced. Despite a great variety of body profiles, especially regarding the opisthosoma, all arachnids conform to this single plan. Merostomata, a small group today, was more extensive (and diverse) in the past. The prosoma bears six pairs of limbs. The chelicerae are followed by four sets of modest-sized walking limbs with specialized gnathobases, the first of which in the males is modified for

7

8

Functional Morphology and Diversity grasping the female during copulation, and a somewhat larger fifth set effective in grooming the underside of the prosoma. The next somite, the pregenital segment, is reduced but bears modified limbs, the chilaria. Unique among living cheliceriforms, the opisthosome of the merostomes bears six pairs of limbs posterior to the genital segment. All merostomes conform to this plan. From this short review, we can see that all these groups of arthropods have concise diagnoses, with distinct sets of apomorphies that characterize all members of the group. Crustacea, if viewed as a single group, simply does not have this.

CAN WE DIAGNOSE A MONOPHYLETIC CRUSTACEA? Any invertebrate zoology textbook can provide a set of characters for Crustacea. When I was asked to provide such a diagnosis 25–30 years ago (Schram 1979, 1982), I certainly did not hesitate. However, we now realize it is not sufficient to simply string together any list of characters. Ideally, as in the arthropod examples cited above, these characters should be unique derived features that diagnose all members of the group. To rephrase this in contemporary terms, a diagnosis should offer synapomorphies that together uniquely delineate a monophyletic group. We strive for natural taxonomies, classifications that reflect evolution. It is critical to determine if this is possible for Crustacea. A commonly accepted diagnosis of Crustacea consists of the following: (1) head of five somites, each bearing a set of appendages consisting of two pairs of antennae, a pair of mandibles, and two pairs of maxillae; (2) body consisting of three regions: head, thorax, and “abdomen”; (3) trunk appendages primitively multiramous; and (4) development consisting of a series of discrete larval and/or juvenile stages, initiated by a stage termed a nauplius. Let us inspect these features one by one in order to determine if these characters provide that unique set of descriptors we require for a diagnosis of Crustacea. In the discussion below, I restrict the term Crustacea to mean a monophyletic group and the term crustaceomorph to connote the amalgam of arthropod types that we generally and broadly refer to as “crustaceans” (fossil and recent) but that may or may not be monophyletic. Table 1.3 will assist the reader in following along the taxa and many of the relevant features discussed below. “Head of five somites, each bearing a set of appendages consisting of two pairs of antennae, a pair of mandibles, and two pairs of maxillae” These are not a unique set of features. A head consisting of five somites is shared with insects and the myriapods (see above; Meglitsch and Schram 1991), and, as would follow, most of the head appendages of these somites are shared among the three groups, namely, the first set of antennae, mandibles, and two sets of maxillae. It is only regarding the so-called second set of antennae that we might have a distinctive crustaceomorph feature since myriapods and insects lack a limb in this position. The second antennae are generally perceived as specialized sensory limbs and as such could serve as a defining apomorphy. This descriptor arises from the mental image of Crustacea conjured up by thinking of a shrimp or a lobster (a malacostracan), and this image without a doubt presents us with an icon of an arthropod with a set of sensory limbs at this position, albeit with slight anatomical variations, depending on group (Fig. 1.3A). Nevertheless, sensory second antennae are not characteristic of all groups of crustaceomorphs. Chapter 7 provides additional details concerning antennae, but a short overview will suffice here to make a point. Remipedes have quite distinctive limbs in this position (Fig. 1.3B) that I suspect serve equally as a hydrofoils to direct currents of water that f low

Table 1.3. Comparative morphology of various aspects that define body plans among various groups of crustaceomorphs. Taxa

Oligo-crustacea Branchiura Mystacocarida Skara Martinssonia Branchiopoda Laevicaudata Notostraca Spinicaudata Cyclestheria Cladocera Anostraca Lepidocaris Rehbachiella Waptia Eucrustacea Cephalocarida Maxillopoda Copepoda Cirripedia Ostracoda Bredocaris Malacostraca Eumalacostraca Hoplocarida Phyllocarida Remipedia

Head (no. somites)

5

Trunk regions

Postthorax

Thorax length (no. somites)

Segment no. of gonopore position

Trunk limb rami

First larval segments

Male

Female

4 5 1 3

4 4 ? ?

4 4 ? ?

2 1 2 2

? 4 ? 4

10–12 11 16–32 16–32 4–6

11 11 11 11

11 11 11 11

M M M M

12

12 11

12 ? ? ?

M 2, M 2

3 3 3 3 3 3

5 4

2 2 2 2

– a a a

5 5 5 5 5 5 5 5 5

1 ?2 1 1 2 2 2 2 2

– ?p – – – a a a a

4

2

a

8

6

6

M

5

5 5

2 1

a a

7 7

7 1 ?7

2 2

5

2

a

7

7 7 ?7 ?

2

3 3 3 4

5(+) 5(+) 5 6

2 2 2 1

p p p –

8 8 8 –

8 8 8 15

6 6 6 8

2 2 M 2

3 3 3 3

Abbreviations: a, abdomen; M, multiramous; p, pleon. Neither the list of taxa nor the features are exhaustive of all possibilities but cover the major elements discussed in the text.

3 ?

10

Functional Morphology and Diversity

B

A C

C'

D palp

anterior spine antennule

F E

antenna

posterior spine

G

Fig. 1.3. Various functional types of “antennal” limbs found in crustaceans. (A) Sensory: Apseudes hermaphroditicus, tanaid (from Lang 1953). (B) Swimming hydrodynamic plane: Lasionectes entrichoma, a remipede (from Schram et al. 1986). (C and C′) Locomotory/feeding: Derocheilocaris ingens, a mystacocarid (from Hessler 1969), antenna (C) and mandible (C′). The arrow indicates gnathal lobes. (D) Part of an attachment complex: Argulus foliaceous, a branchiuran (from Martin 1932). (E) Swimming: Archimisophria discoveryi, a copepod (from Boxshall 1983). (F) Host penetration: Gorgonolaureus muzikae, an ascothoracidan (from Grygier 1981). (G) Swimming/feeding: Bredocaris admirabilis, a Cambrian maxillopodan (from Mü ller and Walossek 1988).

around the head and perhaps also to aid in creating some of those currents by f lapping the hydroplane-like exopods. Although no work has yet been done on functional morphology of this limb, I believe it safe to say that the remipede antenna is not a purely sensory appendage. Mystacocarids have a pair of limbs behind the first antennae that are virtually identical in form to the mandibles (Fig. 1.3C,C′), except these so-called second antennae lack the gnathal armament at the base of the limb that occurs on the mandibles. The mystacocarid “antennae” are locomotory limbs. Branchiurans possess broad plates in this position with basal hooks and terminal, recurved spines—nothing sensory at all but rather serving to assist with attachment (Fig. 1.3D). Tantulocarids lack head appendages altogether. Copepods have well-developed limbs in this position that serve for the most part as the primary organs of swimming (Fig. 1.3E). It is difficult to specify what this limb does in ascothoracicans, where

Crustacean Biodiversity and Disparity of Body Plans all head limbs are highly modified to achieve attachment to a host or to penetrate host tissues (Fig. 1.3F). In cirripedes, the adults lack the antennae, but the nauplius and cypris larvae have limbs in this position to assist in swimming; the second antennae disappear at the time of attachment prior to metamorphosis to the adult cirripede. Finally, within the wide array of Cambrian microarthropods that are considered to bear some relationships to modern groups (see chapter 2), such as Bredocaris (Fig. 1.3G), Martinssonia , Rehbachiella , Skara , and Walossekia , the so-called second antennae are more often than not locomotory limbs, similar in structure to the mandibles and maxillae of these fossils. We can conclude from this brief survey that the only character that Crustacea share at this position, that is, the first segment posterior to the true antennae, is simply the presence of a pair of limbs. However, this is to say nothing—the mere presence of limbs on the first postantennal segment, or any postantennal segment for that matter, is a generalized, primitive, or plesiomorphic feature. As noted above, merostomes, pycnogonids, and arachnids also have a limb in this position, but that does not make them crustaceomorphs. In arthropods, all segments generally carry limbs, at least on the head and thorax; it is only when limbs are particularly specialized, or even missing, that things become more interesting and can serve to help diagnose a group. For example, the presence of a limb on this first somite posterior to the antennae in crustaceomorphs stands in contrast to what occurs in myriapods and insects. In these latter groups, the limb buds on the first postantennal somite are diverted from forming a limb into producing the special labrum seen in these groups. We know this is so because, at least for insects, developmental gene expression studies reveal that the labrum is the “appendage” of the so-called intercalary (first postantennal) segment (Boyan et al. 2002). This diversion of the first postantennal anlagen into forming the upper lip rather than a set of limbs clearly is a derived feature. It is the lack of limbs on the first postantennal segment of insects and myriapods that is a noteworthy and significant apomorphy, not the mere presence of a limb on that segment as occurs in crustaceomorphs, cheliceriforms, and many fossil groups such as trilobites. Crustacea are generally said to have a five-segment head. However, many crustaceomorph groups include at least one pair of maxillipeds and the associated “thoracic” somite into the head, and we thus speak of a cephalothorax. Most of the time, it is clear that these maxillipeds are obviously modified anterior thoracic limbs. Development in the many crustaceomorph groups that have maxillipeds allows us to document successive stages wherein the maxillipeds become specialized and their associated somites through successive molts become incorporated into the cephalon during ontogeny. However, at least one group of crustaceomorphs, the remipedes, does not exhibit such a transition. Koenemann et al. (2007, 2009) observed no biramous precursor state to the uniramous maxilliped in the earliest larval stages—the remipede maxilliped and its segment are part of the head in the earliest recognized ontogenetic stages. Consequently, we could say that the remipedes, for all intents and purposes, have a six-segment head (Koenemann et al. 2009). In summary, this first part of the diagnosis of Crustacea (head of five somites, each bearing a set of appendages consisting of two pairs of antennae, a pair of mandibles, and two pairs of maxillae) is not informative. “Body consisting of three regions: head, thorax, and ‘abdomen’” Body tagmosis is often an important component of defining an arthropod body plan. For example, as noted above, among the chelicerates a discrete head is lacking because the anterior segments associated with feeding and sensation are fused with the segments bearing the walking limbs to form a solid unit, the prosoma, a very distinctive feature.

11

12

Functional Morphology and Diversity The possession of a head, thorax, and abdomen is certainly distinctive, but it is also a feature shared with insects. Hence, while we might appear to have, with tagmosis, another argument for seeking some kind of relationship between crustaceomorphs and insects, that is, within a monophylum Pancrustacea or Tetraconata (Wheeler et al. 2004, Giribet et al. 2005), we do not have an effective component for a definition that seeks to uniquely define Crustacea. Furthermore, crustaceomorphs themselves vary considerably in this regard, as we will pursue in more detail below. The number of thoracic segments can be characteristic, but only for individual crustaceomorphs and not for Crustacea as a whole. Remipedes have a long, homonomously developed trunk with no differentiation between anterior and posterior sectors. Mystacocarids have five and branchiurans have four thoracomeres. Large-bodied branchiopods often have 11 or 12 thoracomeres. Many maxillopodans and the malacostracans have seven or eight thoracomeres: maxillopodans sensu stricto have seven, while cephalocarids and malacostracans have eight. Moreover, the possession of an abdomen is not a uniting feature. This variability is true not only regarding external, gross anatomical features such as total numbers of segments and those with and without paired limbs on the segments, but also for the underlying expression of Hox (homeobox) family genes as well (Fig. 1.4). In connection with the latter, Abzhanov and Kaufman (2004) and Schram and Koenemann (2004a) surveyed the available information concerning Hox gene expression in crustaceans. There are two fundamentally different types of posterior tagmata: the abdomen, a region without expression of the abd-A (abdominal A) Hox gene; and the pleon, a region with the expression of abd-A . Species with the latter type, the malacostracans, possess appendages on the segments and also display a well-differentiated central nervous system in that body region, whereas species with the former type, which lack abd-A expression in that body region (branchiopods and maxillopodans), lack appendages on these segments and do not have a well-differentiated central nervous system in these segments. It is for this reason that Schram and Koenemann (2004a) concluded that the old term pleon, as applied to the posterior region of the trunk of malacostracans, is not just an equal and interchangeable alternative for the term abdomen; the use of pleon as a descriptor is an absolute necessity. Hox gene expression indicates that the pleon of malacostracans and the abdomen of other crustaceans exhibit fundamentally different developmental pathways. Admittedly, the amount of available data is limited. As is the case with developmental work, researchers focus on the study and manipulation of model organisms. Among malacostracans, Porcellio scaber and Procambarus clarkii provided the model systems of preference for studies of Hox patterning, and among entomostracans, Artemia franciscana and Mesocyclops edax have served as the models, and the latter has been only incompletely investigated. The determination of Hox gene expression in diverse arthropods was a leading line of research in arthropod evo-devo studies in the late 1990s and early 2000s, but such investigations have waned, at least for now. In light of the above phylogenetic usefulness of this line of research, we should look forward to more animals being investigated in this regard. Nevertheless, this part of the definition of crustaceans (body consisting of three regions: head, thorax, and “abdomen”) is not a particularly informative statement. “Trunk appendages primitively multiramous” This descriptor is also not very informative. The presence of bi- and/or multiramous limbs is widely accepted to be a primitive condition in Arthropoda; most authorities would concede that uniramy is derived. However, here, too, the devil is in the details. Schram and Koenemann (2001) reviewed the information available concerning early development of crustacean limbs, and Williams (see chapter 3) delves into this subject more deeply.

Crustacean Biodiversity and Disparity of Body Plans lap

pb

z2

Antp

zen

Dfd

Scr

ftz

Antp

Anostraca (Branchiopoda)

Ubx abd-A

lab pb Dfd Scr

Abd-B

Porcellio scaber (Malacostraca)

Antp Ubx Abd-B

Scr Antp

abd-A

Abd-B

Procambarus clarkii (Malacostraca)

Ubx Abd-B

Ubx abd-A

Ubx abd-A abd-B

abd-A

Abd-B

Abd-B

Mesocyclops edax (Maxillopoda)

Fig. 1.4. Hox gene expression pattern for various crustaceans. Shaded areas denote thorax or thorax/pleon. Note the different patterns from an abdomen (no Hox) and a pleon (with abd-A). Modified from Schram and Koenemann (2004a).

We can summarize here, nevertheless, a few basic patterns of limb development. One, in which the proximal pedestal of the limb carries distally a tubular, segmented telopod, is sometimes referred to as the Drosophila model because it was first recognized and studied in detail using the fruit f ly Drosophila (Cohen 1990). It is the most common pattern of limb development seen in all biramous crustacean limbs that have been examined, particularly using Mysidopsis bahia (Panganiban et al. 1995). The limb anlage becomes forked, leading

13

14

Functional Morphology and Diversity eventually to the exopodal and endopodal rami. The gene distalless (dll) is expressed at the tips of the developing rami. A rather different pattern, however, prevails in Branchiopoda, often referred to as the Artemia model and documented with studies on Artemia and Triops (Williams and Mü ller 1996, Williams 1998). Rather than a uni- or biramous limb anlage, limb development begins with a mediolaterally directed ridge upon which eight lobes subsequently appear. The expression of dll occurs in varying patterns on these eight lobes, which proceed to form the unarticulated, leaf like limb, or corm, characteristic of the branchiopods. Similar gross anatomical sequences of limb development (though without the related gene expression patterns) have been documented for Cyclestheria (Olesen 1999) and the cladocerans (Olesen 1998). Hence, the multiramous limb of branchiopods has a fundamentally different mode of development from that seen in the crustaceans bearing biramous or uniramous limbs. Thus, this part of the definition of Crustacea (trunk appendages primitively multiramous) is not an informative statement. The statement equates all crustaceomorph limbs and ignores widely divergent, perhaps incompatible, modes of development. “Development consisting of a series of discrete larval and/or juvenile stages, initiated by a stage termed a nauplius” In examining this characteristic sequence, we possibly come upon firmer ground in seeking a unique set of features to define Crustacea. Many living groups of arthropods exhibit epimorphic development. The animals essentially hatch with the complete set of segments characteristic of the adult; the individuals increase in size only with each molt. Other groups of arthropods (some of the myriapods), although they resemble the adults in general form, hatch with fewer segments than the adults and add segments with each molt. Some Crustacea do this; for example, peracarids brood their young, and some of these are expelled from the marsupium as little “juvenile” forms, called mancas , which eventually molt and add a segment to achieve the adult condition. Many crustaceomorphs, however, hatch as larvae, and these larvae not only possess fewer segments than the adults but also exhibit a distinctive larval form. Successive molts then not only add segments but also metamorphose the form. Does this constitute an apomorphy for Crustacea? Other arthropods have larvae. Extensive larval stages are known for the trilobites, and pycnogonids have a larva; many larvae, both nauplii and other intriguing forms, are known from the fossil record (see Mü ller and Walossek 1986). However, there are distinctive patterns of molting and metamorphosis that serve to absolutely unite some crustaceomorphs. Taxa within Cirripedia are clearly united by the presence of a distinctive nauplius with frontolateral horns and a postnaupliar cypris larva in the life cycle. Branchiopods have a characteristic nauplius with a naupliar process on the antennae. Zoeae are diagnostic larvae of decapod malacostracans. The nauplius stage is often said to represent a phylotypic stage through which in theory all Crustacea passed in the course of the evolution of the group. We need to express some caution here—not all crustaceomorphs begin independent life as a nauplius larva, that is, exhibiting a larva characterized by possession of only three sets of limbs: the first and second antennae and the mandibles. There are crustaceomorphs that do (or did) not begin life as a nauplius but rather have as the initial stage a metanauplius, that is, a stage with more than just the three sets of naupliar limbs and/or more than the three naupliar segments. The issue is confused in the literature with the almost completely interchangeable use of the terms nauplius and metanauplius . This interchangeability implies that it is almost irrelevant as to what the basic structure of the first larva is—if it is tiny, possesses only a small number of limbs and segments, is given to swimming, and may or may not be filter feeding,

Crustacean Biodiversity and Disparity of Body Plans then it is a “nauplius.” We see here the differences between a structural and a functional definition. Which groups have an orthonauplius—a larva with only three pairs of appendages as seen in Branchiopoda, Maxillopoda, Remipedia, and euphausiacean and dendrobranchiate Malacostraca? Each of these orthonauplii bears a distinctive form. As noted above, branchiopod nauplii possess a naupliar process on the second antenna designed to facilitate feeding. Variations occur within the Maxillopoda. Among the most distinctive of nauplii, those of cirripedes bear anterolateral horns, frontal filaments anterior to the first antennae, and a long caudal process. There are four to six naupliar stages, depending on the group. Copepod nauplii exhibit a nauplius in almost its complete and pristine state, although the two orthonaupliar stages are nonfeeding because the gut is not developed until the metanaupliar phase. Ostracodes pass through a single nauplius stage, but the limbs are not completely developed, and in some species the early developmental stages (nauplius and the metanauplii) are retained within the mother’s shell until they are shed near the end of their development. The Cambrian fossil Rehbachiella had an orthonauplius. Finally, the free nauplii of the euphausiaceans (two) and dendrobranchiates (one) are very simple in form and do not feed, and even the succeeding metanauplii can be nonfeeding, also true of remipedes. Most of the other eumalacostracans pass through a clear egg-nauplius phase within the egg (Schram 1986). The diversity of naupliar form and function led Scholtz (2000) to suggest that we should distinguish between primary and secondary nauplii, that is, between nauplii that are indeed primitive and an original part of the life cycle, and nauplii that are secondarily reevolved. Scholtz believes that the primitive stage for malacostracans is the embryonized egg-nauplius and that the nonfeeding, free nauplii of euphausiaceans and dendrobranchiates actually evolved from ontogenetic sequences without a free nauplius. One consequence of Scholtz’s observations is that the nauplius larva would not be a phylotypic stage for all crustaceomorphs. Other groups of crustaceomorphs exhibit a variety of first stages in their development. Cephalocarida begin as a metanauplius, the first stage of which has five limbs and a variable number of limbless segments. Mystacocarida hatch as metanauplii with four sets of limbs and five additional limbless segments. The significance of these metanaupliar stages becomes evident when we consider the larval development in certain of the Cambrian Orsten microarthropods. The larval sequences for many of the Cambrian Orsten taxa are known; Bredocaris, Martinssonia, and Phosphatocopina all had four sets of limbs in the earliest phases, what Walossek has referred to as a “head larva” (Walossek and Mü ller 1990). Agnostus and the other trilobites in their earliest stages also bear four. There seems to be a basis for concluding that the naupliar stage, with its three sets of limbs, is derived from forms with four (possibly five) sets of limbs. Larvae are features of aquatic arthropods, but the nauplius larva is undoubtedly a derived form. Unfortunately, not all crustaceomorphs have a nauplius, which is perhaps a problem whose full implication remains to be determined; some groups may have lost it, but other groups probably never had it. At the beginning of this section I asked the question, What is Crustacea? It appears that we cannot use an unambiguous set of apomorphic descriptors to diagnose a monophyletic Crustacea. Developmental patterns and the nauplius larva appear to offer the best chance of doing so. However, since we have crustaceomorphs that do not exhibit the naupliar stage, we might conclude that the nauplius has evolved independently several times in the evolution of crustaceomorphs or has been lost several times; otherwise, if one demands that the nauplius be treated as diagnostic, Crustacea is not a monophyletic group. It would appear from the above discussion that we must conclude that crustaceomorphs are whatever is left over among the arthropods after we have assigned everything else to other clearly defined monophyletic groups.

15

16

Functional Morphology and Diversity

WHAT ARE THE CRUSTACEOMORPH BODY PLANS THAT MIGHT BE MONOPHYLETIC? We now have a conundrum. If we cannot define a monophyletic Crustacea with a single, consistent set of derived characters, can we perhaps diagnose smaller monophyletic groups within the current array of crustaceomorphs? I do believe that there are groups within this assemblage that are monophyletic (Schram and Koenemann 2004a). Short-Bodied Forms (Oligo-Crustacea) Branchiura Two groups of short-bodied forms at first glance would not appear to be at all alike (Fig. 1.5) but share a similarity regarding gonopore location. The living branchiurans are parasites of fish with highly modified mouthparts, but their gonopores open on the fourth thoracic somite. While there appears to be an abdomen, it is not differentiated into segments and is little more than a single or bilobed sac (Fig. 1.5A–C). Of special note is Pentastomida, the sister group of the branchiurans. Comparative sperm ultrastructure (Wingstrand 1972) and molecular sequence studies (Abele et al. 1989) revealed a close link of Branchiura with Pentastomida, odd wormlike parasites of the respiratory system in higher vertebrates (Fig. 1.5E). This pairing of branchiurans and pentastomids might appear peculiar, but it is a group of great age; pentastomid fossils exist from the early Paleozoic Orsten faunas (Walossek and Mü ller 1994, Walossek et al. 1994). The several species of Cambrian/Ordovician pentastomids (Fig. 1.5D) can convincingly be compared to living pentastomids (see Walossek and Mü ller 1994, their fig. 21), although the fossils have trunk limbs but lack the proboscis bearing the mouth. Walossek and colleagues interpret these fossils as parasites, but there is no direct evidence of this. These fossils could have been ordinary free-living members of the infauna. Nevertheless, what the fossils do show without any debate is that, in combination with the sperm and sequence data above, the ancestry of branchiurans is very ancient. Mystacocarida In contrast to the branchiurans, the mystacocarids are microscopic members of the beach meiofauna, almost wormlike in form, with a well-developed set of mouthparts, including maxillipeds, but with four pairs of rudimentary thoracic limbs (Fig. 1.5F). The gonopores are located on the fourth thoracic somite. Mystacocarids, too, may be of great age because, in some respects, they are not unlike Skaracarida, the Cambrian fossil group from the Orsten of Sweden (Mü ller and Walossek 1985) (Fig. 1.5G). Schram and Koenemann (2004a), using morphologic analysis tempered by Hox gene expression, found mystacocarids and branchiurans to be sister taxa. The results of molecular studies for both of these groups are confusing because long-branch attraction has been a persistent problem in these analyses; for example, Spears and Abele (1997) encountered this phenomenon when their results placed mystacocarids, remipedes, and cephalocarids together and in some proximity to chelicerates (a strange array), and the branchiurans emerged in a clade with podocopan ostracodes. Giribet et al. (2005) increased both the number of taxa sampled and genes sequenced but obtained a confusing collection of results depending on variant runs of taxa sampled (with and without fossils): mystacocarids and branchiurans sometimes appear alongside copepods and ostracodes; under other circumstances, branchiurans emerge elsewhere. Although the taxon sampling of Giribet et al. (2005) is impressive for all arthropods (and especially for hexapods), it is not particularly broad within crustaceomorphs. More recently, Regier et al. (2008) using nuclear

Crustacean Biodiversity and Disparity of Body Plans

A C B

D m

E

g

t

F

G

mp

Fig. 1.5. Body types of “short-bodied” crustaceans. (A–C) Diverse types of Branchiura (from Schram 1986). (A) Argulus. Note the highly modified mouthparts for attachment. (B) Dipteropeltis, with highly reduced body and winglike carapace. (C) Chonopeltis, displaying weak trunk and limb segmentation. (D and E) Diagrammatic Pentastomida (modified from Walossek and Mü ller 1994, their fig. 21). (D) Diagram similar to the Cambrian genus Heymonsicambria. Note the reduced trunk limbs (t). (E) Diagram of a generalized living pentastomid. m, mouth; g, gonopore; light gray, anterior trunk; dark gray, posterior trunk or abdomen. (F) Derocheilocaris, a mystacocarid. The arrow indicates approximate location of gonopore on fourth trunk limb. (G) Skara minuta Mü ller and Walossek, 1985, a Cambrian fossil crustacean that might represent a mystacocarid stem form. mp, maxilliped.

17

18

Functional Morphology and Diversity A

Malacostraca Maxillopoda Branchiopoda Cephalocarida Remipedia

B Hexapoda Xenocarida

Remipedia Leptocarida Malacostraca Maxillopoda Sensus stricto

Branchiopoda Brachiura Mandibulata

Oligostraca

Mystacocarida Ostracoda

Arthropoda

Myriapoda Chelicerata + Pycnogonida Onychophora Tardigrada

Fig. 1.6. (A) A classic understanding of crustacean phylogenetic relationships, based on morphology. (B) A summary version of one of the more recent molecular phylogenies. Modified after Regier et al. (2010).

protein coding genes also found branchiurans as a sister taxon to podocopan ostracodes. While the breadth of their molecular sample was impressive, the taxon sample was again selective; for example, no mystacocarid was included. To remedy the situation, Regier et al. (2010) have expanded the taxon base and increased the number of genes sequenced; their results identified a clear clade with Mystacocarida, Branchiura, and Pentastomida within a group they termed “Oligostraca” (Fig. 1.6). The analysis by Koenemann et al. (2010) also placed these short-bodied groups together these short-bodied groups. Oddly, these clades also contained Ostracoda (see below). The shortness of the body in these orders imposes definite constraints. The lack of an elaborated abdomen in branchiurans and pentastomids undoubtedly limits their ability to move around. One could speculate whether this lack was a factor in both groups adapting parasitic lifestyles. So, too, with mystacocarids: the lack of a well-developed abdomen could have constrained adapting a vermiform, interstitial existence where abilities to swim or otherwise move around are minimized. Branchiopoda Living Branchiopoda This large and fascinating group is almost exclusively restricted to freshwater, with a few exceptional cladocerans that are marine. One is tempted to speculate that they might have been

Crustacean Biodiversity and Disparity of Body Plans marine to begin with and then shifted to fresh waters. However, the evidence for that is not robust, and even most of the fossils, such as they are (mostly conchostracans), are preserved in freshwater to brackish water situations. The branchiopods do exhibit a distinctive set of features that outline a body plan for the group. We have noted above in passing the distinctive mode of limb formation—from horizontal ridges that subsequently become multilobed, rather than uni- or biramous limb bud anlagen—and the nauplius larva. Consideration of the large-bodied branchiopods adds some further depth to our knowledge of the branchiopod bauplan, of which Anostraca serves as a model. Traditionally (see Calman 1909, Schram 1986), anostracans were conceived as having an 11-segment thorax and a 9-segment abdomen. The first two legless, abdominal segments formed a fused genital complex. However, Hox gene expression studies reveal that the genes Antennapedia (Antp), Ultrabithorax (Ubx), and abd-A all are expressed in the thorax of Artemia, with a residual expression of Antp in the genital segments (Fig. 1.4). Abdominal B (abd-B), a marker for “end of thorax” and the genital segments, occurs in the genital segments (Abzhanov and Kaufman 2004, Schram and Koenemann 2004a). Hence, the genital segments are better considered thoracic, rather than abdominal, with the gonopores being carried on the twelfth segment of the thorax. The abdominal segments posterior to the genital complex do not exhibit Hox gene expression. Notostraca and the conchostracans have the gonopores opening on the eleventh or between the eleventh and twelfth trunk segments. Notostraca (Fig. 1.7A,B) carry a welldeveloped pair of limbs on each of the “thoracic” segments (the first two being somewhat modified from that seen on the others), but posterior to this region the limbs become increasingly smaller as one moves posterior in the sequence of somites, and there is little correspondence between the number of limbs and the segment boundaries—there are many more limbs than apparent segments. Regrettably, as yet no Hox gene expression studies have been performed on notostracans. The conchostracans are now generally divided into three monophyletic groups: Laevicaudata (Fig. 1.7C), Spinicaudata (Fig. 1.7D), and Cyclestherida (Fig. 1.7E) (Martin and Davis 2001), but they, too, appear to carry the gonopore in a position similar to that of Notostraca. The first of these, the laevicaudatans, or Lynceidae, have fewer trunk segments than the other two, but at least the female gonopore opens near the base of the eleventh, or penultimate, appendage (Linder 1945). The location of the male pore still must be confirmed (Martin et al. 1986). The other conchostracan groups have many more trunk segments, up to 32, with no differentiation between segments and limbs posterior to the genital openings, which are said to occur on the eleventh somite. Thus, most Branchiopoda feature thin, foliaceous, unjointed limbs, with trunks divided into an anterior section with well-developed limbs and good Hox expression, and with gonopores at or near the eleventh or twelfth trunk somite. The four groups of Cladocera (Fig. 1.7F–I), while they are clearly branchiopods, exhibit extreme forms of body reduction, or oligomery. All these branchiopods bear either a well-developed carapace or a derivative thereof. Only the anostracans (Fig. 1.7J) lack a carapace, and most authorities place the fairy shrimp as a sister group to all other branchiopods (Richter et al. 2007). The restriction of branchiopods to freshwater habitats entices one to wonder why these groups have become so limited. The unique mode of limb development perhaps precludes the development of anything other than thin, foliaceous, cormlike appendages. This in turn might have engendered an overall body habitus that lacks well-sclerotized and/or calcified body somites. Under these constraints, freshwater habitats, especially transient ones, provide satisfactory refugia.

19

20

Functional Morphology and Diversity

A F

B

C G

H

D

I E

J

Fig. 1.7. Body types of Branchiopoda. (A and B) Lepidurus arcticus, Notostraca (after Sars 1896): with carapace removed (A) and with carapace intact (B). (C–E) Various conchostracans. (C) Lynceus gracilicornis, Laevicaudata (modified from Martin et al. 1986). (D) Limnadia lenticularis, Spinicaudata (after Sars 1896). (E) Cyclestheria hislopi, Cyclestherida (after Sars 1887). (F–I) Various infraorders of Cladocera (after Lilljeborg 1901, Birge 1918). (F) Sida crystalline, Ctenopoda. (G) Bosmina longispina, Anomopoda. (H) Podon intermedius, Onychopoda. (I) Leptodora kindtii, Hoplopoda. (J) Branchinecta lindahli, Anostraca (after Lynch 1964).

Fossil Stem-Branchiopods All authorities accept crown group Branchiopoda as a monophyletic group, based on the distinctive nauplius larva and the form and ontogeny of the trunk limbs. However, the branchiopods are also noteworthy in that a number of fossil forms are known that either occupy a stem position to the branchiopod clade or in some instances actually stand within the group.

Crustacean Biodiversity and Disparity of Body Plans

A B

D

C

F ?

E

Fig. 1.8. Fossil species that might have some relationship to Branchiopoda, either near the base of that group or as stem forms. Arrows indicate the twelfth thoracomere, the segment at or just posterior to the end of the thoracic limb series, which might bear the gonopores. (A) Lepidocaris rhyniensis, Devonian (from Scourfield 1926). (B) Rehbachiella kinnekulensis, Upper Cambrian (from Walossek 1993). (C–E) various Cambrian waptiids. (C) Chuandianella ovata (from Chen and Zhou 1997). (D) Pauloterminus spinodorsalis (from Taylor 2002). (E) Waptia fieldensis (from Briggs et al. 1994). (F) Castracollis wilsonae (from Fayers and Trewin 2003).

The lipostracan Lepidocaris rhyniensis, a unique Devonian fossil (Fig. 1.8A), is preserved in great detail within nodules of chert. The thoracic limbs are both multiramous (thoracopods 1 and 2) and biramous (thoracopods 3–5). However, the presence of fully developed maxillae (Schram 1986), rather than the vestigial form characteristic of true branchiopods, indicates possibly only a sister-group relationship with crown Branchiopoda.

21

22

Functional Morphology and Diversity Rehbachiella (Fig. 1.8B) has figured prominently in discussions of branchiopod origins (Walossek 1993). However, Schram and Koenemann (2001) took exception with this and concluded that Rehbachiella, while possibly a stem form, was not a branchiopod sensu stricto since not only do they possess biramous thoracopods, but also these limbs arise from biramous anlagen and not the multilobed ridge of true branchiopods. Hence, I believe Rehbachiella, at best, is a stem form. Another fossil group from the Cambrian could be relevant to understanding stem evolution of branchiopods, the waptiids (Fig. 1.8C–E). The genus Waptia from the Burgess Shale of Canada is probably the most famous. However, several genera are known (Briggs et al. 1994, Chen, and Zhou 1997, Taylor 2002) and all appear to have a subdivided thorax with apparently four anterior telopodous limbs and six posterior foliaceous limbs. The gonopores have yet to be identified for waptiids, but I would venture a guess that they probably occurred on the eleventh trunk segment. Although the recent large-scale molecular analyses of Giribet et al. (2005) and Wheeler et al. (2004) typically find Hexapoda as a sister group to all crustaceomorphs, there is an alternative hypothesis. Schram and Koenemann (2004b), VanHook and Patel (2008), Lartillot and Philippe (2008), and Dell’Ampio et al. (2009) obtained trees with insects and Branchiopoda as sister groups. These results were based on developmental gene expression patterns and molecule sequences, and these trees serve to propose alternative hypotheses concerning branchiopod relationships. Finally, there are fossils such as Castracollis wilsonae (Fig. 1.8F) that exhibit body plans that are complex but nevertheless place them within branchiopods, in this case 11 large, foliaceous, cormlike limbs on the anterior thorax followed by another series of similar limbs but much reduced in size (Fayers and Trewin 2003). Castracollis might or might not have had a carapace. Eucrustacea What remains of the crustaceomorph taxa after clades of short-bodied and branchiopodan types are isolated is a confederation of diverse forms: Cephalocarida, Malacostraca, Remipedia, and Maxillopoda. When viewed as a whole, these taxa are divergent in terms of both habitus and habitat; nevertheless, all these groups bear gonopores on the sixth through eighth thoracic somites. There are a couple of interesting exceptions to this rule, which I note below. Cephalocarida This group of hermaphrodites is small both in size and in species numbers. It has a thorax of eight segments and a limbless abdomen of 12 segments (Fig. 1.9A). The form of the maxillae is very similar to that seen for the thoracopods. The gonopores are located on the sixth thoracomere. Nothing is known of Hox gene expression in cephalocarids. The body plan of cephalocarids might exhibit the results of the same sorts of constraints we saw above with mystacocarids. In this case, the elongate, limbless abdomen with extended terminal caudal rami at best probably functions like the tail on a kite: a stabilizer to minimize drag and the effect of turbulence as the animals swim. The long series of thoracic limbs developed as swimming paddles provide more locomotory abilities than that seen in the tiny thoracopods of mystacocarids, but nonetheless, competition from larger and more mobile forms probably forced the cephalocarids to retreat to f locculent bottom sediments in order to make a living.

Crustacean Biodiversity and Disparity of Body Plans

B

A

C E

F

D

G

Fig. 1.9. The four major groups among the core bauplan of Crustacea, with gonopore-bearing segments indicated by arrows. (A) A cephalocarid, Hutchinsoniella macracantha, a hermaphrodite with pores on the sixth thoracomere (modified from Schram 1986). (B) A hoplocarid Malacostraca, a male Squilla mantis. The male pore (long arrow) would be on the eighth thoracomere; the female pore (short arrow) would be on the sixth thoracomere (modified from Calman 1909). (C and D) Two types of eumalacostracan Malacostraca: with (C) and without (D) a carapace. (C) A euphausiid, Meganyctiphanes norvegica . The male pore is on the eighth thoracomere (long arrow); the female pore, on sixth thoracomere (sort arrow) (modified from Mauchline and Fisher 1969). (D) The syncarid Anaspides tasmaniae. The male pore is on the eighth thoracomere (long arrow); the female pore, on sixth thoracomere (short arrow) (modified from Schminke 1978). (E and F) Two types of remipede, hermaphrodites with the male pore on the eighth thoracomere (long arrow) and the female pore on the fifteenth thoracomere (short arrow). (E) Medium-length body, Speleonectes gironensis (modified from Yager 1994). (F) Short-length body, Micropacter yagerae (modified from Emerson and Schram 1991). (G) A typical maxillopodan, Calanus finmarchicus. Male and female pores open on seventh thoracomere (modified from Calman 1909).

23

24

Functional Morphology and Diversity Malacostraca This most variable of crustacean groups nevertheless has a fundamentally uniform structural plan. The trunk is divided into an anterior thorax of eight segments and a posterior pleon of six or seven segments, sometimes fewer. All trunk somites generally bear appendages, but noteworthy variations can occur, such as one or more thoracopods serving as maxillipeds or posterior thoracopods and/or pleopods being greatly reduced or absent. Hox genes are expressed throughout the body (Abzhanov and Kaufman 2004, Schram and Koenemann 2004a) with Ubx characteristic of the thorax and abd-A of the pleon (Fig. 1.4). The female gonopores occur in association with the sixth thoracic segment, while the male pores are on the eighth. The malacostracans are typically said to contain three groups: the small nectobenthic leptostracans (not illustrated), the obligate carnivorous hoplocaridans (Fig. 1.9B), and the extremely diverse caridoid eumalacostracans that have forms both with a carapace (Fig. 1.9C) and without (Fig. 1.9D). The diversity of this group is examined in greater detail in other chapters in this volume. We might say that the great versatility imparted by the malacostracan body plan is responsible for its success. The long series of limbs, extending through both the thorax and pleon, allows a great degree of variation and specialization that undoubtedly has allowed the group to radiate to the extent it has, with great numbers of species and remarkable variations in structure. Maxillopoda With the problematic Mystacocarida and Branchiura removed from the maxillopodans, where textbooks often place them, there remains a core set of taxa that appear to conform to a single body plan. The old formula of 5–6-5 or the newer viewpoint of 5–7-4—five cephalic, seven thoracic, and four abdominal somites (see Newman 1987)—has great consistency throughout the group. The old interpretation was of a thorax with six limb-bearing segments and an abdomen of five segments always lacking limbs. An alternative interpretation of the gonopore-bearing segment as actually part of the thorax (Newman 1987) leaves only four abdominal somites; this interpretation makes more sense not only in terms of what we can see in other groups, for example, the free gonopore-bearing segment of the anostracans that occurs just posterior to a set of trunk limbs mentioned above, but also in terms of what limited information we have concerning Hox gene expression, with Ubx and Abd-A expression in the thorax and abd-B in the genital segment (Averof and Patel 1997). Copepoda (Fig. 1.9G) most clearly present the pattern of 5–7-4. The gonopores of both sexes occur on the seventh thoracomere. Thecostraca conform to the basic maxillopodan pattern with some variations. Ascothoracica exhibit 5–7-4. Facetotecta appear to manifest 5–7-3, based on the anatomy of the Y-cypris. Cirripedia exhibit 5–7-0, considering the cypris larva as a stand-in model for the highly derived adults. While male gonopores in the cirripedes appear on the seventh thoracic segment, the female pore has shifted forward onto the first thoracic segment. Furthermore, the cirripedes lack an abdomen and coincidently also lack any expression of Abd-A (Mouchel-Vielh et al. 1998). The parasitic Tantulocarida present problems since these microscopic forms have an extremely aberrant life cycle. However, recent advances in elucidating that life cycle (Huys et al. 1993) allow us to conclude that the tantulocarids express a 5–7-2 pattern, with the male gonopore appearing on the seventh thoracic somite and the single median female pore occurring on the first. This latter feature clearly unites tantulocarids and thecostracans as sister groups.

Crustacean Biodiversity and Disparity of Body Plans The constraints exerted by a limbless abdomen on lifestyle may explain much of what we see in maxillopodan evolution. The maxillopodans certainly thrive under unusual conditions. Parasitism is widespread in the group, especially among thecostracans, and those thecostracans that are not parasites have lost the abdomen altogether and settled (literally) into the completely sedentary, highly aberrant body plan seen in the barnacles. Only the copepods possess the kind of biodiversity and habitat variability we associate with “successful” groups. Even so, the small sizes of copepods could be related to the limits engendered by an abdomen lacking limbs. Ostracoda These animals remain the most vexing of arthropods to place phylogenetically and, if molecular sequences are to be believed, may not be a monophyletic group. Their extreme reduction of body plan (oligomery), complete enclosure within a calcareous shell, and specializations directed at life carried on at a microscale have hindered attempts to link them to other crustaceomorphs. There are contentious debates about homologies within Ostracoda (Horne et al. 2005), and ostracodes do not appear to share obvious apomorphies with other crustaceomorphs. Most textbooks and reference books consign ostracodes to the maxillopodans (see Schram 1986), but that is more of a default placement. K. Martens (personal communication, 2004) and R.A. Jenner (personal communication, 2009) expressed an informal view of at least some researchers that Ostracoda might not be a monophyletic group. This possibility obtains some support from molecular data that sometimes finds Podocopa and Myodocopa in different parts of cladograms (see Spears and Abele 1997, Regier et al. 2008, 2010, Koenemann et al. 2010). However, most of these analyses have a very limited taxon sample with sequences from only a handful of ostracode species. There is much variation in form in ostracode limbs, but there is a consensus at least that both the myodocopes and the podocopes are themselves monophyletic (Horne et al. 2005). However, debates about the number of somites in each group are not settled. At first glance, one perceives that only very few thoracic segments bear limbs, but Schulz (1976) presented some evidence that indicates Cytherella pori, a podocopan, might have 11 trunk somites (a 5–7-4 pattern) and that the penis appears to be associated with the sixth or seventh of these segments (see Schram 1986, their fig. 33–1C). Tsukagoshi and Parker (2000) confirmed this in other species of podocopans. No similar information is available yet for Myodocopa. From this, it might appear that podocopan ostracodes are possibly maxillopodans. Some authorities classified Ostracoda as a subclass of Maxillopoda (Schram 1986), but others maintain them as an independent class (Martin and Davis 2001). Most recently, Koenemann et al. (2010) and Regier et al. (2010), on the basis of molecular sequences, obtained both podocopan and myodocopan ostracodes as a sister group to a clade of mystacocarids, branchiurans, and pentastomids. This arrangement would then unite all the short-bodied “oligostracans” into a single clade near the base of the crustaceomorph tree (Fig. 1.6). However, there existed Paleozoic, especially Cambrian, taxa that may have some bearing on eventually determining ostracode affinities. Several such groups are under active study, such as the bradoriids, Phosphatocopida (Maas et al. 2003), and perhaps even the thylacocephalans. These groups will eventually have to be integrated into any classification of the crustaceomorphs, and undoubtedly they will prove very interesting in this regard. Remipedia This most recently discovered group of crustaceomorphs is noteworthy for several reasons. The trunk is not differentiated into a thorax and abdomen/pleon. If we consider the

25

26

Functional Morphology and Diversity maxilliped-bearing segment as a modified trunk somite (even though it is completely merged into the cephalon), then the female gonopore occurs on the eighth postmaxillary segment, and the male gonopore, on the fifteenth. However, as noted above (Koenemann et al. 2007, 2009), the distinctive maxillipeds, virtually identical in general form to the maxillae, display no developmental evidence that this limb is modified from a thoracopod format. In addition, the number of trunk segments is not fixed, either within or between species (Koenemann et al. 2006), with many long-bodied forms recognized (Fig. 1.9E)—although there appears to be at least a lower limit of 16 trunk segments in the adults (Fig. 1.9F). The significance of all this variability remains to be explored. The remipede body plan ensured that these animals are excellent swimmers, on a par with anything seen among the malacostracans. Even so, their habitat restrictions are quite profound; they prefer anchialine cave habitats in low-oxygen conditions.

CLASSIF ICATION The above review indicates there have to be changes in our concepts of crustaceomorph classification, but this is not the place to present any new or radical higher taxonomy. In principle, we want our taxonomies to reflect phylogeny, but that is not always possible. There is much conflicting evidence from molecular analyses, which along with morphological data often suffers from limited taxon sampling, and the latter often ignores or minimizes input from fossils. We still need to more effectively integrate data from gross morphology, molecular sequencing, and paleontology into a coherent whole. Nevertheless, we should extend some effort to recognize the monophyletic groups about which we are certain (Fig. 1.10); there are patterns that should be acknowledged. To these ends, we can make good use of the concept of the plesion, a particular taxon that does not fit well into another category and that eventually might be assigned to its own higher category. I believe that, in this instance, we should begin to think of the infraphyla below as monophyletic groups on a par with other well-established arthropod monophyla such as Hexapoda, Chelicerata, Trilobita, and Pycnogonida. What fossils and where they will fall within or between these monophyletic groups will be explored elsewhere. The scheme is not complete in terms of all possible fossil plesions but does include most of those mentioned in the text above (Table 1.4).

WHAT MIGHT THE CRUSTACEOMORPH ANCESTOR HAVE LOOKED LIKE? At one time, there was a fair consensus as to what the ancestor of Crustacea might have looked like. Hessler and Newman (1975) devised an ancestor with a long, homonomously segmented body, each segment bearing a set of limbs not unlike a cephalocarid, for which Newman preferred a form with a carapace, and Hessler one without (Fig. 1.11A). Cisne (1982) believed that crustaceans arose from a trilobite-like ancestor. Schram (1982) concurred with Hessler and Newman (1975), although he would have preferred a somewhat more foliaceous limb, intermediate between cephalocarids and branchiopods (Fig. 1.11B). However, Schram’s 1982 paper had been written in 1978 (delayed due to a delay in the publication of the book in which it appeared), and in the intervening years the remipedes had come to light. By 1983, Schram had altered his views as to the form of an ancestor, which, while still in possession of a long homonomous body, was viewed as equipped with biramous, paddlelike limbs (Schram 1983). Schram positioned this biramous theory as an alternative hypothesis to the mixopodial theory of Hessler and Newman, and this then postulated a remipede-like alternative ancestor as opposed to a cephalocarid-like forebear.

Crustacean Biodiversity and Disparity of Body Plans A

4 Mystacocarida

B

6-8

Maxillopoda

Malacostraca

Remipedia

Cephalocarida C

12 Branchiopoda

Fig. 1.10. Major crustaceomorph body plans based on gonopore position. (A) Mystacocarida, an “oligostracan.” (B) “Eucrustaceans.” (C) Branchiopoda. From Schram and Koenemann (2004a).

The debate outlined above was based on morphology. Some information derived from molecular sequences now suggests that hexapods could factor into this mix. One fossil that might have facilitated a visual understanding of how this transition might have occurred is Wingertschellicus backesi Briggs and Bartels, 2001 (= Devonohexapodus bocksbergensis Haas et al., 2003). A recent reexamination of all available fossils of this species from the famous Devonian Hunsr ück Shale (K ü hl and Rust 2009) synonymized the two names, but the original reconstruction of Haas et al. (2003) presented a strange chimera—it appears to have a dragonf ly anterior end and a very long myriapodous posterior end (Fig. 1.11C). Although the interpretation of Haas et al. (2003) of D. bocksbergensis offered a head (of possibly four segments) and a short three-segment thorax followed by a long abdomen, the new interpretation presents a six- or seven-segment head, with the posteriormost three pairs of cephalic limbs as long, possibly prehensile appendages and followed by a long trunk with biramous limbs. Neither Briggs and Bartels (2001) nor Kü hl and Rust (2009) offer a reconstruction of W. backesi, but the latter believe that this species is neither a stem hexapod nor within crown group Malacostraca. Of these I am not so sure, having once had the opportunity to examine D. bocksbergensis courtesy of Dieter Walossek. Even though what this Devonian species might represent remains uncertain, nevertheless, it does demonstrate that there is an abundance of long-bodied forms in the Paleozoic that may have significance for understanding the early evolution and possible origins of surviving groups of arthropods. Another source of information that is relevant for understanding crustacean ancestry is derived from the study of the Cambrian Orsten microfossils (a few of which were mentioned

27

28

Functional Morphology and Diversity Table 1.4. A classification of tetraconate arthropods with inclusion of fossil plesions. Subphylum: Tetraconata (= Crustaceomorpha = Pancrustacea) Infraphylum: Hexapoda Infraphylum: unnamed (short-bodied crustaceomorphs—“Oligostraca”) Class Branchiura Order: Arguloida Order: Pentastomida Class: Mystacocarida Plesion: Skaracarida Plesion: Ostracoda (one possible position; includes Myodocopa and Podocopa) Infraphylum: Branchiopoda Class: Phyllopoda (= Calmanostraca) Order: Laevicaudata Order: Notostraca Order: Spinicaudata Order: Cyclestherida Order: Cladocera Class: Sarsostraca Order: Anostraca Plesion: Lipostraca (= Lepidocaris) Plesion: Rehbachiellida (= Rehbachiella) Plesion: Waptiidae Infraphylum: Crustacea Class: Cephalocarida Class: Maxillopoda Subclass: Copepoda Subclass: Thecostraca Infraclass: Ascothoracica Infraclass: Cirripedia Infraclass: Facetotecta Infraclass: Tantulocarida Plesion: Ostracoda (one possible position; includes Myodocopa and Podocopa) Class: Malacostraca Subclass: Eumalacostraca Subclass: Hoplocarida Subclass: Phyllocarida Class: Remipedia Cephalocarida and Remipedia might constitute a single class, Xenocarida, based on molecular evidence. Ostracoda could occupy two possible positions: among oligostracans, based on molecular data, or within maxillopodans, based on some morphological data. Hoplocarida and Eumalacostraca (sensu stricto) could be arranged as a single subclass Eumalacostraca (sensu lato) with infraclasses Caridoida and Hoplocarida.

above), and these have raised the possibility of alternative hypotheses. Incompletely understood in the 1980s, the depth of knowledge about these animals is now astounding, extending as it does to even developmental stages for many of these species (see chapter 2). The full impact of these studies remains to be assessed within the larger framework of the anatomy of modern forms, molecular sequences, and gene expressions, but much of this work suggests a

Crustacean Biodiversity and Disparity of Body Plans

A

B

C

A'

Fig. 1.11. Crustaceomorph ancestors (see text for details). (A and A′) Without and with a carapace, according to Hessler and Newman (1975). (B) According to Schram (1982). (C) Devonohexapodus bocksbergensis (from Haas et al. 2003).

possible alternative hypothesis: a short-bodied ancestor rather than a long-bodied one. This merits consideration (Schram and Koenemann 2004a), but it is not possible or appropriate to examine here.

CONCLUSIONS It would appear that we are little closer to understanding the origin of crustaceomorphs than we were 30 years ago. While the larger assemblage of the crustaceomorphs (or pancrustaceans, or tetraconatans, if you prefer) might be in some way monophyletic, just how it can (or even if it can) be diagnosed with a single set of apomorphies is not clear at this point. There are, however, good monophyletic groups within this vast array that can be clearly defined. Furthermore, these body plans appear to be constrained regarding biodiversity, functional morphology, and habitats they can occupy. In addition, we have a growing array of fascinating fossil taxa scattered within and between these monophyla, but how these are related to the monophyletic groups for which they may serve as stem forms remains to be determined. But take heart! It is a time not to mourn the demise of the monophylum Crustacea but to embrace what will be a new world order and a better understanding of this whole branch of the crustaceomorph arthropods.

ACKNOWLEDGMENTS I am grateful for the invitation from Profs. Martin Thiel and Les Watling to write this chapter. I also thank Dr. Ronald Jenner (Natural History Museum, London) for reading a draft and offering some valuable comments and suggestions. Prof. Stefan Koenemann (Hannover) was involved in producing some of the earlier papers from which parts of this chapter developed. Any and all faults, however, are my own.

29

30

Functional Morphology and Diversity

REFERENCES Abele, L.G., W. Kim, and B.E. Felgenhauer. 1989. Molecular evidence for the inclusion of the phylum Pentastomida in the Crustacea. Molecular Biology and Evolution 6:685–691. Abzhanov, A., and T.C. Kaufman. 2004. Hox genes and tagmatization of the higher Crustacea (Malacostraca). Crustacean Issues 15:43–74. Averof, M., and N.H. Patel . 1997. Crustacean appendage evolution associated with changes in Hox gene expression. Nature 388:682–686. Birge, E.A. 1918. The water fleas (Cladocera). Pages 676–740 in H.B. Ward and G.C. Whipple, editors. Freshwater biology. Wiley, New York. Bouchet, P. 2006. The magnitude of marine biodiversity. Pages 33–62 in C.M. Duarte, editor. The exploration of marine biodiversity: Scientific and technical challenges. Fundación BBVA, Bilbao, Spain. Boxshall, G.A. 1983. Three new genera of misophrioid copepods, with special reference to Benthomisophria palliata . Philosophical Transactions of the Royal Society of London Series B 297:125–181. Boyan, G.S., J.L.D. Williams, S. Posser, and P. Bäunig. 2002. Morphological and molecular data argue for the labrum being non-apical, articulated, and the appendage of the intercalary segment in the locust. Arthropod Structure and Development 31:65–76. Briggs, D.E.G., and C. Bartels . 2001. New arthropods from the Lower Devonian Hunsr ück Slate (Lower Emsian, Rhennish Massif, Western Germany). Palaeontology 44:275–303. Briggs, D.E.G., D.H. Erwin, and F.J. Collier. 1994 . The fossils of the Burgess Shale. Smithsonian Institution Press, Washington, DC. Brusca, R.C., and G.J. Brusca . 1990. Invertebrates. Sinauer, Sunderland, MA. Brusca, R.C., and G.J. Brusca . 2003. Invertebrates, 2nd ed. Sinauer, Sunderland, MA. Calman, W.T. 1909. Crustacea. Part VII, 3rd fasc., in E.R. Lankester, editor. A treatise on zoology, Vol. 7. Adam and Charles Black, London. Chen, J.-Y., and G.-Q. Zhou. 1997. Biology of the Chengjiang fauna. Bulletin of the National Museum of Natural Sciences (Taiwan) 10:11–105. Cisne, J.L. 1982. Origin of Crustacea. Pages 65–92 in L.G. Abele, editor. Biology of the Crustacea, Vol. 1. Academic Press, New York . Cohen, S.M. 1990. Specifications of limb development in the Drosophila embryo by positional cues from segmentation genes. Nature 343:173–177. Dell’Ampio, E., N.U. Szucsich, A. Carapelli, F. Frati, G. Steiner, A. Steinacher, and G. Pass . 2009. Testing for misleading effects in the phylogenetic reconstruction of ancient lineages of hexapods: Influence of character dependence and character choice in analyses of 28S rRNA sequences. Zoologica Scripta 38:155–170. Emerson, M.J., and F.R. Schram . 1991. Remipedia. Part 2. Proceedings of the San Diego Society of Natural History 7:1–52. Fayers, S.R., and N.H. Trewin. 2003. A new crustacean from the Early Devonian Rhynie chert, Aberdeenshire, Scotland. Transactions of the Royal Society of Edinburgh, Earth and Environmental Science 93:355–382. Giribet, G., S. Richter, G.D. Edgecombe, and W. Wheeler. 2005 . The position of crustaceans within Arthropoda—evidence from nine molecular loci and morphology. Crustacean Issues 16:307–352. Groombridge, B., and M.D. Jenkins. 2000. Global diversity: Earth’s living resources in the 21st century. World Conservation Press, Cambridge. Grygier, M.J. 1981. Gorgonolaureus muzikae sp. nov. parasitic on a Hawaiian gorgonian, with special reference to its protandric hermaphroditism. Journal of Natural History 15:1019–1045. Haas, F., D. Waloszek , and R. Hartenberger. 2003. Devonohexapodus bocksbergensis, a new marine hexapod from the Lower Devonian Hunsr ück Slates, and the origin of Atelocerata and Hexapoda. Organic Diversity and Evolution 3:39–54. Hessler, R.R. 1969. A new species of Mystacocarida from Maine. Vie et Milieu 22:105–116. Hessler, R.R., and W.A. Newman . 1975 . A trilobitomorph origin for the Crustacea. Fossils and Strata 4:437–459.

Crustacean Biodiversity and Disparity of Body Plans Horne, D.J., I. Schön, R.J. Smith, and K. Martens . 2005 . What are the Ostracoda? A cladistic analysis of the extant superfamilies of the subclasses Myodocopa and Podocopa (Crustacea: Ostracoda). Crustacean Issues 16:249–273. Huys, R., G.A. Boxshall, and R.J. Lincoln. 1993. The tantulocaridan life cycle: The circle closed? Journal of Crustacean Biology 13:432–442. Koenemann, S., R.A. Jenner, M. Hoenemann, T. Stemme, and B.M. von Reumont. 2010. Arthropod phylogeny revisited, with a focus on crustacean relationships. Arthropod Structure and Development 39:88–110. Koenemann, S., J. Olesen, F. Alwes, T.M. Iliffe, M. Hoenemann, P. Ungerer, C. Wolff, and G. Schultz. 2009. The post-embryonic development of Remipedia (Crustacea)—additional results and new insights. Development Genes and Evolution 219:119–129. Koenemann, S., F.R. Schram, A. Bloechl, T.M. Iliffe, M. Hoenemann, and C. Held . 2007. Post-embryonic development of remipede crustaceans. Evolution and Development 9:117–121. Koenemann, S., F.R. Schram, and T.M. Iliffe. 2006. Trunk segmentation patterns in Remipedia. Crustaceana 79:607–631. Kü hl, G., and J. Rust . 2009. Devonohexapodis bocksbergensis is a synonym of Wingertshellicus backesi (Euarthropoda)—no evidence for marine hexapods in the Devonian Hunsr ück Sea. Organisms Diversity and Evolution 9:215–231. Lang , K. 1953. Apseudes hermaphroditicus, a hermaphroditic tanaid from Antarctica. Arkiv för Zoologie (2) 9:341–350. Lartillot, N., and H. Philippe 2008. Improvement of molecular phylogenetic inference and the phylogeny of Bilateria. Philosophical Transactions of the Royal Society of London Series B 363:1463–1472. Lilljeborg , W. 1901. Caldocera Suiciae. Nova Acta Regiae Societatis Scientarium Upsaliensus. Uppsala 3:1–701. Linder, F. 1945 . Affinities within the Branchiopoda, with some notes on some dubious fossils. Arkiv för Zoologie 37:1–28. Lynch, J.E. 1964 . Packa’s and Pearse’s species of Branchinecta: Analysis of nomenclatural involvement. American Midland Naturalist 71:466–488. Maas, A., D. Waloszek , and K. Mü ller. 2003. Morphology, ontogeny, and phylogeny of the Phosphatocopina (Crustacea) from the Upper Cambrian “Orsten” of Sweden. Fossils and Strata 49:1–238. Martin, M.F. 1932. On the morphology and classification of Argulus. Proceedings of the Zoological Society of London 1932:771–806. Martin, J.W., and G.E. Davis. 2001. An updated classification of the recent Crustacea. Science Series, Natural History Museum of Los Angeles County 39:1–124. Martin, J.W., and G.E. Davis. 2006. Historical trends in crustacean systematics. Crustaceana 79:1347–1368. Martin, J.W., B.E. Felgenhauer, and L.G. Abele. 1986. Redescription of the clam shrimp Lynceus gracilicornis (Packard) (Branchiopoda, Conchostraca, Lynceidae) from Florida, with notes on its biology. Zoologica Scripta 15:221–232. Mauchline, J., and L.R. Fisher. 1969. The Biology of Euphausiids. Advances in Marine Biology 7:1–454. May, R.M. 1988. How many species are there on Earth? Science 241:1441–1449. Meglitsch, P.A., and F.R. Schram. 1991. Invertebrate zoology, 3rd ed. Oxford University Press, New York . Minelli, A. 1993. Biological systematics. Chapman and Hall, London . Minelli, A., G. Fusco, and S. Sartori. 1991. Self-similarity in biological classifications. BioSystems 26:89–97. Mouchel-Vielh, E., C. Rigolot, J.-M. Gibert, and J.S. Deutsch . 1998. Molecules and the body plan: The Hox genes in cirripedes (Crustacea). Molecular Phylogenetics and Evolution 9:382–389. Mü ller, K.J., and D. Walossek. 1985 . Skaracarida, a new order of Crustacea from the Upper Cambrian of V ä stergotland, Sweden. Fossils and Strata 17:1–65. Mü ller, K.J., and D. Walossek. 1986. Arthropod larvae from the Upper Cambrian of Sweden. Transactions of the Royal Society of Edinburgh, Earth and Environmental Science 77:157–179. Mü ller, K.J., and D. Walossek. 1988. External morphology and larval development of the Upper Cambrian maxillopod Bredocaris admirabilis. Fossils and Strata 23:170.

31

32

Functional Morphology and Diversity Newman, W.A. 1987. Evolution of cirripedes and their major groups. Crustacean Issues 5:3–42. Olesen, J. 1998. A phylogenetic analysis of the Conchostraca and Cladocera (Crustacea, Branchiopoda, Diplostraca). Zoological Journal of the Linnean Society 122:491–536. Olesen, J. 1999. Larval and post-larval development of the branchiopod clam shrimp Cyclestheria hislopi (Crustacea, Branchiopoda, Conchostraca, Spinicaudata). Acta Zoologica 80:163–184. Panganiban, G., A. Sebring , L. Nagy, and S. Caroll . 1995 . The development of crustacean limbs and the evolution of arthropods. Science 270:1363–1366. Regier, J.C., J.W. Shultz , A.R.D. Ganley, A. Hussey, D. Shi, B. Ball, A. Zwick, J.E. Stajich, M.P. Cummings, J.W. Martin, and C.W. Cunningham . 2008. Resolving arthropod phylogeny: Exploring phylogenetic signal within 41 kb of protein-coding nuclear gene sequence. Systematic Biology 57:920–938. Regier, J.C., J.W. Shultz , A. Zwick, A. Hussey, B. Ball, R. Wetzer, J.W. Martin, and C.W. Cunningham. 2010. Arthropod relationships revealed by phylogenomic analysis of nuclear protein-coding sequences. Nature 463:1079–1083. Richter, S., J. Olesen, and W.C. Wheeler. 2007. Phylogeny of Branchiopoda (Crustacea) based on a combined analysis of morphological data and six molecular loci. Cladistic 23:1–36. Ruppert, E.E., and R.D. Barnes. 1994 . Invertebrate zoology, 6th ed. Saunders College Publications, Fort Worth, TX. Sars, G.O. 1887. On Cyclestheria hislopi (Baird), a new generic type of bivalved phyllopod raised from dried Australian mud. Christiania Videnskabernes Selskab Fordhandlinger 1:1–65. Sars, G.O. 1896. Phyllocarida and Phyllopoda. Fauna Norvegiae, Vol 1. Stock, Christiania, Aschehoug. Scholtz , G. 2000. Evolution of the nauplius stage in malacostracan crustaceans. Journal of Zoological Systematics and Evolution Research 38:175–187. Schminke, H.K. 1978. Die phylogenetische Stellung der Stygocarididae—unter besonderer Berucksichtigung morphologischer Ahnlichkeiten mit Larvenformen der Eucarida. Zeitschrift f ü r zoologische Systematik und Evolutionsfoschung 16:225–239 Schram, F.R. 1979. Crustacea. Pages 238–244 in R.W. Fairbridge and D. Jablonski, editors. The encyclopedia of paleontology. Dowden, Hutchinson, and Ross, Stroudsberg, PA . Schram, F.R. 1982. The fossil record and evolution of Crustacea. Pages 93–147 in L.G. Abele, editor. The biology of Crustacea, Vol. 1. Academic Press, New York . Schram, F.R. 1983. Remipedia and crustacean phylogeny. Crustacean Issues 1:23–28. Schram, F.R. 1986. Crustacea . Oxford University Press, New York . Schram, F.R. 2003. Our evolving understanding of biodiversity through history and its impact on the recognition of higher taxa in Metazoa. Pages 359–368 in A. Legakis, editor. The new panorama of animal evolution, Proceedings of the 18th International Congress of Zoology. Pensoft Publishers, Sofia, Bulgaria. Schram, F.R., and S. Koenemann . 2001. Developmental genetics and arthropod evolution: Part I, on legs. Evolution and Development 3:343–354 Schram, F.R., and S. Koenemann. 2004a . Developmental genetics and arthropod evolution: Part II, on body regions of crustaceans. Crustacean Issues 15:75–92. Schram, F.R., and S. Koenemann. 2004b. Are the crustaceans monophyletic? Pages 319–329 in J. Cracraft and M.J. Donaghue, editors. Assembling the tree of life. Oxford University Press, New York . Schram, F.R., J. Yager, and M.J. Emerson. 1986. Remipedia. Part 1. Systematics. San Diego Society of Natural History Memoires 15:1–60. Schulz , K. 1976. Das Chitinskelett der Podocopida (Ostracoda, Crustacea) und die Frage der Metamerie dieser Gruppe. PhD dissertation, University of Hamburg. Scourfield, D.J. 1926. On a new type of crustacean from the Old Red Sandstone—Lepidocaris rhyniensis. Philosophical Transactions of the Royal Society of London Series B 214:153–187. Spears, T., and L.G. Abele. 1997. Crustacean phylogeny inferred from 18S rDNA. Pages 169–186 in R.A. Fortey and R.A. Thomas, editors. Arthropod relationships. Chapman and Hall, London . Taylor, R.S. 2002. A new bivalved arthropod from the early Cambrian Sirius Passet fauna, North Greenland. Palaeontology 45:97–123. Tsukagoshi, A., and A.R. Parker. 2000. Trunk segmentation in some podocopine lineages in Ostracoda. Hydrobiologia 419:15–30.

Crustacean Biodiversity and Disparity of Body Plans VanHook , A.M., and N.H. Patel. 2008. Crustaceans. Current Biology 18:R547–R550. Walossek , D. 1993. The Upper Cambrian Rehbachiella and the phylogeny of Branchiopoda and Crustacea. Fossils and Strata 32:1–202. Walossek , D., and K. Mü ller. 1990. Stem-lineage crustaceans from the Upper Cambrian of Sweden and their bearing upon the position of Agnostus. Lethaia 23:409–427. Walossek , D., and K.J. Mü ller. 1994 . Petastomid parasites from the Lower Palaeozoic of Sweden. Transactions of the Royal Society of Edinburgh, Earth and Environmental Science 85:1–37. Walossek , D., J.E. Repetski, and K.J. Mü ller. 1994 . An exceptionally preserved parasitic arthropod, Heymonsicambria taylori n. sp. (Arthropoda incertae sedis: Pentastomida), from Cambrian-Ordovician boundary beds of Newfoundland, Canada. Canadian Journal of Earth Sciences 31:1664–1671. Wheeler, W., G. Giribet, and G.D. Edgecombe. 2004 . Arthropod systematics: the comparative study of genomic, anatomical, and paleontological information. Pages 281–295 in J. Cracraft and M. Donoghue, editors. Assembling the tree of life. Oxford University Press, New York . Williams, T.A. 1998. Distalless expression in crustaceans and the patterning of branched limbs. Development Genes and Evolution 207:427–434. Williams, T.A., and G.B. Mü ller. 1996. Limb development in a primitive crustacean, Triops longicaudatus: Subdivision of the early limb bud gives rise to multibranched limbs. Development Genes and Evolution 206:161–168. Willis, J.C., and G.U. Yule. 1922. Some statistics of evolution and geographical distribution in plants and animals, and their significance. Nature 109:177–179. Wingstrand, K.G. 1972. Comparative spermatology of a pentastomid, Raiallietiella hemidactyli, and a branchiuran crustacean, Argulus foliaceous, with a discussion of pentastomid relationships. Biologiske Skrifter af det Kongelige Danske Videnskabernes Selskab 19(4):1–72. Yager, J. 1994 . Speleonectes gironensis, new species (Remipedia: Speleonectidae), from anchialine caves in Cuba, with remarks on biogeography and ecology. Journal of Crustacean Biology 14:752–762.

33

2 EVOLUTION OF CRUSTACEAN APPENDAGES

Joachim T. Haug, Andreas Maas, Carolin Haug, and Dieter Waloszek

Abstract The evolutionary history of the postantennular appendages of Crustacea is reviewed, including information on limb development early in the evolutionary lineage of this taxon. This is particularly well demonstrated in the exceptional three-dimensionally preserved Cambrian fossils of the “Orsten” type (~500 million years old). Crustaceans started with serially similar limbs obtained from euarthropod ancestors, the “euarthropodium” comprising a biramous limb with a joint membrane and a prominent rigid stem portion, the basipod , carrying two rami, called endopod and exopod , of different phylogenetic origin. However, as a key innovation, they used the anterior three appendages (the “still” food-gathering antennulae and two more limbs) to collaborate as a set for feeding and swimming. For this collaboration, the two postantennular limbs had special outer lateral rami, exopods with fine annulation, and swimming setae inwardly positioned; a setiferous “proximal endite” developed medioproximal to the basipod. Further changes of the appendages and the effects on the feeding and locomotory system are followed along the evolutionary lineage of the crustaceans into the crown group, Eucrustacea. Acknowledging these changes is crucial to understand the high degree of variation of modern crustacean limb morphology and to overcome difficulties in recognizing their common features in terms of homology and relationships. The high plasticity of crustacean limb morphology is, in fact, not surprising since the major branchings along the crustacean lineage had already occurred far back in the Cambrian. It also means that there is no general “crustaceopodium”; instead, each limb must be viewed individually in terms of its history and its fate. Moreover, restructuring affected virtually all limb parts, including the possibility of formation of epipod(ite)s/gills (not mentioned in detail herein) and development of numerous types of surface outgrowths (spines, setae, and subordinate structures; discussed elsewhere in this volume) for various duties. Also, the rami underwent

34

Functional Morphology and Diversity. Edited by Les Watling and Martin Thiel. © 2013 Oxford University Press. Published 2013 by Oxford University Press.

Evolution of Crustacean Appendages significant changes regarding functional adaptations. One example of this change is that the rami became paddle shaped and/or symmetrical (of same morphology) as an adaptation to swimming, and sometimes multiannulated or even lost. Other strategies that evolved subsequently in eucrustacean ingroups include the arrangement of limbs into functional units and consequent changes in their morphologies, and the high modification of limbs for very specific purposes, for example, reproduction/copulation. Lastly, limbs are also lost repeatedly in various taxa. Reconstructing the evolutionary history of limbs along different crustacean lineages is still a major task for future research.

HOW DID THE CRUSTACEAN APPENDAGES EVOLVE? Noncrustacean euarthropod taxa such as myriapods, insects, and the arachnid chelicerates have comparatively uniform, exclusively uniramous appendages (except for appendages 2–4) that function mainly for walking. Moreover, they are made of one long, segmented element and are restricted, at least in arachnids and insects, to either the head region (in arachnids called prosoma) or the thorax. The many thousands of species of Crustacea, however, exhibit a remarkably large variation regarding the morphology of their appendages, particularly the initially biramous postantennular appendages, which are the focus of this chapter. At one extreme, crustacean appendages may display high multifunctionality (Swiss Army knife effect); at the other extreme, they or parts of them may be specialized for a particular function. To highlight the problem of comparability, limbs that are used exclusively for walking, as in several malacostracan taxa, appear, at first sight, strikingly similar to those of the (mostly) terrestrial euarthropod taxa mentioned above; that is, they are made of one long, segmented element. This superficial similarity is misleading, however, because even this design is based on a quite different original morphology. Yet, it is fairly easy to distinguish crustacean limbs from those of noncrustacean taxa, mainly because of the morphology of their proximal parts, original constitution, and details of joint morphology—but only when we have knowledge of the morphologies in the surrounding taxa and of the historical traits. When tracing limb morphology back in time, postantennular limbs in the ground pattern of the crown arthropods, the Euarthropoda, were indeed all “biramous.” This means that they consisted of a rigid, large, platelike basal structure, the “basipod,” that carried one rod-shaped ramus mediodistally, the so-called endopod, and a paddle-shaped structure on its sloping lateral edge, the exopod (basipod = carrier of the rami; e.g., Waloszek et al. 2005, 2007). Consequently, uniramy within euarthropods constitutes the apomorphic state. Development of uniramy in arachnid chelicerates is readily reproducible since postcheliceral prosomal arachnid appendages possess the basipod as a small proximal element and an elongate endopod made of several articles, possibly more than eight originally. The disappearance of the exopod is evidenced by the presence of this ramus on opisthosomal limbs and on the last walking limb (called flabellum; see Boxshall 2004) of xiphosurids. Moreover, within arachnids, the basipod even becomes fixed to the body. This results in a situation where the endopod-basipod joint is the limb-mover joint, and the main part for walking is the endopod: arachnids are “endopod” walkers (and head walkers). In all uniramous myriapods, insects, and crustaceans, the main joint remains the body-limb joint, so they are “whole-limb” walkers. Uniramy in crustaceans has apparently developed many times convergently, such as in various taxa of malacostracan crustaceans, for example, amphipods, isopods, and different ingroups of decapods. In these “thoracic walkers” (rather, walking on limbs of thorax I sensu Walossek and Mü ller 1998a), the proximal limb portion is a coxa (see below for its origin), while the

35

36

Functional Morphology and Diversity (plesiomorphically retained) basipod is the second limb portion and about as small as the coxa. This basipod then gives rise to the basically five-segmented and astonishingly uniform, elongate endopod. In all cases, the original state as deducible from the related taxa is the possession of an exopod stemming from the basipod, so there is convergence in all cases. Uniramy also occurs in Entomostraca, such as in some ostracodes and in the predatory water fleas (onychopods and haplopods). The latter do not really walk on their thoracopods, but use them for many tasks, including grasping prey. The number of limb portions is also different from that of malacostracans because a coxa is lacking from the beginning. The situation is even more complex for myriapods and insects because we cannot apply, at present, the terms coxa and basipod to the proximal limb elements due to lack of reference structures, for example, the exopod. The question therefore arises of how this large variety of appendage morphologies evolved in crustaceans. Moreover, what was the original condition and morphology that crustaceans received from their ancestors to start with? With this and a better knowledge of what the appendages looked like in the ground pattern of Crustacea, we may understand what they were used for, how they evolved subsequently, and what the probable driving forces were. It is, in this context, also important to clearly homologize the different limb parts—to use an appropriate and consistent terminology—in order to trace them from the beginning of limb formation in arthropods to modern crustacean taxa. Clearly, the basic form was laid down much earlier, and crustacean postantennular limbs retained (as a “historical burden”) much of the morphology that was present even in the stem taxa of Arthropoda sensu stricto (= s. str.) and Euarthropoda, at least initially (see below). It seems useful, therefore, to start this review with a look at our current understanding of postantennular limbs before Crustacea. This might help us to better understand the specific changes in the early evolutionary lineage of crustaceans that led to the conditions developed within the modern ingroups. This might also allow us to overcome the difficulties with morphologies and terminologies people have had in the past when addressing the question of the general morphology of crustacean appendages and their origin—particularly those attempts made before data became available about the early to late Cambrian fossils of the Chengjiang and “Orsten” fossil deposits (lagerstätten). Examples of such historical studies are Hansen (1925), Størmer (1939), Heegaard (1945), Snodgrass (1958), and Kaestner (1967). Neglecting the fossil data even after this information became available is not much better. And sometimes it appears that there are as many different answers as there are researchers in this field—all researchers seem to have developed their own ideas and terminologies (examples of more recent attempts: Boxshall 2004, Williams 2004, Boxshall and Jaume 2009). Accordingly, there is still no consensus in sight today. Our review cannot cover all of the different morphologies present in Crustacea but instead seeks to focus on some major issues, such as origin and fate of the crustacean postantennular appendages and their major components. Also, some examples of the capabilities crustaceans have achieved are given to demonstrate that crustaceans are the euarthropod taxon that made the most use of its appendages—and did so very successfully.

THE “ORSTEN” PERSPECTIVE Our overview is based on data from studies of extant larval and adult Crustacea, including our own investigations, and particularly on the Cambrian fossil record of the Crustacea, which is evident exclusively from Orsten crustaceans with an age of 520–495 million years. There are multiple reasons for considering this fossil evidence. First, Orsten fossils are old, but superbly three-dimensionally preserved, thus interpretable almost as living forms. Second,

Evolution of Crustacean Appendages because of their ancient age, these fossils exhibit characters or character combinations that may not, or should no longer, exist in extant crustaceans, therefore helping to identify early evolutionary pathways. Third, different ontogenetic stages have been uncovered from several of the Orsten-type fossils, particularly taxa of the crustacean lineage, permitting interpretations of the sequential morphogenesis of structures. Three-dimensional Orsten preservation in full detail, down to even submicrometer scale and in topologically correct position, also facilitates functional-morphological interpretations, which represent another helpful tool in understanding evolutionary pathways. Fourth, these Cambrian crustacean fossils comprise species that have been demonstrated to represent different evolutionary levels. Accordingly, their data can aid in sorting features in the order of evolutionary and ontogenetic appearance. The resulting phylogenetic tree directly ref lects the appearance of particular features on certain nodes, with some species already possessing certain features that others (still) lack. This knowledge can be readily used for reciprocal illumination sensu Hennig (1965). Fifth, these fossils help us to progressively establish a consistent terminology for crustacean and even arthropod appendages, including a coloration scheme for homologous parts of limbs based on the fossil material, which can be followed up from Arthropoda s. str. (the evolutionary level of sclerotized arthropods; see Maas et al. 2004) into the different crustacean ingroups (e.g., Walossek 1993; see also Walossek and Mü ller 1990, 1998a, 1998b, Maas et al. 2003, 2004, Waloszek 2003a, Waloszek et al. 2007, Zhang et al. 2007, Stein et al. 2008, Haug et al. 2009a). Our approach permitted, for the first time, large-scale comparability between the various crustacean and euarthropod taxa. This helped not only to sort in new taxa, including two-dimensionally preserved forms (e.g., from Chengjiang and Burgess lagerstätten) but also to identify modifications of known morphologies and to correct past errors in interpretation. This allowed for various new better-known taxa to be added to our knowledge base and significantly improved our understanding of the evolutionary pathways of structures and structural systems in arthropods and Crustacea. It is particularly the evidence of structures and structural systems in the early fossil record that we consider useful for understanding phylogeny and evolution—sometimes even more than any other data, because the once living morphologies are the proof of existing structures and are always superior to models of any kind. Altogether, research on fossils in Orsten-type preservation and the huge character data set provided by the other sources has challenged traditional assumptions on the early evolution of Crustacea and has significantly stimulated this subject. The result of this is, it is hoped, a straightforward, consistent view of the evolution and phylogeny of the Crustacea and their appendages. Sources for Fossilized Appendages In studying the early radiation of arthropods in the Cambrian (first geological period for the fossil record of animals), we made use (mainly) of four lagerstätten: the North American Burgess Shale, the Sirius Passet in northern Greenland, the Chinese Chengjiang biota, and the world-wide occurring Orsten. The name Orsten has been applied to limestone nodules, which are embedded within alum shales. They were first found in Sweden in rocks dated to the middle to upper Cambrian, but they occur worldwide (North America, Europe, Siberia/Russia, China, and Australia) from the Terreneuvian series (former lower Cambrian; ca. 520 mya) to the Lower Ordovician (ca. 490 mya) (Maas et al. 2006). While fossils from Burgess Shale, Sirius Passet, and Chengjiang provide us with data from the level of Arthropoda sensu lato (= s. l.) up to Euarthropoda (unassignable taxa, chelicerates, trilobites and allied), only the Orsten has yielded confirmed crustacean fossils from the Cambrian (besides lobopods, chelicerates, and agnostid euarthropods).

37

38

Functional Morphology and Diversity Several Cambrian species have been claimed to be Crustacea. A few prominent examples of so-called “crustaceans” from the Burgess Shale are Branchiocaris pretiosa Resser, 1929 (see Briggs 1976), Waptia fieldensis Walcott, 1912 (Heldt 1954), and Canadaspis perfecta Walcott, 1912 (Briggs 1978). There are further examples from the Chengjiang biota, such as Pectocaris spatiosa Hou, 1999 (Hou et al. 2004), Ercaia minuscula Chen et al., 2001 (Chen et al. 2001), and Isoxys auritus Jiang, 1982 (Shu et al. 1995). Although all these species have been assigned to Crustacea or even to ingroups, many if not all of them may be euarthropods or, in some cases, not even members but stem derivatives (e.g., Dahl 1984, Maas and Waloszek 2001a, Taylor 2002, Waloszek et al. 2005, Budd 2008). Indeed, interpretations of the mentioned species as crustaceans suffered from a rather incomplete knowledge of their morphologies, especially concerning the segmental composition of the head or the (important) proximal parts of the appendages. Likewise, the Cambrian record of malacostracans (based on the stem euarthropod Canadaspis perfecta; see Maas and Waloszek 2001a) and ostracodes (bradoriids and phosphatocopines misidentified as eucrustacean ostracodes) is based simply on misinterpretations, even though it is mentioned in modern textbooks. Until now, no unequivocal malacostracan and no ostracode had been uncovered in the Cambrian (Siveter 2008). Orsten-type preservation, by contrast, provides a rather complete view of historic (crustacean) animals, because Orsten fossils are uncompressed and exhibit the finest details, such as setae carrying minute setules, eye structures with visible facet patterns, and weakly sclerotized membranous areas such as the arthrodial membranes of the appendages, with the smallest visible structures being only 0.2 μm (for a summary, see Maas et al. 2006). The reason for this may be the fine impregnation of the epicuticular surface by phosphate (apatite), which conserves the finest denticles and pores. Strangely, and for unknown reasons, the fossils do not exceed a length of 3 mm, and the best-preserved ones are only 100–200 μm in length. Thus, a large amount of larval and young developmental stages occurs in the material. Given the small sizes of the preserved organisms, species from the Orsten are rarely represented by adults. If preserved, these adults are small in size, such as for the species of Skara Mü ller, 1983 (e.g., Mü ller and Walossek 1985, Liu and Dong 2007) and Dala peilertae Mü ller, 1982 (Walossek and Mü ller 1998b), or possibly Bredocaris admirabilis Mü ller, 1983 (Mü ller and Walossek 1988). A larger number of taxa are known only from certain early postembryonic (larval) stages or sets of them (e.g., Mü ller and Walossek 1986a, Walossek and Mü ller 1989). Some species are even represented by a larger set of ontogenetic stages that allow the reconstruction of more or less complete ontogenetic sequences. Examples are the euarthropod Agnostus pisiformis Wahlenberg, 1818 (Mü ller and Walossek 1987) and various crustacean taxa such as Hesslandona unisulcata Mü ller, 1982 (Maas et al. 2003; see Fig. 2.1E,F), Henningsmoenicaris scutula (Walossek and Mü ller 1990) (Haug et al. 2010a), and Rehbachiella kinnekullensis Mü ller, 1983 (Walossek 1993; see Fig. 2.1I,J). This large set of exceptionally preserved fossils also provides us with a wealth of morphological features and allows us to study the ontogeny and, particularly, the morphology and morphogenesis of appendages in much detail. In a combined approach using data from extant taxa, fossils, and data from ontogenies of both sources (although size limitation might cause a slight bias), Orsten-type fossils have yielded key information for understanding early crustacean evolution and might be regarded as the principal source. While Schram (1986) initially considered them “larval f lotsam,” and Lauterbach (particularly Lauterbach 1986) tried to promote the view that Orsten fossil taxa are nothing but a bunch of stem mandibulates, intense study of Orsten species over almost 30 years uncovered them as a set of taxa derived from very different evolutionary levels. This differentiated view and understanding of their morphologies in the light of changes along the evolutionary lineage of Crustacea permitted the recognition of the specific fate of particular appendages, their morphology, and their association

Evolution of Crustacean Appendages

Fig. 2.1. Examples of “stem crustaceans” and labrophorans (Phosphatocopina + Eucrustacea) in the Orsten material (scanning electron microscopic images and reconstructions). (A and B) The cambropachycopid Goticaris longispinosa Walossek and Mü ller, 1990. (C and D) The “stem crustacean” Martinssonia elongata Mü ller and Walossek, 1986. (E and F) The phosphatocopine Hesslandona unisulcata Mü ller, 1982. (G and H) The entomostracan Yicaris dianensis Zhang et al., 2007. (I and J) The branchiopod Rehbachiella kinnekullensis Mü ller, 1983. In this figure and those that follow, repository numbers and museum names are given so that the reader can locate these specimens: BM, Natural History Museum, London; MB.A, Museum f ü r Naturkunde, Berlin, Germany; UB, University of Bonn, Germany; YKLP, Yunnan Key Laboratory for Paleontology, Yunnan University, Kunming, China. (A) UB 98; (C) UB 752; (E) UB 1570; (G) YKLP 10841 (Zhang et al. 2007, their fig. 1a); (H) from Zhang et al. 2007, their fig. 2; (I) UB 644. All figures used with permission of the authors.

39

40

Functional Morphology and Diversity within the series of limbs along the body. These fossils could also be used to illuminate the early branchings of the crustacean lineage into lines with living descendants and the evolution of structural systems such as locomotory and feeding apparatus (e.g., Stein et al. 2005, Waloszek et al. 2007). In fact, several species from the Cambrian Orsten fossil sites could thus be assigned to the crown group of Crustacea, Eucrustacea (Müller 1983, Müller and Walossek 1985, 1988, Walossek 1993, Zhang et al. 2007) and are even deeply nested within extant taxa. Others belong to the Phosphatocopina, a group of small bivalved forms that could be identified as the sister group of Eucrustacea (Phosphatocopina + Eucrustacea = Labrophora; see Maas et al. 2003, Siveter et al. 2003; see also Waloszek 2003a, 2003b, Maas and Waloszek 2005). Lastly, a set of taxa turned out to represent derivatives of earlier branchings (see below). Until now, the Orsten is the only paleontological source providing species that are derivatives of the stem lineage of Labrophora, thus extending the view of the crustacean stem lineage further back in time phylogenetically (Walossek and Müller 1990, Stein et al. 2005, 2008, Haug et al. 2009a, 2010a, 2010b; see Fig. 2.1A–D). Appendage Morphologies The Basis, 1: Appendages at the Level of Arthropoda sensu lato and sensu stricto Before discussing the appendage morphologies of crustaceans, we need to know more about the evolution and functional history of these limbs. This will also ensure that we understand the respective homologies and apply appropriate terminology to these appendages. Accordingly, we need to know about the ground pattern status of limbs before the crustacean level, that is, the postantennular limbs in the ground pattern of Euarthropoda and, even prior to this, of the Arthropoda s. str. and s. l. (see Maas et al. 2004, Waloszek et al. 2005). The arthropod system as it is used throughout this chapter has been developed over a period of almost 30 years and laid down in many of our papers. It is presented in Table 2.1 in a simplified written version, following the notation proposed by Hennig (1965) and Ax (1995).

Table 2.1 Phylogenetic system of Arthropoda. Arthropoda sensu lato Maas et al., 2004 (= Aiolopoda Hou and Bergström, 2006 [see Bergström et al. 2008]) Arthropoda sensu stricto Maas et al., 2004 (= Arthropoda sensu Bergström et al. 2008) “Fuxianhuiidae” Hou and Bergström, 1997 Euarthropoda Walossek, 1999 Chelicerata Heymons, 1901 Myriapoda Latreille, 1796 Insecta Linnaeus, 1758 Crustacea sensu lato Stein et al., 2008 (= Crustacea Brünnich, 1772) N.N. 2 = unnamed sister taxon Labrophora Siveter, Waloszek, and Williams, 2003 Phosphatocopina K.J. Müller, 1964 Eucrustacea Kingsley, 1894 (sensu Walossek 1999) Entomostraca O.F. Müller, 1785 Malacostraca Latreille, 1802

Evolution of Crustacean Appendages A

B

C mem

basipod

endopod

exopod

D mem

E mem

proximal endite / coxa

Fig. 2.2.

Postantennular appendages from Arthropoda sensu stricto to Eucrustacea. mem, membrane. (A) Arthropoda sensu stricto: Fuxianhuia protensa Hou, 1987. (B and C) Euarthropoda. (B) Leanchoilia illecebrosa Hou, 1987 (modified after Liu et al. 2007). (C) Agnostus pisiformis Wahlenberg, 1818. (D) Crustacea sensu lato: Martinssonia elongata Müller and Walossek, 1986. (E) Eucrustacea: Skara anulata Müller, 1983. Color scheme modified from Walossek (1993): black, proximal endite/coxa; light gray, basipod; medium gray, endopod; dark gray, exopod. All figures used with permission of the authors. The crustacean postantennular limbs consist of four major parts, for which we have to identify the traits in earlier morphologies. The ground pattern of the soft cuticle-bearing arthropods in the broad sense (Arthropoda s. l.) forms the start, because we have some knowledge from extant onychophorans and about 10 Cambrian so-called lobopodians (Maas et al. 2007). Legs of such forms (and autapomorphy of the taxon Arthropoda s. l.) were tubular, possibly ending in a distal pair of claw hooks, and as soft as the body proper. Arthropods in the strict sense (Arthropoda s. str.) have, among other features, a much more sclerotized dorsoventrally flattened body, each segment bearing a tergitic dorsal part connected by softer membranous cuticle (the evolutionary process is called arthrodization). This morphology is known now from three two-dimensionally but well-preserved, several-centimeters-long lower Cambrian taxa: Fuxianhuia protensa Hou, 1987, Chengjiangocaris longiformis Hou and Bergström, 1991, and Shankouia zhengei Chen et al., 2005 (Waloszek et al. 2005; see also Figs. 2.2A, 2.3A), traditionally combined in the taxon Fuxianhuiidae with unclear phylogenetic status. These fossils also helped to reconstruct the appendages in the ground pattern of Arthropoda s. str. The main portion of their limbs, or limb stem, is tubular as before, but it is apomorphically annulated consisting of approximately 20 sclerotic rings with membranes between. The articles of these “arthropodia” articulate against each other via two opposing so-called pivot joints, that is, knoblet-against-depression joints (evolutionary process: arthropodization). This construction might have guaranteed flexibility as before but also provided much more stability for walking on the limbs with the body lifted up to some degree. Yet the joint angle of bending between the articles is limited—possibly the high number of articles counterbalanced this. Another new limb structure is a paddle-shaped structure on all postantennular limbs arising from the outer proximal edge of the limb stem (Fig. 2.2A, 2.3A), which is missing, however, on the first appendage, the antennula. The new postantennular arthropodium (Fig. 2.2A) consists, therefore, of two major elements and one significant detail: the multiannulated stem (originating from the lobopodium, but now segmented), the pivot joints (new), and the flaplike outer ramus (new). In a way, this is a biramy induced at this level, but it is different from the next step (the Euarthropoda level)

41

42

Functional Morphology and Diversity A

B

C

D

E

F

G

H

? I

basipod endopod exopod proximal endite / coxa

Fig. 2.3. Schematic evolution of the arthropod limb apparatus from Arthropoda sensu stricto into Eucrustacea: ground pattern status of the limb series. (A) Arthropoda sensu stricto based on Shankouia zhenghei Chen et al., 2005 in Waloszek et al. (2005). (B) Euarthropoda, based on different taxa; number of endopod segments and exopod morphology mainly based on Leanchoilia illecebrosa Hou, 1987 (Liu et al. 2007, used with permission). (C) Crustacea sensu lato based on Oelandocaris oelandica Mü ller, 1983 (Stein et al. 2008). (D) Crustacea sensu lato excluding Oelandocaris oelandica, based partly on Martinssonia elongata Mü ller and Walossek, 1986; ? indicates that some aspects are still unclear, such as exopod morphology. (E) Labrophora (Waloszek 2003b, Maas et al. 2003, Waloszek et al. 2007). (F) Phosphatocopina (Maas et al. 2003, Maas and Waloszek 2005); ? indicates unclear condition of coxa-basipod border in second appendage. (G) Eucrustacea (Maas et al. 2003, Waloszek 2003b, Waloszek et al. 2007); ? indicates that the exact ground pattern condition of maxillula is unclear. (H) Entomostraca (Waloszek 2003b). (I) Malacostraca (Waloszek 2003b). Gray shading scheme is as described in Fig. 2.2. All figures used with permission of the authors.

because only the flap is a ramus, in the sense of subsequent evolutionary steps. The stem is in fact the main part that may be subdivided into functional portions and not a ramus. It also demonstrates that the two portions, stem and flap, are of different age phylogenetically and by no means symmetrical. It may be, as known from subdivided limb parts of living arthropods and particularly crustaceans, that all articles were interconnected by fine muscle strands operating as intrinsic and extrinsic musculature, with the possible exception of the terminal article (see Waloszek et al. 2007, their fig. 4G). The flap may have served to some degree as a respiratory surface but more likely as a locomotive aid, permitting the animal to produce some water flow around the body. Thus, it may have initiated a swimming mode of life for arthropods in addition to their crawling lifestyle. It is also important to note the body parts that the early arthropodium lacks: there is, most likely, no basal limb joint; there are only a few shorter annules; and no setae or spines (immobile stronger outgrowths) developed, either on the stem or along the margin of the flap. It

Evolution of Crustacean Appendages seems that the limbs were exclusively used for locomotion—and were all very similar to each other (serially similar). The only structure possibly useful for food grasping and transport to the mouth was the uniramous 15-segmented antennula (Waloszek et al. 2005; see Fig. 2.3A), the first head appendages associated with the deutocerebrum. And at least in Fuxianhuia protensa it may have carried one fine spine medially on each of the articles (Hou and Bergström 1997, Waloszek et al. 2005). The Basis, 2: Appendages at the Level of Euarthropoda Various species, such as Canadaspis perfecta (Briggs 1978), have the potential of helping researchers document the evolutionary lineage to the next level, Euarthropoda. However, our knowledge of these species is still too incomplete for a more conclusive discussion. While the antennula apparently has not changed morphologically or functionally (grasping, food collecting aid) in the ground pattern of Euarthropoda, two major events toward this level had taken place—most likely achieved by more than just one single stem species. One event is the development of a larger head unit, including the fact that now all postantennular limbs, though remaining serially similar, are grouped into three cephalic and a set of trunk limbs. The other event is the development of a significantly different postantennular limb, which in the ground pattern of Euarthropoda possesses three major elements and another important structure (Figs. 2.2B,C, 2.3B). All postantennular limbs are anteroposteriorly compressed, therefore extended in mediolateral direction and not circular in cross section like the limb stem of the arthropodium. The first is a large subrectangular to triangular basal portion. A rodlike, distally tapering and possibly nine-segmented structure (articles interconnected by pivot joints) arises mediodistally from this portion. Another, paddle-shaped structure, marginally adorned with setae, inserts on the outer sloping margin of the stem portion. Importantly, the stem portion, as we know from several Cambrian fossil arthropods, shows more details, demonstrating the significance of this level for the evolution of arthropod and crustacean limbs. All rigid basal portions arise from the ventral body proper like bricks placed abaxially on their narrower edges in a regular series on the left and right sides at some distance between the pair, leaving a median path for food transport. Another morphological detail is that the basal portions articulate with the body in an ample folded arthrodial or joint membrane. Such folds are even visible in two-dimensionally preserved fossils (e.g., Leanchoilia illecebrosa Hou, 1987; see Liu et al. 2007). This joint membrane may have facilitated movability of the limb, while the rigidity of the basal portion may have permitted attachment of stronger muscles within the stem in order to enhance operability of the limb. The elongation of the limb in mediolateral direction limits movability to a largely forward-backward swing, but this was even enhanced by a second feature: anterior and posterior sides of the limb stems are of different morphology. The anterior edge is straight proximally, while the posterior edge is excavated. This condition may have permitted wider backward than anterior swinging (e.g., Liu et al. 2007). The next noteworthy feature is that the elongate narrow median edge of the stem portion carries a row of spines, or pairs of spines, from proximal to distal pointing medially into the interlimb space or food path. It is likely that these structures aided in food transport toward the mouth, but it is possible that they (and others mentioned here) developed earlier, as such structures are developed in the putative stem lineage derivate Canadaspis perfecta. It should be noted that early in the evolution of Arthropoda and Euarthropoda, it is very difficult to differentiate between immobile spines and spinelike setae that have a joint at their base. Fossils are not preserved well enough to verify this. In our Orsten forms, this detail can be observed more clearly, but it seems that because of size, many more robust structures are simply lacking a joint, although they have a socket; thus, they should be named spines. By contrast, marginal outgrowths of exopods used as swimming aids should continue to be named setae, although

43

44

Functional Morphology and Diversity they are not necessarily mobile. Outgrowths of small entomostracan crustaceans are not easily compared with those of Malacostraca (see below). Terminological classification may not be easily applied across larger taxa because such typification may also be phylogenetically misleading. We have called this basal portion of Euarthropoda basipod (e.g., Waloszek et al. 2005). The basipod not only has a characteristic shape but also carries the two rami (see above). The endopod arises mediodistally and in line with the spine-bearing inner edge of the basipod. It is rodlike and roughly circular in cross section. Its portions, called podomeres, are all about the same size and are slightly humped mediodistally, each hump bearing one or two spines or setae, very similar to those along the median edge of the basipod. The distal smaller element bears a tuft of one or few spines or setae. It remains unclear if there were eight, nine, or even more endopodal podomeres originally, and it is also unclear if and how much the endopod was proximally partly fused with the outer f lap (Liu et al. 2007, their fig. 5, Stein et al. 2008, their fig. 7D–F). The flap arising from the sloping outer margin of the basipod is fringed with marginal setae. The oblique orientation of this joint permits the flap to be held more laterally, therefore facilitating or improving the use of this flap as a locomotory device. Subdivision of the flap into a triangular proximal part and a distal portion might even enhance locomotory abilities (Liu et al. 2007, Stein et al. 2008, Haug et al. 2010a), though we cannot yet validate this conclusively due to lack of more evidence. We understand this setose flap, the exopod, as homologous to the lateral flap of the arthropodium in the ground pattern of Arthropoda s. str. It remains unclear if the exopod was individually movable by musculature at this stage, as demonstrated in extant paddle-shaped exopods of crustaceans (e.g., branchiopods). The entire morphology of the “euarthropodium,” as named herein, characterizing the level of Euarthropoda, must be understood as derived from prior morphologies. It may hence be speculated that the folds of the arthropodial membrane and the sets of spines along the inner basipod margin are indicators of an original subdivision, possibly corresponding to annules of the limb rod of the “arthropodium.” Consequently, the cuticle of the proximal articles may have become softer to form the basal joint area, while a set of more distal articles should have elongated in abaxial aspect and fused to form the brick-shaped basipod. The only intermediate situation must have been the development of spines, possibly in the course of achievement of a feeding function of the proximal part of the limb stem. The limbs of the Cambrian species Canadaspis perfecta (Briggs 1978, his fig. 108) could indeed give a hint to such a pathway (Maas and Waloszek 2001a), but this has to be verified in more detail. Continuing this line of logic, the remaining part of the original multiannulated stem of the arthropodium may have evolved into the endopod of the euarthropodium. Again, the limb of Canadaspis perfecta can help us understand this change, because it has spines mediodistally on all its distal articles. A kind of functional split seems indicated, which lets us view the evolutionary path to a complex multifunctional arthropodium, with its proximal part serving for feeding and the distal for locomotion plus food intake. While the basipod muscles may have become somewhat concentrated, the endopod retained interior muscles running individually from podomere to podomere, which indeed can still be observed in the endopods of extant crustaceans. In any case, it becomes quite evident also for this euarthropodium that endopod and exopod are very different structures in terms of origin and morphology. Likewise biramy is a problem because the original arthropodium consisted in fact of only a stem and a single extra ramus, while the euarthropodium has a basipod that carries two real rami (endopod and exopod). Yet, the endopod is simply not the old stem. In fact, the stem has very likely been split into one piece forming the membrane, one forming the basipod and one forming the endopod. The only conservative trait in this scenario is the f lap-shaped exopod, which from its appearance in the arthropodium did not change its shape and function in locomotion

Evolution of Crustacean Appendages (swimming), although, in addition, setae appear along the margin. The exopod has therefore almost as deep phylogenetic roots as the stem, but its first appearance can be traced only to the level of Arthropoda s. str. at present. It may, however, be possible that it had appeared earlier, but informative fossils in these surroundings are still too poorly understood (e.g., Budd 1998, Liu et al. 2007). The stem, on the other hand, can be traced down to the “lobopodium” of Arthropoda s. l. The complex limb morphology of the euarthropodium indicates that it had developed into a multifunctional tool, using the whole set of appendages in a simple heterochronal movement cycle that aided in locomotion as much as in food intake and manipulation. Moreover, the two processes were rather neatly coupled. The basipod spines could assist in food transfer to the mouth, while the endopod aided in locomotion and, with its spines, also aided in food manipulation. The distal spines might even have permitted some scratching and sorting of food. Though everything was (still) strictly serial, it was a huge innovation for arthropods to have such a three-part euarthropodium comprising a basipod carrying the endopod and exopod. Indeed, it was so successful that much of it—for example, the slope for the exopod—is still present in certain appendages particularly of early larval stages of crustaceans (e.g., cirriped nauplii), as well as in the Orsten crustaceans (see below). This morphology, particularly the slope, can be used as a nice reference when searching for homologies of structures in different taxa and for testing hypotheses about limb evolution in arthropods (see also Fig. 2.3). The morphology must have been so effective that this tripartite euarthropodium (not two or four parts! see Fig. 2.2B) and the strict seriality of the postantennular limbs was transferred into the various descending evolutionary lineages. Various Paleozoic representatives of the Euarthropoda such as trilobites and allied taxa exhibit this morphology and seriality (though some head limbs may be smaller than the rest, as in Emeraldella brocki Walcott, 1912 or Parapeytoia yunnanensis Hou, Bergström, and Ahlberg, 1995). Parts of this complex limb morphology may well have developed earlier. Chelicerates are another example taxon with extant descendants that retained this morphology during their evolution, changing little more than the grasping antennula into a short chelicera (e.g., Chen et al. 2004, esp. their fig. 5). The morphology of the entire apparatus suggests that it was a conjointly feeding and locomotion system: stopping locomotion meant stopping feeding. This may have been the reason that this strategy was not taken over completely by all successive euarthropod lineages, at least in those reducing the food-intake possibilities of more posterior limbs, as in the modern chelicerates, myriapods, and insects. In a way, Crustacea are the ones among the taxa with living representatives that retained most of the morphology of the euarthropodium in their ground pattern—also Agnostus pisiformis Wahlenber, 1818, a minute euarthropod (Mü ller and Walossek 1987) with supposed affinities to crustaceans (Walossek and Mü ller 1990, Stein et al. 2005, Haug et al. 2010a). And they made the most use of it. Initially, however, much of the seriality was retained. Appendages at the Level of Crustacea sensu lato The appendages of crustaceans indeed exhibit such various plesiomorphic traits. This is most evident in several Orsten fossil crustaceans that we refer to as stem derivatives—more precisely, derivatives of the evolutionary lineage toward the Labrophora, comprising the Phosphatocopina and the Eucrustacea (e.g., Waloszek et al. 2007). Examples of plesiomorphies are the seriality and the gross morphology of the postantennular limbs, with the ample joint, the basipod (posteriorly more excavated) with median spines on a rather straight median edge, the mediodistal endopod with sets of spines, and the paddle-shaped exopod with marginal setae on the sloping edge of the basipod (Fig. 2.4A).

45

46

Functional Morphology and Diversity However, this holds true only for the more posterior limbs because all three anterior appendages, that is, the antennulae and the following two limbs, are more specialized in the ground pattern of Crustacea s. l. than are the corresponding appendages in the euarthropod ground pattern and therefore are more like those in modern crustaceans. Nevertheless, these fossil species combined as stem derivatives are not stem species themselves but exhibit also autapomorphies indicative of their own evolutionary lineage (the reason that we usually speak of stem-lineage derivatives). The following interpretations are therefore based not on single species or specimens but on more than half a dozen known species. Moreover, they apparently represent different evolutionary levels, and several of them are preserved as a set of instar stages. Based on this evidence, we have identified several evolutionary novelties of Crustacea in the wider sense. Many if not most affected the morphology particularly of the appendages and are coupled with the locomotory and feeding apparatus, apparently an important issue for this euarthropod group. These are, for example, • the deviation from a strict seriality of all postantennular body limbs (state of Arthropoda s. str. and Euarthropoda; Fig. 2.3A,B); • changing tagmosis of the head-trunk system, for example, by incorporation of more trunk segments into the head, and specializations of limbs; • changes in the locomotory and feeding system, again with differentiated development in the head and the trunk, likewise independent of the changing tagmosis; • the development of associated features with the changes in the functional system; and • changes in the morphology of the appendages, individually as well as in sets (further development of tagmata). The first significant changes of the appendages and their morphologies along the crustacean evolutionary lineage based on the stem derivatives are interpreted as autapomorphies in the ground pattern of Crustacea s. l.: 1. Antennulae uniramous, composed of a few tubular portions or articles, though large, limblike, and not feelerlike, but involved in both feeding and locomotion; several long setae along the posterior side of the appendage (originally the median side) and on the tip (see, e.g., Haug et al. 2009a for Cambropachycopidae or Stein et al. 2008 for Oelandocaris oelandica Mü ller, 1983). Comments: Plesiomorphies associated with the antennula include the insertion anterolaterally at the hypostome, and the function of the antennula possibly only for food gathering (a similar specialization has also been found in Agnostus pisiformis, which, by contrast, has a 15-segmented antennula with short spines). A sensorial antennula seems to have evolved convergently in several evolutionary lineages of the Euarthropoda, for example, in malacostracan crustaceans and atelocerate taxa (myriapods and insects). In the evolutionary lineage of chelicerates, the antennula changed into a highly shortened grasping element, the chelicera (Chen et al. 2004). 2. Anterior three appendages morphologically different from each other and from the more posterior ones. Comments: In the ground pattern of Arthropoda s. str., only the antennula differs from the serial biramous trunk limbs in being uniramous and the major food-raking element

Evolution of Crustacean Appendages

Fig. 2.4. Scanning electron microscopic images of different postantennular limbs of Orsten fossil crustaceans. (A) Isolated trunk limb of Henningsmoenicaris scutula Walossek and Mü ller, 1990, stage 10 with paddle-shaped exopod. No proximal endite is present at this ontogenetic stage. (B) Third appendage of Hen. scutula, stage 8, multiannulated exopod with few articles, distal part of endopod broken off. The proximal endite is still small at this evolutionary level. (C) Second appendage of Goticaris longispinosa Walossek and Mü ller, 1990, stage 2 with multiannulated exopod. (D) Proximal endite of a trunk limb of a very late ontogenetic stage of Hen. scutula. At this ontogenetic stage, trunk limbs also possess proximal endites. Scale bar, 10 μm. (E) Postmandibular appendage of the phosphatocopine Hesslandona unisulcata Mü ller, 1982, growth stage III. Note the large proximal endite, the enditic protrusions on basipod and endopod drawn out medially, and the exopod being paddle-shaped proximally and multiannulated distally. (F) Mandible of Hes. unisulcata with multiannulated exopod, short endopod, small basipod, and large coxa. (G) Cephalic appendages of Martinssonia elongata Mü ller and Walossek, 1986. Multiannulation of exopod is only weakly developed. (H) Postmandibular appendage of growth stage V of Hes. unisulcata. The exopod is paddle-shaped proximally and multiannulated distally as in Fig. 2.4E, but the multiannulated part is smaller in this specimen due to later ontogenetic stage. (I) Subequal (second) antenna (right) and mandible (left) of Skara anulata Mü ller, 1983, with multiannulated exopods. Abbreviations: bas, basipod; cox, coxa; en, endopod; ex, exopod; pe, proximal endite. Repository numbers (see Fig. 2.1 for abbreviations): (A) UB W 338; (B) UB W 335; (C) UB W 124; (D) UB 103; (E) UB 103; (F) UB W 106; (G) UB 780; (H) UB W 156; (I) UB 692.

47

48

Functional Morphology and Diversity (Fig. 2.3A). Also in the euarthropods, all postantennal limbs are serial. The second appendage is slightly shifted anteriorly in its position and may be smaller than the more posterior limbs but still resembles them (Fig. 2.3B). Here in crustaceans, all three anterior appendages move in accord and are somewhat set off from the posterior ones with their heterochronous beat (Fig. 2.3C). This points to a new type of combined locomotion and feeding in Crustacea s. l., the so-called “sweep-net feeding” (Waloszek 2003b). 3. Exopods of the first and second postantennular limbs comprising a tubular proximal part and a multiannulated distal part with each annulus carrying a long seta on the median side, which is directed mediodistally against the endopod (Fig. 2.4C); the proximal socket element may have served to raise the setae-bearing part slightly away from basipod and endopod in order to facilitate the swing of the setae. Comments: Plesiomorphically, the exopod is a simple flap with setae around its entire free margin (compare Fig. 2.3B,C). In the putatively closest relative to Crustacea s. l., Agnostus pisiformis (Stein et al. 2005), the exopods of the first and second postantennular limbs are also multiannulated, but bear setae on their lateral = outer side (Mü ller and Walossek 1987, their plates 18:1–5, 19:1). These two limbs are generally similar to each other, except that because of the different position regarding the mouth, the basipod and its median enditic armature were most likely already different. However, the most apparent new feature of Crustacea s. l. is the following: 4. Possession of a small setiferous sclerotized humped area, the “proximal endite,” within the body-basipod membrane of the postantennular limbs and clearly proximomedially of the basipod (Fig. 2.2D). Comments: This endite was termed as such by Walossek and Mü ller (1990; see also Walossek 1993) because of its morphological similarity to median basipodal endites of extant entomostracan eucrustaceans. Much earlier, Calman (1909, 51) had pointed to its existence in branchiopods and remarked upon it being most likely a very old feature. In contrast to Walossek and Mü ller (1990), who assumed its presence on all postantennular limbs originally, recent investigations of the stem derivatives demonstrated its first appearance only on the third appendage in young larvae, that appendage later called mandible in Labrophora (e.g., Stein et al. 2005, 2008, Waloszek et al. 2007, Haug et al. 2009a, 2010a, 2010b; Fig. 2.4B). The proximal endite is regarded as one of the key evolutionary characters of Crustacea (Walossek and Mü ller 1990, Walossek 1993, Waloszek 2003a, 2003b), possibly appearing only on the third appendage at first (Stein et al. 2005, 2008, Waloszek et al. 2007; but see Haug et al. 2010a). Further along the evolutionary lineage to the Eucrustacea, the proximal endite occurs on all postantennular appendages, having become larger and more setiferous (Figs. 2.3D, 2.4D). Its possible main purpose was food manipulation and transport, as seen still today, for example, in cephalocarids and mystacocarids (Fig. 2.5G,H). It is still unclear if the proximal endite was already movable at this early stage, as described, for example, for the proximomedial endite (and the more distal ones) of Cephalocarida (see Sanders 1963, Hessler 1964, his fig. 11). It also remains unclear if it served as a food-raking device from the very beginning. Its functionality as an aid in food transport is, however, clear from its later shape, position, and use in living taxa, indicative of its continuing significance in the

Evolution of Crustacean Appendages

Fig. 2.5. Scanning electron microscopic images of proximal and basipodal endites and appendage armatures. (A) Mandible of an undetermined phosphatocopine. Anterior and posterior sets of setae are not yet differentiated from each other. Numbers indicate different sets of setae: 1, anterior retention setae; 2, median set, usually developed as stronger spines; 3, posterior sieving setae. (B) Trunk limb of Rehbachiella kinnekullensis Mü ller, 1983; numbers are as in A. While the retention setae (1) are arranged in parallel rows, the sieving setae (3) form more or less a triangle. (C) Trunk appendage of Yicaris dianensis Zhang et al., 2007 with about the same arrangement of setal sets as R. kinnekullensis; numbers are as in A. (D) Trunk limbs of fossil R. kinnekullensis . (E) Trunk limbs of extant Cyclestheria hislopi (Baird, 1859) (Olesen 2007). Enditic arrangement very similar to that of R. kinnekullensis (see panel D). (F) Anterior head region of the phosphatocopine Hesslandona unisulcata Mü ller, 1982. Note the enlarged coxa of the mandible as an extreme variation of the proximal endite. (G) Postmaxillulary appendage of the cephalocarid Lightiella monniotae Cals and Delamare-Deboutteville, 1970. (H) Maxillula of the mystacocarid Derocheilocaris remanei Delamare-Deboutteville and Chappuis, 1951. Abbreviations: cox, coxa; end, endite; pe, proximal endite; ste, sternum. Repository numbers (see Fig. 2.1 for abbreviations): (A) UB W 255; (B) UB W 380; (C) Spec. 5/YKLP 10846; (D) UB W 86; (F) UB W 381. All figures used with permission of the authors.

49

50

Functional Morphology and Diversity food-gathering apparatus of head and thorax. Its absence in other euarthropods, including Agnostus pisiformis (Mü ller and Walossek 1987, Waloszek et al. 2007), is interpreted as primary lack, while other general appendage features are shared between A. pisiformis and Crustacea s. l., that is, multiannulated exopods on appendages 2 and 3, suggestive of closer alliance. One less obvious feature of Crustacea s. l. concerns the number of endopod podomeres. In the ground pattern of euarthropods, there may have been 9, possibly up to 11. Not least regarding its armament with median spines, this may point to the importance of the endopod in both food gathering and locomotion (possibly walking). The exclusively Paleozoic trilobites had constantly seven endopod podomeres, as also Agnostus pisiformis (Mü ller and Walossek 1987, also for data on trilobite limbs). In crustaceans the maximum number observed is six (Zhang et al. 2007 for Yicaris dianensis), five or six in Cephalocarida depending on the authors (e.g., Jones 1961), or always five in Malacostraca (specializations not counted). The maximum observed in Orsten stem derivatives was originally thought to be five, but maybe six or seven (Haug et al. 2010a). Most fossil and extant taxa have fewer endopod podomeres, at times four or even as few as one portion (which may not be a simple single podomere, however). This may point to a rather early initiation of different uses of endopods, likely in accordance with differences in lifestyles of their carriers. Walkers among malacostracans, for example, have five podomeres, but these are tubular and may have enormous lengths so that in the extreme a limb can be more than a meter long (for a size comparison of large arthropods, see Rudkin et al. 2003, their fig. 5). We cannot, however, follow up this interesting question in more detail at present. Seriality of limbs is also retained in crustaceans principally in all appendages behind the third limb (Fig. 2.3C,D, level with more proximal endites), but these limbs share the proximal endite with the preceding two limbs, and they differ from those in that they have paddle-shaped exopods (Fig. 2.4A). Although partly retained from the euarthropod ground pattern, this indicates the tagmotic break and the different use of anterior and posterior limbs. We must admit that reconstructing the ground pattern status of exopods at this level remains problematic because a multiannulated state (although with inwardly pointing setae, a status not known from other arthropods) is known from a number of Orsten stem derivatives, such as the cambropachycopid Goticaris longispinosa Walossek and Mü ller, 1990 (Haug et al. 2009a; Fig. 2.1A,B) and Martinssonia elongata Mü ller and Walossek, 1986 (Fig. 2.1C,D) (unclear for Cambropachycope clarksoni Walossek and Mü ller, 1990). In M. elongata, the annulation of its exopods is only weakly developed (Mü ller and Walossek 1986b, their fig. 2.4G; see also Haug et al. 2010b). As mentioned earlier, a triangular plate may be set off from the exopods basally, as in Oelandocaris oelandica and Henningsmoenicaris scutula. In both, the distal part forms the main setae-bearing paddle (Stein et al. 2005, 2008, Haug et al. 2010a). This morphology is remarkably similar to what has also been described for a derivative of the stem lineage toward Euchelicerata, Leanchoilia illecebrosa from the Terreneuvian Chengjiang fauna of China (Liu et al. 2007). Whether these two exopod articles are homologous and how they relate to subdivisions developed in other Cambrian euarthropods such as trilobites and naraoiids remain unclear at present. Two noteworthy plesiomorphies have been taken over into the crustacean ground pattern. One is the retention of a fairly rigid basipod with a straight inner rim carrying robust medially pointing spines. This feature is recognizable in two of the Orsten stem derivatives, Oelandocaris oelandica and Henningsmoenicaris scutula (Fig. 2.4A,B). All others have shortened this edge in proximodistal aspect but elongated it in mediolateral aspect to a more humplike or enditelike shape with a central major spine and some flanking spines or setae (e.g., in Goticaris longispinosa; Fig. 2.4C). It seems that this feature may be useful to discriminate within the set of early crustacean taxa and to resolve the early evolutionary lineage of Crustacea in more detail. The other is an often-neglected plesiomorphy retained from Euarthropoda, namely, that the head comprises only four appendage-bearing segments (Maas et al. 2003, Waloszek 2003b)—as in

Evolution of Crustacean Appendages Agnostus pisiformis. When exactly this change to a larger head unit occurred, and why, remains unclear because several of our stem derivatives exhibit this condition. Other taxa have five appendage-bearing head segments (H. scutula and O. oelandica). This cannot be evaluated any further at present, but other characters indicate that the longer head may have been achieved in parallel to the evolutionary development in the remaining crustacean taxa—as much as the number six has been achieved various times in parallel in the different eucrustacean ingroups. The shield cannot be the reason because H. scutula has a large shield and O. oelandica a smaller one. There is also no change in the tagmotic pattern on the ventral side. In summary, crustaceans have retained much of the limb morphology of the euarthropodium, but evolved (1) a new type of exopod that is multiannulated and bears setae along the inner side and (2) a fourth limb element, the proximal endite, that has no counterpart in any other arthropod taxon. Up to the next evolutionary level, the described condition remains stable, except for the occurrence of more proximal endites along the limb series. This also implies that the fourth and fifth appendages resemble trunk appendages—regardless of whether they are already included into the head. Appendages at the Level of Labrophora Major novelties along the crustacean lineage characterize the evolutionary level of Labrophora (Maas et al. 2003, Siveter et al. 2003). As before, most if not all of them are allied with significant changes in the locomotory and feeding apparatus, more specifically the cephalic one (Waloszek et al. 2007). However, only few changes affected the limb morphology. Examples of innovations (autapomorphies) not associated with the appendages include the following: • The appearance of an enormous fleshy “labrum” with slime glands and chemoreceptors on the posterior side of the hypostome (note that labrum and hypostome are both present and not synonymous structures!) • The occurrence of a “sternum” as a fusion product of the sternites of the mandibular and first postmandibular segments (no sternite belonging to the second appendage visible) • A pair of humps on the mandibular sternite, the “paragnaths” • Fine hairs or denticles occurring in rows on the labral flanks and like a carpet on the sternum and paragnaths (for details, see, e.g., in Maas et al. 2003, Waloszek et al. 2007). It is very likely that this, again, did not happen in a single stem species, but up to now we have not discovered more fossil species to be able to split up this set more precisely. Regarding appendage morphology, labrophoran autapomorphies occur, as far as we could detect them in our material, only on the anterior two postantennular limbs but in two important ways: • The proximal endite of the second appendage, from this level on called antenna, is enlarged to a ring-shaped sclerotic structure on the proximal end of the limb, long ago named coxa; its original enditic surface with setae is prolonged and points, due to the special position of the antenna laterally, at the labrum posteromedially around the labral flanks toward the mouth; the coxa now carries the basipod with its median enditic hump and armature, which itself carries the rami. • The proximal endite of the third appendages, now called “mandible,” is also enlarged to a ring-shaped sclerotic structure on the proximal end of the limb; also here this coxa is medially prolonged but points medially because of the position of the mandible lateral to the paragnaths; different from the antenna, the surface of the coxal endite is slightly flattened and tilted against the mouth, a morphology often called gnathobase (Figs. 2.3E, 2.5F; see Fig. 2.6F for an extreme variation).

51

52

Functional Morphology and Diversity

Fig. 2.6. Scanning electron microscopic images and schematic drawings of malacostracan and entomostracan trunk limbs. (A) Pereopod of euphausiid malacostracan, not to scale. In contrast to Entomostraca, the thoracic limbs of Malacostraca do not have endites. (B–D) Postmaxillulary appendages of different entomostracan species. (B) Lightiella monniotae Cals and Delamare-Deboutteville, 1970 (Cephalocarida). Note the two-parted exopod. (C) Yicaris dianensis Zhang et al., 2007. The proximal part of the basipod is concealed. (D) Dala peilertae Mü ller, 1982. Endites are bent toward the reader. (E) Schematic cross section of thorax I (sensu Walossek and Mü ller 1998a) at the level of the first thoracopod of the malacostracan Speonebalia cannoni Bowman, Yager, and Iliffe, 1985, modified after Bowman et al. (1985, their fig. 2a). Each sternite has a median keel for food manipulation (Walossek 1993). (F and G) Cross sections at the level of postmaxillulary appendages of two entomostracan species. Note the large number of endites and the proximal endite, all used for feeding (Walossek 1993). (F) Y. dianensis, the only entomostracan with the unequivocal (high) number of six endopod articles. (G) The branchiopod Rehbachiella kinnekullensis Mü ller, 1983 with the ventral food groove, autapomorph for Branchiopoda and used for feeding (Walossek 1993). Gray shading scheme is as described in Fig. 2.2. Abbreviations: bas, basipod; cox, coxa; en, endopod; ex, exopod; pe, proximal endite. Repository numbers (see Fig. 2.1 for abbreviations): (C) YKLP 10859 (Zhang et al. 2007, their fig. 1i–k); (D) UB W 310. All figures used with permission of the authors.

Often the coxal endite is called a gnathite and is even thought to serve for grinding or cutting. At this early evolutionary stage, nothing of that kind is even remotely present. Equipped with fine, setalike outgrowths distally, it is not capable of biting. This may even be a major misunderstanding of arthropods in general when applying functional terms from other animals, such as jawed vertebrates. An example is the molar surface of gammarid amphipods, which is made of

Evolution of Crustacean Appendages a large number of extremely soft, densely packed outgrowths. The molars act, for example, to destroy the cells of diatoms or plants (Watling 1993), and they improve the ability to hold food (Mayer et al. 2009, their fig. 3E,F; G. Mayer, personal communication). Plesiomorphically, the antenna and mandible are very similar to each other (see Fig. 2.4I) (except for the tilted endite surface of the mandibular coxa), and the spines and setae-bearing enditic protrusions of coxae and basipods served to stuff food into the mouth (Labrophora are in a way “Di-Mandibulata,” as Waloszek 2003b pointed out). Also, the distal elements are little different from the state before; both limbs retain the multiannulated exopod for sweeping. This situation has even been retained in all those eucrustacean taxa with a larval development that includes feeding nauplii/metanauplii (e.g., cephalocarids, mystacocarids, copepods, branchiopods). Slight differences between the two appendages refer to the different positions and orientation along the body axis regarding the mouth opening. While the posteromedially oriented antenna is inserting laterally to the mouth, the mandible inserts posterolaterally, being orientated medially to anteromedially. Also, the more posterior appendages appear to have remained largely unaltered in their gross morphology; that is, they are serial and composed of four elements: proximal endite, basipod, endopod, and (here) paddle-shaped exopod. Even so, all postantennular limbs show another evolutionary novelty, clearly recognizable in our material: • A special arrangement of the spines and setae on the median enditic protrusions of the proximal endite, the basipod, and likewise, all endopodal podomeres. This armature is clearly different from that developed in all known stem derivatives and also all available outgroups. It comprises basically one central spine and a set (crescentic row) of setae flanking the spine anteriorly and posteriorly (e.g., Walossek 1993, Maas et al. 2003, Waloszek 2003b; Fig. 2.5A–C; for more details, see below). Remarkably, at this level of Labrophora, all median structures of the appendages are rather similar to each other, that is, the proximal endite, the basipod protrusion, and the mediodistal humps on the endopodal podomeres. Despite these similarities, the proximal endite is often easily detectable. It is usually larger and better armed than the other enditic protrusions, and it is slightly anteriorly tilted (e.g., Walossek 1993, his plate 2–2). This tilting of the proximal endite against the axis of the limbs can be almost 90°, as in triopsid branchiopods (Fryer 1988, his fig. 102). The enditic origin remains visible even if the proximal endite is modified as in the mandibular coxal median prolongation (e.g., in Figs. 2.4F, 2.5A), the so-called gnathobase—best seen in phosphatocopine larval stages (see Maas et al. 2003, their fig. 65A for the antenna and fig. 66A for the mandible). Another example of extreme modification is that of the maxillulae (or first maxillae) and maxillae (or second maxillae) of eubranchiopods, which are nothing but the retained proximal endites of the otherwise strongly reduced limbs (e.g., Martin and Cash-Clark 1995, their figs. 6, 11A–C, Olesen et al. 2003, their fig. 9E). The entire feeding apparatus of Labrophora may be regarded as a posteriorly open food-path system, with the anterior appendages as much engaged in feeding and locomotion as the posterior set that brought in food from the posterior along the interlimb space. The more specialized appendages were located in the vicinity of the mouth. Here food could be checked (chemosensed at rear of labrum; Waloszek 2003b, his fig. 3B), sorted, and directed into the mouth by means of the strong enditic spines of antenna and mandible. As before, the posterior feeding system works while moving, but the anterior system was decoupled and could operate much more individually. During rotation of the limbs around their basal joints, the limbs produced moving currents by their exopods as much as inwardly directed food flow (even more so by the antennae and mandibles). From this, nutrients were raked off

53

54

Functional Morphology and Diversity and shoveled against the mouth. It is important to note that the basipod, at this level, possesses only one major enditic protrusion and that the median edge was not subdivided into soft lobate endites. Within the Labrophora, phosphatocopines deviate from this in having significantly modified their postantennular limbs in at least two aspects (Fig. 2.3F): • All endopods are basically three segmented (Maas et al. 2003; endopod of antennae and mandibles may be only two segmented in ingroups); each article of the otherwise serial postmandibular limbs is drawn out into long endites (Fig. 2.4E,H) (yet with the triplet of setae as described above). • Within the taxon, the endopods may be further modified from the ground pattern for special feeding purposes as observed from isolated appendages from the middle Cambrian of Australia. Here the endopod is involved in the formation of a kind of gnathic edge together with the coxa and the basipod, comprising more or less a “whole-limb jaw” (Walossek et al. 1993, their fig. 3C). • The mandibular basipod becomes progressively shorter and smaller during ontogeny until it almost disappears, which gives the superficial impression of a coxa carrying the rami. The fate of the basipod was documented at least for Vestrogothia spinata Mü ller, 1964 (Maas et al. 2003, their fig. 59A,C and plate 45A,B, Maas and Waloszek 2005, their fig. 2A,B). Eventually, the basipod can no longer be recognized as a larger subtriangular unit carrying the two rami and is nothing more than a separate setae-bearing endite below the proximal endopodal endite, while the exopod seems to stem directly from the coxa (Fig. 2.4F; see Maas et al. 2003). This final morphology is quite misleading because the basipodal endite seems very much like an endopodal article and gives the appearance of three articles altogether, as in the endopods of all posterior limbs and in the ground pattern of Phosphatocopina. The developmental sequence can be recognized only in intermediate (= younger) stages that possess a narrow ring around the exopodal basis with connection to the median protrusion (Maas et al. 2003, their fig. 59C). Thus, it becomes clear that coxa and basipod do not really fuse. A number of phosphatocopine species also have multiannulated exopods on the postmandibular appendages, for example, Vestrogothia spinata, Falites fala Mü ller, 1964, and Hesslandona neocopina Mü ller, 1964 (Maas et al. 2003). In Phosphatocopina, the exopod morphology is very special, because there are species that change during ontogeny from one morphology, that is, multiannulated exopods on the posterior appendage series with inner setation, to the other morphology, that is, paddle-shaped exopods with marginal setation (Maas et al. 2003 for Hesslandona unisulcata). Phosphatocopina are also a good model for the understanding of the ontogenetic change of the proximal endite into a huge coxa with its median gnathic edge. Also, the armature, still present in the earliest stages, is a nice reference for the recognition of the transition (see Fig. 2.5A). Lastly, Phosphatocopina also demonstrate the retention of the large excavation of the proximal margin of a limb (Fig. 2.4F here referring to the coxal posterior side), as initiated in the euarthropodium. Appendages of the Eucrustacea The ground pattern of Eucrustacea includes the four-element postantennular limb, with coxae on the antenna and mandible as well as proximal endites on all posterior limbs (Figs. 2.2E, 2.3G). Some doubt remains regarding the development of the proximal endite into the coxa in the antenna and mandible. Only a few taxa exhibit the morphogenetic change of a proximal

Evolution of Crustacean Appendages endite to a coxa on the mandible in very early larval stages, such as the Cambrian Rehbachiella kinnekullensis, while the antenna clearly has a coxa from the beginning (Walossek 1993). It is also difficult to determine the ground pattern of phosphatocopines, though it could be demonstrated for one species from the lower Cambrian, Klausmuelleria salopensis Siveter et al., 2003 (Siveter et al. 2003, their text fig. 4). At present, we therefore favor the hypothesis that the morphogenetic switch in the antenna occurred in the labrophoran ground pattern. The only notable change concerning limb morphology in the ground pattern of Eucrustacea seems to have affected the third postantennular or fourth cephalic limb, which from this evolutionary level on is termed maxillula and may represent an autapomorphy of Eucrustacea: • First postmandibular limb is dissimilar in morphology to the preceding and succeeding limbs, functioning as a feeding aid. The word seems is used to express our cautiousness, because the morphology of the maxillula remains unclear. This is because (a) its shape differs between the two sister taxa Malacostraca and Entomostraca (see below; examples given in Walossek and Mü ller 1998a, their fig. 12.11), and (b) the maxillulary shape of these taxa differs significantly from that of the fourth limb of phosphatocopines and that of “stem crustaceans.” Therefore, we cannot reconstruct a ground pattern state for the morphology of the maxillula. We can only construct a ground pattern state for its function because function is generally the same in all ingroup eucrustacean taxa. What seems clear is only that the maxillula was already reduced in size compared to that of the more posterior limbs. This situation can be seen in all known Cambrian Orsten taxa referred to Eucrustacea, for example, Bredocaris admirabilis (Mü ller and Walossek 1988), Rehbachiella kinnekullensis (Walossek 1993), and Yicaris dianensis (Zhang et al. 2007). (The same seems to be the case in Dala peilertae and Walossekia quinquespinosa Mü ller, 1983, but detailed descriptions of these are still under way.) We also have difficulties reconstructing the morphology of the fourth postantennular or fifth cephalic limb, traditionally called the maxilla. We can only clearly state that it was not a “mouthpart”—a still repeated misunderstanding of crustacean morphology. It was included in the head, and it was acting in accord with the trunk limbs, a plesiomorphic trait. So this holds only for its feeding and locomotory duties in the ground pattern of Eucrustacea. Clearly, the maxilla at that level did not have the morphology of previous levels, as we learned for other limbs and for earlier nodes. And it is also not a mixture of the morphology of ingroup maxillae, but the maxillae have adopted the specific trunk-limb morphology of one of the two sister taxa within Eucrustacea. As discussed below, these limbs are also very differently developed, in terms of specific changes in the locomotory and feeding apparatus and tagmotic changes, for example, in the Malacostraca. For both maxilla and the trunk limbs, we cannot state that the malacostracan morphology should be plesiomorphic, but we cannot state this for the entomostracan morphology either. Likewise, it is difficult to argue that entomostracans are a paraphylum “along the evolutionary lineage toward Malacostraca” because this would intimately violate the data around limb morphology. The retention of the maxilla within the posterior limb system appears plesiomorphic, while the specific similarity in shape, flatness, and softness of the maxillula and maxilla in malacostracans only adds to the large number of other autapomorphies validating this monophylum. The condition that the maxilla has a trunk-limb morphology is developed in several Cambrian fossil eucrustacean taxa such as Bredocaris admirabilis, Dala peilertae, Rehbachiella kinnekullensis, Walossekia quinquespinosa, and Yicaris dianensis, and in extant cephalocarids (e.g., Sanders 1957, 1963). It may hence be likewise interpreted as a specific feature of Entomostraca and not as an ancient trait.

55

56

Functional Morphology and Diversity All more posterior appendages remain basically unaltered in their gross morphology. They are serially homonomous and composed of the proximal endite, the basipod, the endopod, and the paddle-shaped exopod (= euarthropodium + proximal endite). It appears likely that epipods on these limbs belong to the ground pattern of Eucrustacea, but their number remains uncertain, being three in Entomostraca and no more than two in Malacostraca. Their exact origin also remains unclear at the moment (Maas et al. 2009). (For a discussion of the second autapomorphy, the ontogeny via a short-segmented [ortho]nauplius, see Maas et al. [2003].) Lastly, it must be stressed that coxa and proximal endite exclude each other because a coxa, as far as we can state at present, originates from this endite below the basipod. The joint between coxa and basipod originates from the fact that the proximal endite lies embedded within the proximal limb joint membrane. When the coxa develops, the distal portion of the membrane is simply retained between the final sclerotic stem portions. Different Evolution of Limb Morphologies in Entomostraca and Malacostraca The appendages remain more or less unchanged in the ground pattern of Eucrustacea, but substantial further evolutionary modifications of them occurred in the evolutionary lineages of the two sister taxa. This is fully understandable since all branchings mentioned so far, and possibly even a few more already on the lines of the two sister taxa, had already occurred in the Cambrian. The others had, accordingly, another 500 million years to make changes. Significant differences in appendage morphology already in the ground patterns of Entomostraca (Fig. 2.3H) and Malacostraca (Fig. 2.3I) indicate that, as in the maxillulae and maxillae, it is often difficult to reconstruct well-founded ground pattern states, because the differences between the taxa to be compared are not traceable down to a common ancestral morphology. Up to now, only the evolution of Entomostraca has been documented (as we think) by well-preserved Cambrian fossils in Orsten-type preservation, for example, Yicaris dianensis (Zhang et al. 2007, Fig. 2.1G,H), Bredocaris admirabilis (Mü ller and Walossek 1988), and Rehbachiella kinnekullensis (Walossek 1993; see Fig. 2.5D,E for comparison with an extant branchiopod). The early phase of malacostracan evolution, in contrast, is still completely unknown. The earliest reliable malacostracan fossils occur in the Lower Ordovician, and even the status of these so-called phyllocarids must be considered uncertain at best: living representatives of phyllocarids, having leaf-shaped limbs in the anterior thoracic portion (thorax I after Walossek and Mü ller 1998a), are indeed very different from the well-sclerotized large fossil forms possibly having rather stenopodial limbs. This bias may have led to the impression of a “well-defined” taxon Malacostraca (admittedly the two-section thorax region present in all living taxa is a striking feature), while entomostracans seem to express—another misunderstanding—“morphological instability as an expression of plesiomorphy,” because their ingroup taxa exhibit very different morphologies. In fact, this is exactly their “trick.” The high number of autapomorphies of the Malacostraca in its ground pattern means nothing but the result of a long, though unknown, evolutionary lineage in which these characters have been accumulated. These are assigned now to the last common ancestor of all descendants with living representatives. The morphological conservatism of Malacostraca may then just be the retention of plesiomorphies, not expression of evolutionary success. The bias also results, in our opinion, in the superficial view that Entomostraca should have retained more plesiomorphic traits than the malacostracans (the often-used terms lower and higher crustaceans reflect this interpretation) and hence should not be monophyletic. Indeed, some body parts retained plesiomorphic traits, such as the anterior appendages. Antennular and antennal morphologies of malacostracans are easily identified as their evolutionary novelties,

Evolution of Crustacean Appendages while Entomostraca apparently retained a plesiomorphic appearance of these two appendages. However, all further posterior appendages underwent significant evolutionary changes in both lineages, and in both, the morphology of these limbs deviates from the ground pattern of Eucrustacea or Labrophora. This can even be stated for the mandibles, which lose their basipod plus the rami during the ontogeny in entomostracans, and for the maxillulae and maxillae as well. Accordingly, many features in the stem species of Entomostraca and Malacostraca, as far as they can be reconstructed, can be well understood as modifications in accordance with adaptations to specific life habits achieved in each group separately. This makes any evolutionary transition from the one to the other morphology unlikely; that is, in these cases, the features represent autapomorphic states of the two taxa. Therefore, we view the assumption of a paraphyly of Entomostraca, as commonly proposed (see Waloszek 2003b for discussion), to be improbable and less parsimonious and uphold the hypothesis of their monophyly. One area of evidence for this is the specific morphology of appendages, which is detailed next. Anterior Cephalic Appendages The antennula of adult malacostracans comprises a prominent tubular basal part and two distal flagella (three in some taxa) and has sensory function. Only in the nauplii of Euphausiacea and Dendrobranchiata (the only malacostracan taxa with free-living nauplius larvae) is the antennula uniramous and limblike as in their ancestors; it even bears a number of long swimming setae on its distal end (e.g., Hirota et al. 1984a, 1984b, Kidd 1991). Thus, it appears that in these early larvae the antennula recalls at least part of its original state. The original functions of the antennulae, swimming and helping in sweeping in food particles, may have been abandoned early during evolution and ontogeny. Nauplii of these groups only swim and do not feed; having no feeding aids at all may even point to a very early loss and change to lecithotrophy. This loss may have been in line with the modification of the other appendages taking these functions over later during development, that is, from the protozoea onward. Similarly, a plesiomorphy exposed in larvae is recognizable in the antenna of Euphausiacea and Dendrobranchiata (e.g., Cockcroft 1985 for the penaeid Macropetasma africanum Balss, 1913, Maas and Waloszek 2001b for the Antarctic krill Euphausia superba Dana, 1852). The morphology is very similar to that of feeding entomostracan nauplii, but those of malacostracan taxa lack any median feeding structures because they do not feed. Later in ontogeny the antennal endopod is transformed into a sensorial “feeler” with a long multiannulated flagellum, and the exopod becomes the paddle-shaped “scaphocerite” (e.g., Maas and Waloszek 2001b for E. superba as an example of euphausiids). Lastly, the naupliar mandible of malacostracans with early larvae is very similar to those of early larval entomostracans, including coxa, basipod, and two rami. The retention of the parts distal to the coxa, called palp in later stages and the adult, may be plesiomorphic, but this uniramous palp is only remotely similar to the original condition of the Eucrustacea. The adult palp consists of three articles interpreted as basipod and the bipartite endopod (see, e.g., Olesen and Walossek 2000). The morphology of the palp is remarkably similar in all known ingroup taxa and clearly represents an autapomorphy of Malacostraca. In Cephalocarida, Branchiopoda, and, as the ontogenetic path suggests, most likely in a number of Orsten species, such as Yicaris dianensis, Rehbachiella kinnekullensis, and Walossekia quinquespinosa, all parts of the mandible distal to the coxa are lost during ontogeny. This condition is considered an autapomorphy in the ground pattern of Entomostraca. We thus understand the presence of a palpus in Maxillopoda as resulting from a pedomorphic event in their evolutionary history (e.g., Newman 1983, Walossek and Mü ller 1998a, Waloszek 2003b).

57

58

Functional Morphology and Diversity The situation of the exopods of antennae and mandibles is more complicated. In entomostracans, they may retain their multiannulated morphology (e.g., cephalocarids, some maxillopodans), be modified (some phyllopod branchiopods), or become lost (some maxillopodans, anostracan branchiopods). Immature stages retain the original morphology in all taxa. In Malacostraca, the larvae—where developed—have multiannulated antennal exopods, but in adults the exopod has at most two parts and is flattened, that is, the scalelike scaphocerite. Furthermore, these appendages can undergo high modification or complete loss in particular ingroups, but mostly entomostracans. Examples are the tantulocarids (all cephalic appendages), facetotectans (antenna, mandible), and cirripedes (antenna). Not only do the maxillulae of Malacostraca and Entomostraca differ significantly in their morphologies, but they also differ from the equivalent limb in the Phosphatocopina and stem derivatives. The homotope, that is, the positional homologue, of the maxillula in Phosphatocopina is still serial to all more posterior appendages and consists of a lobate proximal endite, a basipod with one median endite, and the rami (Maas et al. 2003). In both Malacostraca and Entomostraca, the maxillula is developed as a further “mouthpart” and differs in morphology from the more posterior appendages. The malacostracan maxillula is uniformly rather thin and has a slight C-shape (concave anteriorly) and a proximal endite modified into a coxa. Both coxa and basipod are medially drawn out into one blade-shaped enditic process with a rim of spines or setae. Both maxillula and maxilla are similar and act as a unit behind the mandible, closing the food chamber posteriorly. In the ground pattern, the entomostracan maxillula has a proximal endite and three more setose endites along the median edge of the basipod (not one as in phosphatocopines or stem derivatives). This state is realized in copepods, in mystacocarids, and in cephalocarids during the larval phase. In the stem-branchiopod Rehbachiella kinnekullensis it occurs until the latest instar stage known, and it is also developed in, for example, the Orsten taxa Yicaris dianensis (Zhang et al. 2007), Walossekia quinquespinosa (unpublished), and Bredocaris admirabilis (Mü ller and Walossek 1988). Crown-group branchiopods (Eubranchiopoda) have reduced maxillulae and maxillae retaining only their proximal endites (see Walossek 1993). In summary, the maxillulae have a different fate from that of the maxillae in entomostracans but are both fully integrated and acting together in a cephalic feeding system in malacostracans. There they are rather similar to each other, and both have a coxal portion below the basipod. Only in cephalocarids among the living entomostracans does the maxillula serve as a single “mouthpart,” while the maxilla is a trunk limb. In all other taxa, the situation is different. In the extreme, maxillula and maxilla are almost lost (eubranchiopods) or completely lost (parasitic tantulocarids). Remarkably, the maxillae retain a paddle-shaped exopod in all taxa where it remains developed—and there is no exception throughout. Postmaxillulary Appendages In Malacostraca the maxilla is much flattened in anteroposterior aspect, in this way looking much like the maxillula. It is also C-shaped (anteriorly concave), and its proximal endite forms a coxa. Coxa and basipod are medially drawn out into bladelike spine- to setae-bearing protrusions, but in contrast to the maxillula, they have a median cleft, so they appear divided in two. Both maxillulae and maxillae have a very soft and fragile appearance and are considerably shorter than all subsequent limbs of the trunk. These postcephalic trunk appendages of Malacostraca not only are different from the maxillae but also cannot easily be discussed because (1) they occur in two very distinctive sets, and (2) they are rather different in the two sister taxa Phyllocarida and Eumalacostraca. Only the first eight thoracic limbs (of thorax I according to Walossek and Müller 1998a) have a limb stem made of a coxal and basipod portion, but the so-called pleopods (limbs of thorax II sensu Walossek and Müller 1998a) have neither a coxa nor a proximal endite.

Evolution of Crustacean Appendages The anterior eight thoracopods of phyllocarids are flattened and appear very superficially similar to the filter appendages of Branchiopoda—though virtually all details are different, the feeding system is a closed one, phyllocarids do not filter feed in the strict sense (see Walossek 1993 for detailed comparisons), and they have fairly large coxae and basipods, from which an elongate endopod and a paddle-shaped exopod arise. The first four pleopods have a rod-shaped basal part, which carries two rami of equal size immediately on its top (e.g., Olesen and Walossek 2000, their fig. 7b). Pleopods 5 and 6 are small and consist only of an elongated lobe. The eight anterior thoracopods of Eumalacostraca have fairly short coxae and basipods but likewise elongate endopods. The exopods are nowhere prominent except in euphausiids and may be small paddles (e.g., Maas and Waloszek 2001b, their fig. 11) or multiannulated rods stemming from a proximal peduncle piece (e.g., Neil et al. 1976, their figs. 3a, 5c, 9)—the sloping articulation area of the exopod recalls the original morphology taken over from the euarthropodium. As in Phyllocarida, eumalacostracan pleopods lack a coxa or a proximal endite, and also their rami rest on top of the stem portion. The variety of morphologies of the rami is, however, large. The maxilla in the ground pattern of Entomostraca is, on the other hand, very different from the morphology in the maxilla or trunk limbs of malacostracans—and it matches that of the series of posterior limbs (Fig. 2.3H). In size and morphology, it virtually equals the trunk limbs at this stage. Its proximal endite is prominent, possibly the largest of all postmaxillulary limbs, and in its setation grossly similar to that developed in Phosphatocopina. The basipod is a large subrectangular element, longer than wide. Its mediodistal extension forms a kind of socket for the transition to the endopod and a sloping distal outer margin from which the exopod is articulated. Its straight median edge is drawn out into several lobate setiferous endites (Waloszek 2003b). Their armature resembles that of the proximal endite but is progressively less elaborate. The basipod body is fairly f leshy and little sclerotized, with the exception of the lateral edge proximal to the exopod insertion, which is slightly better sclerotized but interrupted by two furrows, giving this side a tripartite appearance (Fig. 2.6F,G). For the extant cephalocarids (see Fig. 2.6B), it is clear that the proximal endite and the endites along the inner rim of the basipod have muscles internally and can be moved accordingly (Sanders 1963, Hessler 1964). It is possible that, as the preservation of endites in Rehbachiella kinnekullensis and other such Orsten eucrustaceans suggests, the movability was a particular feature of these endites already in the ground pattern of Entomostraca and an additional autapomorphy. In those taxa where the basipod rims are more or less straight, more rigid, and adorned with rows of setae, this would be a secondary adaptation in line with somewhat modified food intake. The endopod appears to be the direct continuation of the basipod. Not only do its proximal podomeres match the shape of the distal enditic protrusions of the basipod, but also the setation pattern is continued, again being less and less developed toward the distal end. The endopod is maximally six segmented in Eucrustacea and ends in a much smaller distal conelike to caplike piece with a terminal tuft of setae or spines. The exopod is paddle shaped with a marginal row of long setae. This entire morphology matches that of the subsequent limbs and is developed in this way not only in certain Orsten taxa (Yicaris dianensis, Rehbachiella kinnekullensis, Dala peilertae, and Bredocaris admirabilis; see Mü ller 1983, Mü ller and Walossek 1985, 1988, Walossek 1993, Zhang et al. 2007) but also in the extant Cephalocarida (e.g., Carcupino et al. 2006, their fig. 4A). Although this type of appendage does not match that of any of the known stem derivatives, phosphatocopines, or malacostracans (which have a coxa in the maxilla and anterior eight thoracopods), we can find here a mixture of plesiomorphies and apomorphies present on a single structure. In terms of seriality, the maxilla in the ground pattern of entomostracans (and

59

60

Functional Morphology and Diversity ingroups!) is similar, that is, homonomous, to the more posterior limbs. In other words, it is not included into a specific cephalic feeding and locomotory apparatus, so it can be termed maxilla in the strict sense. Superimposed on this plesiomorphy are other, autapomorphic features, such as a subdivision of the median edge of the basipod into several setae-bearing lobate endites (up to seven in Yicaris dianensis; see Zhang et al. 2007). The same applies to the more posterior appendages, the thoracopods, in Entomostraca (up to eight basipodal endites in Yicaris dianensis; see Zhang et al. 2007). Furthermore, the otherwise transversely inserting basipods of all postmaxillulary appendages are elongated in proximodistal and lateral axis and are rather f leshy and slightly C-shaped curved backward, so the posterior side is concave. In cephalocarids this special morphology forms narrow chambers between the limbs in the row. These are opened and closed during the moving cycle (heterochronal beat) of the limbs pressing out water or sucking it in during opening. The system of pumping chambers is known from these and other eucrustaceans but is studied in more detail in anostracan branchiopods, which filter feed (Fryer 1983), a habit not applicable to cephalocarids (Walossek 1993). Suspension feeding is widespread among the Crustacea and is not discussed here any further (for more details, see chapter 8). Yet, for both taxa it is known that the system has two functions: feeding and locomotion. Moreover, it requires the open system, as noted for the Labrophora level. Maxillae forming a chamber as in Malacostraca is, however, a very different operational system. Also, maxillae and a so-called maxilliped operating as a closed feeding system, as in copepods, can be set off, but since this condition occurs in ingroup entomostracans, this system is derived from the open entomostracan type. A similar system with fleshy basipods has also been reconstructed for Orsten entomostracans such as Rehbachiella kinnekullensis and Yicaris dianensis, so it may have operated similarly. The differences in detail (setae, subsetules, endite form, etc.) not only point to differences in feeding and locomotion but also may indicate different development in different evolutionary lineages. The basipodal endites of all postantennular appendages are, as mentioned above, equipped with setae and spines. These occur in a pattern of three different rows of spines/setae, similar to that of Phosphatocopina (ground pattern of Labrophora). However, while the anterior and posterior sets of setae on the enditic protrusions are still undifferentiated in Phosphatocopina (Fig. 2.5A), in Entomostraca the setae are differentiated both in structure and arrangement into anterior retention setae with rows of more backward oriented setulae, median stronger spines, and posterior setae, which have rows of anteriorly pointing setulae (Fig. 2.5B,C), each set forming a small basket. The specific morphology and the so-called sucking chambers between the limbs enable the animals to feed and locomote at the same time using the entire postmaxillulary apparatus (Walossek 1993), as seen today in cephalocarids and branchiopods. This morphology was most likely modified again in Maxillopoda. The tripartite set of setae develops ontogenetically through a two-part set, as exemplified by Rehbachiella kinnekullensis (Walossek 1993). This ontogenetic change from a two-row to a three-row system can also be observed in Yicaris dianensis (Zhang et al. 2007). The presence of a two-row system in Maxillopoda, such as in Bredocaris admirabilis (Mü ller and Walossek 1988), can thus be understood as another effect of pedomorphosis that affected Maxillopoda (Newman 1983, Walossek 1993). None of these special changes of the basipods can be found in Malacostraca. There the maxilla is indeed developed as a specialized mouthpart in the ground pattern of this taxon, more or less an aid to close the oral chamber. The maxilla is extremely f lat, similar to the maxillula, and the proximal endite is also enlarged to form a coxa. Coxa and basipod are medially drawn out into two (not just one) bladelike protrusions each. In both maxillula and maxilla,

Evolution of Crustacean Appendages the endopods are most likely subdivided into at least three articles in the ground pattern of Malacostraca, as exemplified by the euphausiacean Bentheuphausia amblyops G.O. Sars, 1883 (see Maas and Waloszek 2001b). These limbs differ from those of the Entomostraca also in that they surround the mandibles like hands held over the mouth. Accordingly, this apparatus is a closed system, whereas that of entomostracans is an open system. Basipods of more posterior limbs in Malacostraca may carry setation but do not have lobate endites (Fig. 2.6A,E). It has to be noted here that the appendages of the Silurian fossil phyllocarid Cinerocaris magnifica Briggs et al., 2004 have been interpreted as possibly possessing endites (Briggs et al. 2004). Unfortunately, the appendages of this species have been presented as two-dimensional line drawings (Briggs et al. 2004, their fig. 2i) and not, as usual for this preservational type, as three-dimensional reconstructions (see, e.g., Siveter et al. 2007). It is therefore impossible to verify the presence of any enditic subdivision of the basipods of trunk limbs in Malacostraca. Consequently, based on the existing data, a proximodistally elongated basipod with lobate setae-bearing enditic subdivisions of the narrow median edge is regarded here as an autapomorphy of Entomostraca (Fig. 2.6B–D,F,G; Waloszek 2003b). Malacostraca possess further specializations on their postmaxillary appendages, namely, in their division into two series, eight belonging to the first part of the thorax (thorax I sensu Walossek and Mü ller 1998a) and six so-called pleopods belonging to the second part of the thorax (thorax II or pleon) in the ground pattern. Only thoracopods 1–8 appear to bear a true coxa; that is, the proximal endite is laterally enlarged to form a complete enclosed sclerotized ring (Waloszek 2003b). The presence or absence of a coxa or proximal endite could not be shown with certainty in any malacostracan so far. Another evolutionary modification that cannot be followed up here in more detail, however, is the specialization of one or even more limbs as mouthparts, so-called maxillipeds. This feature, possibly even convergently developed, is a gradual process and does not change the general functionality of the closed oral chamber, but just adds the next limb—for example, as a food grasper or holder—to form a larger unit. The trunk limbs progressively lose their ability to provide assistance in food intake, a trait that is still recognizable in phyllocarids and at least distally in the case of the basket-feeding euphausiids (Hamner 1988). It should be noted that similar strategies occur in entomostracans, where a closed oral feeding chamber including maxillipeds exists, and the remaining thoracopods are devoid of feeding structures and function. Summing up regarding limb morphology, the following features are interpreted as autapomorphies of Malacostraca: • • • •

Mandible with a typical tripartite palp (basipod and two endopod elements) Maxillula with a coxa; coxa and basipod each with a bladelike endite Maxilla with a coxa; coxa and basipod with two endites each Trunk limb series divided into two special series (thorax I and II of Walossek and Mü ller 1998a) of eight anterior and six posterior pairs of appendages; appendages of the anterior series with coxae

Limb-related autapomorphies of Entomostraca include the following: • Mandible loses palp late in ontogeny • Maxillula with proximal endite and with basipod bearing three endites • Basipod of “maxilla” (not specialized as a mouthpart) and all trunk limbs elongated in proximodistal axis of the limb, medially equipped with a series of endites (seven to eight originally?); basipod laterally subdivided into three major parts; setation on the endites organized in three sets of setae specialized for sorting of food particles anteriorly and retention setae posteriorly (orientation of subsetules different)

61

62

Functional Morphology and Diversity

VARIATIONS ON A THEME: LIMB MODIF ICATIONS The two rami of an appendage, endopod and exopod, are modified and specialized in many ways in the various eucrustacean ingroup taxa; in the extreme, they are completely lost. In the following we give a number of examples for different morphological variations of the two rami. This is, however, a limited set since the morphological variation is very extensive within such a large taxon as the Crustacea. Endopod Variations The crustacean endopod plesiomorphically has a simple tubelike shape with setae or spines on mediodistal humps along the inner margin (one such hump per podomere) and a setiferous to spine-bearing tip, possibly an adaptation for food gathering and locomotion (much retained from the euarthropod ground pattern; see above). Alterations occur in various ways. On the one hand, the endopod can be, very rarely, partially reduced to a simple bulbous structure within Eubranchiopoda (Olesen et al. 2001). (Another example outside Crustacea is the endopod of the second head appendage of Agnostus pisiformis [see Mü ller and Walossek 1987, their fig. 6B and plates 12.5, 16.1].) Reduction in size of the endopods and/or of their segmentation in living branchiopods led some authors to (among other errors) misidentify the basipodal endites as endopod podomeres of the according appendage (e.g., Wehner and Gehring 1995, caused by Preuss 1957; for correction, see Walossek 1993). On the other hand, the endopod may become elongated by its podomeres, the number of which stays the same. This occurs, for example, in the thoracopods 1–8 of Eumalacostraca, and there even the whole limb may become stenopodous by parallel reduction of the coxa and basipod as well as the reduction or even loss of the exopod (walking legs). This is also independent of the numbers in toto: an endopod made of one or two portions may be as large as others with more portions. Because of this it is extremely difficult to identify functional or evolutionary adaptations. As hinted at before, each limb has to be viewed separately and in the context of the other limbs. Another specialization is that endopod tips may become subchelate or chelate in various lineages, for example, in Stomatopoda (Morgan and Goy 1987), Peracarida (tanaidaceans or amphipods), and Decapoda (Richter and Scholtz 2001). Also within Entomostraca, endopods with a subchelate tip may be found, such as in the form of claspers of males to attach to the females—examples are the first trunk limbs of certain diplostracans (e.g., Olesen et al. 1996) or the antenna of Anostraca (e.g., Dumont and Negrea 2002, their fig. 130I). It is apparently evolutionarily possible even to change the endopod morphology from one extreme (partially reduced endopod) to the other (elongated stenopodous endopod), as proven by the raptorial water fleas within Cladocera. In the ground pattern of Cladocera, the thoracic endopods are, as in the ground pattern of Eubranchiopoda, simple undivided bulbs (Olesen 2004, 2007). Within Onychopoda and Leptodora kindtii Focke, 1844, the whole anterior thoracic appendages are elongated into multisegmented stenopodia to fulfill raptorial functions (Olesen et al. 2001, 2003; Fig. 2.7F). The endopod may also become a paddle-shaped structure for swimming (see “Symmetries in the Morphology of the Rami,” below). Although partial reduction is possible, a complete loss of the endopod is quite unusual within eucrustaceans. Often only the loss of the entire appendage also permits the endopod to be missing as in the pleopods of various interstitial peracarids, or the loss of the distal part of the appendage including the basipod such as in branchiopod mandibles. Investigation of the ontogeny of, for example, the Cambrian branchiopod Rehbachiella kinnekullensis, in which the mandible is complete in early stages (Walossek 1993, his plate 4–1), reveals that the distal part,

Evolution of Crustacean Appendages

Fig. 2.7. Variations of biramous crustacean limbs. (A and B) The branchiopod Lepidocaris rhyniensis Scourfield, 1926. Images show a maximum intensity projection; that is, images of different focal planes are combined (for details, see Haug et al. 2009b). Arrows indicate exopods of adjacent limbs. (A) Anterior postmandibular appendage in anterior view. Exopod and endopod differ strongly in their morphology and in their insertion angle and position on the basipod. (B) Posterior postmandibular appendage in posterior view. Exopod and endopod are rather symmetric morphologically and in their insertion position and angle on the basipod (“copepod-like”). (C) Thoracopod of Lepas sp. Linnaeus, 1758 (Cirripedia). Endopod and exopod (cirri) are morphologically indistinguishable (arrows); only their position in the living animal shows their identity. Image not to scale. (D) Tail fan of the Jurassic stomatopod Sculda pennata Mü nster, 1840. The uropodal exopod is a simple undivided paddle. (E) Tail fan of the extant stomatopod Squilla mantis Linnaeus, 1758. The uropodal exopod is divided into a distal paddle and a proximal article. (F) Ventral view on a late embryo of the phyllopod Leptodora kindtii Focke, 1844. Arrows mark the second and third thoracopods, developed as stenopodous limbs with median setation, whereas the ground pattern of phyllopod branchiopods is characterized by leaf-shaped limbs. (G) Second antenna of the branchiopod Macrothrix laticornis Jurine, 1820. Exopod and endopod are symmetrically composed of three tubular elements. (H) Thoracopod of an undetermined copepod. Endopod and exopod are symmetrically composed of three setose anteroposteriorly flattened elements. Image not to scale. Abbreviations: bas, basipod; cox, coxa; dex, distal paddle of exopod; en, endopod; ex, exopod; pex, proximal article of exopod. Repository numbers (see Fig. 2.1 for abbreviations): (A and B) BM 25698; (D) MB.A.669.

63

64

Functional Morphology and Diversity the “palp,” of the appendage becomes progressively smaller. Finally, a small round scar marks the original insertion of the palp, but the appendage consists solely of the coxa (Walossek 1993, his plate 17–4). One of the rare examples of complete endopod loss in Eucrustacea is the degeneration of the posterior two thoracopods within Euphausiacea. The gill in these two appendages is the most prominent structure. A coxa-basipod subdivision is not identifiable; a uniramous structure with marginal setation is identified as exopod, while an endopod is lacking (see, e.g., Maas and Waloszek 2001b, their fig. 13). Exopod Variations The exopods, often referred to as the swimming branches (e.g., Calman 1909), also exhibit a wide variety of morphologies in the different eucrustacean taxa. As mentioned above, in Eumalacostraca, the antennal exopod is a two-part paddle that is reduced even further to a simple undivided paddle in Caridoida (Richter and Scholtz 2001; caridoids are the sister group to stomatopods). While exopods are paddle shaped in the maxillulae and maxillae where developed, they are multiannulated on the first eight thoracopods of various taxa within the Eumalacostraca. There the exopod is highly movable and is used for locomotion—for example, in larvae of lobsters (Neil et al. 1976) or in mysid peracaridans, where it may move at high speed to keep the animal in position like a helicopter (the German name Schwebegarnele refers to this). In certain entomostracan taxa, the exopod is subdivided into several articles, often three, as in the trunk limbs of Copepoda and Remipedia (Itô 1982, 1989), but this is unlikely to be a hint to relationships but merely a functional enforcement to achieve a more f lexible paddle. Another example of the appearance of an additional joint in an exopod is the uropodal exopod of some Eumalacostraca, for example, the Stomatopoda. All extant species of Stomatopoda have a bipartite uropodal exopod (Fig. 2.7E), but a number of fossil species demonstrate that this condition is most parsimoniously derived from a simple undivided exopod (Schram 2007; Fig. 2.7D). The exopod has become more subject to loss than the endopod, both during evolution and during ontogeny. This is most evident in all land-living bottom-walking crustaceans, as well as in other euarthropods such as arachnids, myriapods, and insects. But this also holds true for limbs of benthic walkers in the marine environment such as thoracopods 4–8 of Eureptantia among the decapod Malacostraca (not the anterior thoracopods 1–3, which are modified into the maxillipeds, specialized limbs that support feeding) or all isopods, in both cases an autapomorphy of these taxa, or outside crustaceans in the pantopod Chelicerata (already observable in fossils from the Devonian; Bergström et al. 1980). Exopod loss in nonlocomotive limbs can be recognized, for example, on the mandibles of land-based isopod peracarids among the Malacostraca (so clearly convergent to Entomostraca) or on the maxillulae and maxillae of Mystacocarida (species of the taxon Ctenocheilocaris Renaud-Mornant, 1976 also lack the exopod on the maxilliped) and Copepoda (unclear if this hints at a close relationship). We consider these true losses, not a lack of subdivision of exopod and endopod. The idea of the loss of exopods through a lack of subdivision of exo- and endopod (and thus the remaining “endopod” being a homologue of former endopod and exopod) has been concluded from the results of a cell lineage investigation of an amphipod (Wolff and Scholtz 2008). It appears to be unfounded to expand this putative mechanism, which is interpreted to exist in a highly derived ingroup malacostracan taxon, to all Crustacea, or even Arthropoda. This also contradicts all data accumulated that depict the evolutionary path from the lobopodium to the arthropodium and euarthropodium and eventually the crustacean limbs, as we explained it here. Even though the cell lineage data may indicate that the walking limbs of the pereopods are the result of “undivided” branches, the structural similarities to the endopods of other malacostracans and the endopod origin from the distal part of

Evolution of Crustacean Appendages the limb stem in Arthropoda s. str. clearly reject this assumption (for discussion, see also Boxshall and Jaume 2009). Other Crustacea also gain uniramous appendages through loss of exopods. In the mystacocarid species of the taxon Derocheilocaris Pennak and Zinn, 1943, for example, the maxillula and maxilla appear very similar to the maxilliped—except that the maxilliped possesses an exopod. Therefore, it appears to be much more parsimonious to assume a simple loss of the exopod in maxillula and maxilla than to homologize their endopods with both rami of the maxilliped. Symmetries in the Morphology of the Rami Despite their clearly different phylogenetic origin, the two rami, endopod and exopod, can be almost identical or symmetrical in certain eucrustacean taxa. There they even serve the same functions. Examples of such symmetrization of rami can be found in various taxa and on different appendages—clearly an example of convergence. Some examples of this are the “cirri” (endopod and exopod) of the thoracopods of adult barnacles (Fig. 2.7C) and, within the Branchiopoda, the antennae of Diplostraca. These bear two very symmetric rami, best known probably from Cladocera (= “branched horns”; i.e., the taxon name established by Pierre André Latreille in 1829 even refers to this symmetrization), including one of the standard laboratory organisms in student courses, Daphnia pulex De Geer, 1778. Both endopod and exopod comprise three articles, with the proximal articles carrying a single median seta each and the terminal article carrying a number of setae concentrated medially (Fig. 2.7G). All these similar-appearing setae are long and setulose and mainly serve as swimming devices. Another example of symmetrization of rami is Lepidocaris rhyniensis Scourfield, 1926, a putatively branchiopod species from the famous Early Devonian Rhynie Chert lagerstätte in Scotland (Scourfield 1926). The anterior trunk appendages exhibit clearly differing morphologies in the endopod and the exopod (Fig. 2.7A). But in the series of posterior trunk appendages, the so-called copepod-like appendages, both rami appear very similar. They are developed as simple paddles with seven to nine setae along their whole margins (Fig. 2.7B). Symmetrization in this case affects not only the shape of the two rami but also their insertion at the basipod— both rami insert distally. In the anterior trunk appendages, the exopods insert laterodistally (Scourfield 1926). The morphology of these posterior limbs has been interpreted as an adaptation to swimming. As the terminology in Lepidocaris rhyniensis (“copepod-like” appendages) already indicates, copepods also have symmetrical rami on the posterior trunk limbs. Unlike in L. rhyniensis, the endopod and exopod are not simple paddles but are more elongated and comprise three articles (Fig. 2.7H). Even so, endopod and exopod have a very similar appearance. A comparable morphology of the two rami as symmetric elongated paddles with three articles on the trunk appendages is also found in Remipedia (Itô 1989). Also, in various malacostracan taxa, appendages with symmetrical rami occur. The pleopods of Eumalacostraca are one example. In Stomatopoda the endopods and exopods of the pleopods (limbs of thorax II sensu Walossek and Mü ller 1998a) are developed as symmetrical paddles bearing a large number of setae along the whole margin. Yet, the symmetry is partly interrupted, as the exopod carries gills on its anterior side (Morgan and Goy 1987, Maas et al. 2009). Seriality of the Limbs and Tagmatization Changes to the appendage morphologies can affect single appendages or complete series and can be coupled to tagmatization pattern changes, but need not. For example, the inclusion of further appendages into a specialized feeding apparatus is not necessarily linked to the inclusion of the corresponding segments into the head tagma. There are, on the other hand, examples

65

66

Functional Morphology and Diversity of the inclusion of segments into the head tagma with unmodified appendages, that is, appendages that look exactly like trunk appendages, as it is reconstructed for the eucrustacean ground pattern and also exhibited by extant cephalocarids. Also, the reverse case can be found, where an appendage is distinctly evolved into a specialized feeding apparatus, while the segment is dorsally clearly set off from the head tagma. This can be seen in the Copepodoida. The autapomorphy of this taxon comprising, in our view, the Copepoda, the tiny interstitial Mystacocarida and the Cambrian Skaracarida is a special cephalothoracic feeding apparatus with one maxilliped involved. The maxilliped segment is incorporated in the cephalothorax in copepods; that is, its dorsal surface is included into the shield. In Skaracarida and Mystacocarida, the segment is set off dorsally: it is movable against the head (Walossek and Mü ller 1998b). Stomatopoda possess a feeding apparatus, which involves five postmaxillary maxillipeds. The posterior four maxillipeds are arranged in a highly condensed area on the ventral surface. Nevertheless, the fifth maxilliped indeed belongs to an extra segment, which is not included into the cephalothorax and which possesses an extra tergite (e.g., Ahyong 2001). Research in developmental genetics has supported this long known morphological phenomenon: dorsal and ventral segmentation patterns need not necessarily be coupled (see Janssen et al. 2004). Also, the postcephalic trunk region of labrophorans or eucrustaceans (head includes here the maxillary = fifth appendage-bearing segment) may be further subdivided into tagmata and with this exhibit differentiated appendage morphologies. A subdivision of the trunk into an appendage-bearing thorax and a limbless abdomen is a possible autapomorphy supporting the taxon Entomostraca (Walossek and Mü ller 1998a, Maas et al. 2003, Waloszek 2003b). An autapomorphic trunk division of Malacostraca is the differentiation into two compartments, called thorax and pleon or thorax I and II (see Walossek and Mü ller 1998a), both with differently specialized appendages (Waloszek 2003b). In Eumalacostraca the pleon is further subdivided as the sixth pair of appendages is transformed in structure and insertion to form a tail fan together with the telson (see Fig. 2.7D,E for two stomatopod species). Functional subdivision of the limb-bearing thorax is rarely known from Entomostraca, such as in the Devonian fossil Lepidocaris rhyniensis. In this species, the anterior thoracopods have well-developed endites and are also used for food transport, while the posterior appendages have symmetric rami and are mainly used for swimming (Scourfield 1926). In many Entomostraca, the last thoracopods may be specialized, often highly modified, for copulation or as an egg carrier (e.g., Sanders 1957 for Cephalocarida; Torrentera and Dodson 1995 for Anostraca). In Malacostraca, the anterior one or two pleopods may be modified as a sperm-transfer appendage (mostly called petasma, though likely comprising various different morphologies; e.g., Martin and Abele 1986), or a thoracopod of the first thoracic limb series bears an appendix for sperm transfer. In summary, many morphological changes occur in conjunction with copulation and the transfer (in males) or the reception (in females) of sperm. The same holds for brood care, which also led to the development or modification of specific structures associated with the thoracopods, such as oostegites in Peracarida (as possible modified epipodites; see Maas et al. 2009) or specific setae that hold the eggs, such as in diplostracans. Within various crustacean groups and possibly in the ground pattern character at least of Eucrustacea, epipodites occur as rather soft outgrowths on the lateral side of the limb stems of thoracopods. They may serve for a respiratory or osmoregulatory function (see Maas et al. 2009 for an extensive essay of these structures).

FUTURE DIRECTIONS One important unsolved riddle is the early evolution of Malacostraca. As mentioned above, entomostracans are well represented in the fossil record as early as the Terreneuvian, while we

Evolution of Crustacean Appendages simply lack malacostracan fossils for this period and even several million years later. All Paleozoic malacostracans show clear malacostracan features, so they are not derivatives of the stem lineage and cannot contribute much to the reconstruction of the early evolution of Malacostraca and their appendages. Here we have to wait for new fossil discoveries, because comparison of extant malacostracan and entomostracan species cannot tell us anything about the evolutionary split of these two taxa. Also, the ground pattern status of Maxillopoda needs further stabilization, because this taxon represents, together with Branchiopoda and Cephalocarida, the taxa with extant derivatives within the Entomostraca. Therefore, we are confident that a closer look at some fossil maxillopods, for example, Dala peilertae, will contribute to the solution of this problem. Another problem awaiting a solution is the phylogenetic position of Insecta, respectively Tracheata. Especially for the appendages, it is difficult to draw out a plausible evolutionary scenario that would show the evolution from a crustacean ingroup taxon to Insecta/Tracheata. Especially difficult would be the derivation via Branchiopoda, as has been suggested quite recently (Glenner et al. 2006). Insecta or Tracheata may well be Labrophora or at least their sister group (Zhang et al. 2007), but still an evolutionary scenario remains hard to reconstruct with the existing data on early insect/tracheate evolution. Leaving the character “appendages” aside for reconstructing phylogenetic trees is not helpful since the animals evolved as a whole and not just their single character complexes. Besides studying fossil specimens, a closer look at the development of several taxa living today will provide additional information of probable use for this question. While many publications in the modern evo-devo field already show promising results, the classic morphological studies of arthropod ontogeny also contribute important data (see, e.g., Liu et al. 2009). Nevertheless, many fossil and extant arthropod species need to be studied to finally solve this section of the tree of life.

CONCLUSIONS Postantennular appendages of Crustacea show an extremely large variety of morphologies—in our view more than in any other taxon within Euarthropoda, and this diversity is considered the driving force of the crustacean success. In living eucrustaceans the head appendages are specifically dissimilar to the trunk appendages. This dissimilarity is also evident in land-living euarthropods such as spiders and insects. As we know now from a number of fossil species, crustacean evolution began, however, with only three specialized head appendages, that is, three anterior appendages differing from the remaining ones—yet more than in the euarthropod ground pattern! Specialization of the two postantennular appendages occurred successively along the crustacean evolutionary lineage, most significantly in the ground pattern of Labrophora. Any coincidence in the morphology of head appendages between Eucrustacea and the insects refers to rather basal features, which would be explainable only when assuming at most a derivation from a basal node within Labrophora. Ingroup eucrustacean affinities, as suggested, for example, by molecular studies (e.g., Regier et al. 2008) suffer from little support by any morphological structures (e.g., one criticism is that the segments bearing the excretory glands in insects differ from those in all eucrustacean taxa). Furthermore, a huge variety of tree suggestions may also be explained as being founded on symplesiomorphies. Again, tagmatization of the body into head and trunk does not correspond to a functional tagmatization of the limb series. A head with four appendage-bearing segments in the crustacean ground pattern has only three specialized limbs originally. The fourth head limb is functionally a trunk appendage. A larger head including the fifth limb-bearing segment in the eucrustacean ground pattern is in conjunction with the specialization of the fourth head limb. However, the fifth head limb is still trunk-limb shaped. Therefore, defining what a trunk limb or thoracopod

67

68

Functional Morphology and Diversity is depends heavily on the evolutionary level. We even think that a differentiation is rather inappropriate, since further trunk segments become involved in the head, although their appendages do not necessarily change their morphology. Lastly, it is important to note that counting (at particular structures and taking the evolutionary level into account) deserves a better reputation. The endopod, for example, originally having more than eight articles in the ground pattern of Euarthropoda, shortened to comprise seven articles at most in the sister taxon to Chelicerata (e.g., trilobites, Agnostus pisiformis). The basal number is unclear for myriapods and hexapods (or Tracheata if monophyly is favored) because of the unclear subdivision of the entire limb but should be six in the endopods of Crustacea. Within this group, many taxa evolved lower numbers, individually on different limbs and groups of limbs, but never more. Probably most important when dealing with crustacean appendages is to recognize that there is no special “crustaceopodium.” The morphology, ontogeny, and evolution of each crustacean appendage need to be studied separately and in combination with the other appendages.

ACKNOWLEDGMENTS First, we thank Martin Thiel and Les Watling for their invitation to contribute to this volume and for their editorial work. Thanks are also due to the Central Facility of Electron Microscopy of the University of Ulm for their support in using their equipment. Several images were taken during research stays in Copenhagen and London by J.T.H. and C.H., which were supported by grants from the European Commission’s (FP 6) Integrated Infrastructure Initiative program SYNTHESYS (DK-TAF-2171, DK-TAF-2652, GB-TAF-4733). During these visits, scanning electron microscopic (SEM) images were taken on a JEOL JSM-840 and JEOL SM-31010 (Copenhagen), and light microscopy was performed on a Leica DFC 480 (London). Jørgen Olesen, Copenhagen, kindly provided access to specimens from his collection (Figs. 2.5G,H, 2.6B, 2.7F,G). We would also like to thank several persons for their permission to use some of their images: Zhang Xi-guang, Kunming (Figs. 2.1G, 2.5C, 2.6C), Jørgen Olesen, Copenhagen (Fig. 2.5E), Yu Liu, Munich (Fig. 2.2B), and Verena Kutschera, Ulm (Fig. 2.7E). In Ulm, images were taken on an SEM Zeiss DSM 962 at the Central Unit for Electron Microscopy and on a Zeiss Axioskop with a mounted DCM 510 ocular camera. Some images had to be enhanced or processed with various computer programs. Therefore, we express our sincere thanks to the people that spent their time in providing open-access and open-source software programs such as ImageJ, CombineZM, GIMP, and Inkscape. Lastly, we thank Klaus J. Mü ller, discoverer of the Orsten, and the German Research Foundation (DFG) for its continuous funding of the Orsten research activities, for which J.T.H. received funding under DFG WA-754/15–1.

REFERENCES Ahyong , S.T. 2001. Revision of the Australian stomatopod Crustacea. Records of the Australian Museum, Supplement 26:1–326. Ax , P. 1995 . Das System der Metazoa I. Ein Lehrbuch der phylogenetischen Systematik. Gustav Fischer, Stuttgart. Bergström, J., X. Hou, X. Zhang , and S. Clausen . 2008. A new view of the Cambrian arthropod Fuxianhuia. Geologiska Föreningens i Stockholm Förhandlingar 130:189–201. Bergström, J., W. Stürmer, and G. Winter. 1980. Palaeoisopus, Palaeopantopus and Palaeothea, pycnogonid arthropods from the Lower Devonian Hunsrück Slate, West Germany. Paläontologische Zeitschrift 54:7–54.

Evolution of Crustacean Appendages Bowman, T.E., J. Yager, and T.M. Iliffe. 1985 . Speonebalia cannoni n.gen., n.sp., from the Caicos Islands, the first hypogean leptostracan (Nebaliacea: Nebaliidae). Proceedings of the Biological Society of Washington 98:439–446. Boxshall, G.A. 2004 . The evolution of arthropod limbs. Biological Reviews 79:253–300. Boxshall, G.A., and D. Jaume. 2009. Exopodites, epipodites and gills in Crustaceans. Arthropod Systematics and Phylogeny 67:229–254. Briggs, D.E.G., 1976. The arthropod Branchiocaris n. gen., Middle Cambrian, Burgess Shale, British Columbia. Geological Survey Bulletin 264:1–29. Briggs, D.E.G., 1978. The morphology, mode of life, and affinities of Canadaspis perfecta (Crustacea, Phyllocarida), Middle Cambrian, Burgess Shale, British Columbia. Philosophical Transactions of the Royal Society of London 281:439–487. Briggs, D.E.G., M.D. Sutton, David J. Siveter, and Derek J. Siveter. 2004 . A new phyllocarid (Crustacea: Malacostraca) from the Silurian Fossil-Lagerstätte of Herefordshire, UK. Proceedings of the Royal Society of London Series B 271:131–138. Budd, G.E. 1998. Stem group arthropods from the Lower Cambrian Sirius Passet fauna of North Greenland. Pages 125–138 in R.A. Fortey and R.H. Thomas, editors. Arthropod Relationships. Systematics Association Special Volume Series, Vol. 55. Chapman and Hall, London . Budd, G.E. 2008. Head structure in upper stem-group euarthropods. Palaeontology 51:561–573. Calman, W.T. 1909. Part VII. Appendiculata. Pages 1–346 in Lankester, R., editor. A Treatise on Zoology. Crustacea, Vol. 3. Adam and Charles Black , London . Carcupino, M., A. Floris, A. Addis, A. Castelli, and M. Curini-Galletti. 2006. A new species of the genus Lightiella: The first record of Cephalocarida (Crustacea) in Europe. Zoological Journal of the Linnean Society 148:209–220. Chen, J.-y., J. Vannier, and D.-y. Huang. 2001. The origin of crustaceans: New evidence from the Early Cambrian of China. Proceedings of the Royal Society of London Series B 268:2181–2187. Chen, J.-y., D. Waloszek , and A. Maas . 2004 . A new “great-appendage” arthropod from the Lower Cambrian of China and homology of chelicerate chelicerae and raptorial antero-ventral appendages. Lethaia 37:3–20. Cockcroft, A.C. 1985 . The larval development of Macropetasma africanum (BALSS, 1913) Decapoda, Penaeoidea) reared in the laboratory. Crustaceana 49:52–74. Dahl, E. 1984 . The subclass Phyllocarida (Crustacea) and the status of some early fossils: A neontologist’s view. Videnskabelige Meddelelser fra Dansk naturhistorik Forening 145:61–76. Dumont, H.J., and S.V. Negrea. 2002. Introduction to the class Branchiopoda. Backhuys, Leiden . Fryer, G. 1983. Functional ontogenetic changes in Branchinecta ferox (Milne-Edwards) (Crustacea, Anostraca). Philosophical Transactions of the Royal Society of London Series B 303:229–343. Fryer, G. 1988. Studies on the functional morphology and biology of the Notostraca (Crustacea: Branchiopoda). Philosophical Transactions of the Royal Society of London Series B 321:27–124. Glenner, H., P.F. Thomsen, M.B. Hebsgaard, M.V. Sørensen, and E. Willerslev . 2006. The origin of insects. Science 314:1883–1884. Hamner, W.M. 1988. Biomechanics of filter feeding in the Antarctic krill Euphausia superba: Review of past work and new observations. Journal of Crustacean Biology 8:149–163. Hansen, H.J. 1925 . On the comparative morphology of the appendages in the Arthropoda. A. Crustacea. Studies on Arthropoda, Vol. 2. Gyldendalske Boghandel, Kopenhagen. Haug , J.T., C. Haug , A. Maas, S.R. Fayers, N.H. Trewin, and D. Waloszek . 2009b. Simple 3D images from fossil and Recent micromaterial using light microscopy. Journal of Microscopy 233:93–101. Haug , J.T., Maas, A., and D. Waloszek. 2009a . Ontogeny of two Cambrian stem crustaceans, Goticaris longispinosa and Cambropachycope clarksoni. Palaeontographica Abteilung A 289:1–43. Haug , J.T., Maas, A., and D. Waloszek. 2010a . Henningsmoenicaris scutula, Sandtorpia vestrogothiensis gen. et sp. nov., and heterochronic events in early crustacean evolution. Transactions of the Royal Society of Edinburgh, Earth and Environmental Science 100:311–350. Haug , J.T., D. Waloszek , C. Haug , and A. Maas. 2010b. High-level phylogenetic analysis using developmental sequences: The Cambrian Martinssonia elongata, Musacaris gerdgeyeri gen. et sp. nov., and their position in early crustacean evolution. Arthropod Structure and Development 39:154–173.

69

70

Functional Morphology and Diversity Heegaard, P. 1945 . Remarks on the phylogeny of arthropods. Arkiv för Zoologi 37A:1–15. Heldt, J.H. 1954 . Waptia fieldensis Walcott et les stades larvaires des Pénéides. Bulletin de la Societe des Sciences Naturelles de Tunisie 6:177–180. Hennig , W. 1965 . Phylogenetic systematics. Annual Review of Entomology 10:97–116. Hessler, R.R. 1964 . The Cephalocarida. Comparative skeletomusculature. Memoirs of the Connecticut Academy of Arts and Sciences 16:1–97. Hirota, Y., T. Nemoto, and R. Marumo. 1984a . Larval development of Euphausia nana (Crustacea: Euphausiacea). Marine Biology 81:311–322. Hirota, Y., T. Nemoto, and R. Marumo. 1984b. Larval development of Euphausia similis (Crustacea, Euphausiaeea) in Sagami Bay, Central Japan. Journal of the Oceanographical Society of Japan 40:57–66. Hou, X.-G., and J. Bergström. 1997. Arthropods of the Lower Cambrian Chengjiang fauna, southwest China. Fossils and Strata 45:1–116. Hou, X.-G., J. Bergström, and G.-H. Xu. 2004 . The lower Cambrian crustacean Pectocaris from the Chengjiang biota, Yunnan, China. Journal of Paleontology 78:700–708. Itô, T. 1982. The origin of “biramous” copepod legs. Journal of Natural History 16:715–726. Itô, T. 1989. Origin of the basis in copepod limbs, with reference to remipedian and cephalocarid limbs. Journal of Crustacean Biology 9:85–103. Janssen, R., N.M. Prpic , and W.G.M. Damen. 2004 . Gene expression suggests decoupled dorsal and ventral segmentation in the millipede Glomeris marginata (Myriapoda: Diplopoda). Developmental Biology 268:89–104. Jones, M.L. 1961. Lightiella serendipita gen. nov., sp. nov., a cephalocarid from San Francisco Bay, California. Crustaceana 3:31–46. Kaestner, A. 1967. Lehrbuch der Speziellen Zoologie. Gustav Fischer, Stuttgart . Kidd , R.J. 1991 . Development of embryonic and naupliar setae and spines and their role in hatching in the penaeid Sicyonia ingentis: A light and electron microscopic study. Journal of Crustacean Biology 11:40–55. Latreille, P. A. 1829. Les crustacés, les arachnides, les insectes. Pages 1–584 in G. Cuvier, editor. Le règne animal distribué d’après son organisation, pour servir de base à l’histoire naturelle des animaux et d’introduction à l’anatomie comparée. Déterville, Crochard, Paris. Lauterbach, K.E. 1986. Zum Grundplan der Crustacea. Verhandlungen des naturwissenschaftlichen Vereins Hamburg (NF) 28:27–63. Liu, J., and X.-P. Dong. 2007. Skara hunanensis a new species of Skaracarida (Crustacea) from Upper Cambrian (Furongian) of Hunan, South China. Progress in Natural Science 17: 934–942. Liu, Y., X.-G. Hou, and J. Bergström. 2007. Chengjiang arthropod Leanchoilia illecebrosa (Hou, 1987) reconsidered. Geologiska Föreningens i Stockholm Förhandlingar 129:263–272. Liu, Y., A. Maas, and D. Waloszek . 2009. Early development of the anterior body region of the grey widow spider Latrodectus geometricus Koch, 1841 (Theridiidae, Araneae). Arthropod Structure and Development 38:401–416. Maas, A., A. Braun, X.-p. Dong , P.C.J. Donoghue, K.J. Mü ller, E. Olempska , J.E. Repetski, D.J. Siveter, M. Stein, and D. Waloszek . 2006. The “Orsten”—more than a Cambrian Konservat-Lagerst ätte yielding exceptional preservation. Palaeoworld 15:266–282. Maas, A., C. Haug , J.T. Haug , J. Olesen, X. Zhang , and D. Waloszek . 2009. Early crustacean evolution and the appearance of epipodites and gills. Arthropod Systematics and Phylogeny 67:255–273. Maas, A., G. Mayer, R.M. Kristensen, and D. Waloszek . 2007. A Cambrian micro-lobopodian and the evolution of arthropod locomotion and reproduction. Chinese Science Bulletin 52:3385–3392. Maas, A., and D. Waloszek. 2001a . Cambrian derivatives of the early arthropod stem lineage, pentastomids, tardigrades and lobopodians—an “Orsten” perspective. Zoologischer Anzeiger 240:451–459. Maas, A., and D. Waloszek. 2001b. Larval development of Euphausia superba Dana, 1852 and a phylogenetic analysis of the Euphausiacea. Hydrobiologia 448:143–169. Maas, A., and D. Waloszek . 2005 . Phosphatocopina—ostracode-like sister group of Eucrustacea. Hydrobiologia 538:139–152.

Evolution of Crustacean Appendages Maas, A., D. Waloszek , J.-y. Chen, A. Braun, X.-q. Wang , and D.-y. Huang. 2004 . Phylogeny and life habits of early arthropods—predation in the early Cambrian sea. Progress in Natural Science 14:158–166. Maas, A., D. Waloszek , and K.J. Mü ller. 2003. Morphology, ontogeny and phylogeny of the Phosphatocopina (Crustacea) from the Upper Cambrian “Orsten” of Sweden. Fossils and Strata 49:1–238. Martin, J.W., and L.G. Abele. 1986. Notes on male pleopod morphology in the brachyuran crab family Panopeidae Ortmann, 1893, sensu Guinot (1978) (Decapoda). Crustaceana 50:182–198. Martin, J.W., and C.E. Cash-Clark. 1995 . The external morphology of the onychopod “cladoceran” genus Bythotrephes (Crustacea, Branchiopoda, Onychopoda, Cercopagididae), with notes on the morphology and phylogeny of the order Onychopoda. Zoologica Scripta 24:61–90. Mayer, G., G. Maier, A. Maas, and S. Waloszek . 2009. Mouthpart morphology of Gammarus roeselii compared to a successful invader, Dikerogammarus villosus (Amphipoda). Journal of Crustacean Biology 29:161–174. Morgan, S.G., and J.W. Goy. 1987. Reproduction and larval development of the mantis shrimp Gonodactylus bredini (Crustacea: Stomatopoda) maintained in the laboratory. Journal of Crustacean Biology 7:595–618. Mü ller, K.J. 1983. Crustacea with preserved soft parts from the Upper Cambrian of Sweden. Lethaia 16:93–109. Mü ller, K.J., and D. Walossek. 1985 . Skaracarida, a new order of Crustacea from the Upper Cambrian of V ä stergötland, Sweden. Fossils and Strata 17:1–65. Mü ller, K.J., and D. Walossek. 1986a . Arthropod larvae from the Upper Cambrian of Sweden. Transactions of the Royal Society of Edinburgh, Earth and Environmental Science 77:157–179. Mü ller, K.J., and D. Walossek. 1986b. Martinssonia elongata gen. et sp. n., a crustacean-like euarthropod from the Upper Cambrian “Orsten” of Sweden. Zoologica Scripta 15:73–92. Mü ller, K.J., and D. Walossek . 1987. Morphology, ontogeny and life habit of Agnostus pisiformis from the Upper Cambrian of Sweden. Fossils and Strata 19:1–124. Mü ller, K.J., and D. Walossek . 1988. External morphology and larval development of the Upper Cambrian maxillopod Bredocaris admirabilis. Fossils and Strata 23:1–70. Neil, D.M., D.L. Macmillan, R.M. Robertson, and M.S. Laverack. 1976. The structure and function of thoracic exopodites in the larvae of the lobster Homarus gammarus (L.). Philosophical Transactions of the Royal Society of London 274:53–68. Newman, W.A. 1983. Origin of the Maxillopoda; urmalacostracan ontogeny and progenesis. Pages 105– 120 in F.R. Schram, editor. Crustacean phylogeny. Crustacean Issues, Vol. 1. Balkema, Rotterdam, The Netherlands. Olesen, J. 2004 . On the ontogeny of the Branchiopoda (Crustacea): Contribution of development to phylogeny and classification. Pages 217–269 in G. Scholtz , editor. Evolutionary developmental biology of Crustacea. Crustacean Issues, Vol. 15. Balkema, Lisse, The Netherlands. Olesen, J. 2007. Monophyly and phylogeny of Branchiopoda, with focus on morphology and homologies of branchiopod phyllopodous limbs. Journal of Crustacean Biology 27:165–183. Olesen, J., J.W. Martin, and E.W. Roessler 1996. External morphology of Cyclestheria hislopi (Baird, 1859) (Crustacea, Branchiopoda, Spinicaudata), with a comparison of male claspers among the Conchostraca and Cladocera and its bearing on phylogeny of the “bivalved” Branchiopoda. Zooogica Scripta 25:291–316. Olesen, J., S. Richter, and G. Scholtz . 2001. The evolutionary transformation of phyllopodous to stenopodous limbs in the Branchiopoda (Crustacea)—is there a common mechanism for early limb development in arthropods? International Journal of Developmental Biology 45:869–876. Olesen, J., S. Richter, and G. Scholtz . 2003. On the ontogeny of Leptodora kindtii (Crustacea, Branchiopoda, Cladocera), with notes on the phylogeny of the Cladocera. Journal of Morphology 256:235–259. Olesen, J., and D. Walossek. 2000. Limb ontogeny and trunk segmentation in Nebalia species (Crustacea, Malacostraca, Leptostraca). Zoomorphology 120:47–64. Preuss, G. 1957. Die Muskulatur der Gliedma ßen von Phyllopoden und Anostraken. Mitteilungen des Zoologischen Museums Berlin 33:221–256.

71

72

Functional Morphology and Diversity Regier, J.C., J.W. Shultz , A.R. Ganley, A. Hussey, D. Shi, B. Ball, A. Zwick , J.E. Stajich, M.P. Cummings, J.W. Martin, and C.W. Cunningham . 2008. Resolving arthropod phylogeny: Exploring phylogenetic signal within 41 kb of protein-coding nuclear gene sequence. Systematic Biology 57:920–938. Richter, S., and G. Scholtz. 2001. Phylogenetic analysis of the Malacostraca (Crustacea). Journal of Zoological Systematics and Evolutionary Research 39:113–136. Rudkin, D.M., G.A. Young , R.J. Elias, and E.P. Dobrzanski. 2003. The world’s biggest trilobite— Isotelus rex new species from the Upper Ordovician of Northern Manitoba, Canada. Journal of Paleontology 77:99–112. Sanders, H.L. 1957. The Cephalocarida and crustacean phylogeny. Systematic Zoology 6:112–128. Sanders, H.L. 1963. The Cephalocarida: Functional morphology, larval development, comparative external anatomy. Memoirs of the Connecticut Academy of Arts and Sciences 15:1–80. Schram, F.R. 1986. Crustacea . Oxford University Press, New York . Schram, F.R. 2007. Paleozoic proto-mantis shrimp revisited. Journal of Paleontology 81:895–916. Scourfield, D.J. 1926. On a new type of crustacean from the Old Red Sandstone (Rhynie Chert Bed, Aberdeenshire)—Lepidocaris rhyniensis gen. et sp. nov. Philosophical Transactions of the Royal Society of London 214:153–187. Shu, D., X.-L. Zhang , and G. Geyer. 1995 . Anatomy and systematic affinities of the Lower Cambrian bivalved arthropod Isoxys auritus. Alcheringa 19:333–342. Siveter, David J. 2008. Ostracods in the Palaeozoic? Senckenbergiana lethaea 88(1):1–9. Siveter, David J., D. Waloszek , and M. Williams. 2003. An early Cambrian phosphatocopid crustacean with three-dimensionally preserved soft parts from Shropshire, England. Special Papers in Palaeontology 70:9–30. Siveter, Derek J., R.A. Fortey, M.D. Sutton, D.E.G. Briggs, and David J. Siveter. 2007. A Silurian “marrellomorph” arthropod. Proceedings of the Royal Society of London Series B 274:2223–2229. Snodgrass, R.E. 1958. Evolution of arthropod mechanisms. Smithsonian Miscellaneous Collections 138:1–77. Stein, M., D. Waloszek , and A. Maas. 2005 . Oelandocaris oelandica and its significance to resolving the stem lineage of Crustacea. Pages 55–71 in S. Koenemann and R. Vonck , editors. Crustacea and arthropod relationships. CRC Press, Boca Raton, FL. Stein, M., D. Waloszek , A. Maas, J.T. Haug , and K.J. Mü ller. 2008. The stem crustacean Oelandocaris oelandica re-visited. Acta Palaeontologica Polonica 53:461–484. Størmer, L. 1939. Studies on trilobite morphology. Part I. The thoracic appendages and their phylogenetic significance. Norsk Geologisk Tidsskrift 19:143–273. Taylor, R.S. 2002. A new bivalved arthropod from the Early Cambrian Sirius Passet fauna, North Greenland. Palaeontology 45:97–123. Torrentera, L., and S.I. Dodson. 1995 . Morphological diversity of populations of Artemia (Branchiopoda) in Yucatá n. Journal of Crustacean Biology 15:86–102. Walossek , D. 1999. On the Cambrian diversity of Crustacea. Pages 3–27 in F.R. Schram and J.C. von Vaupel Klein, editors. Crustaceans and the biodiversity crisis, Proceedings of the Fourth International Crustacean Congress, Amsterdam, The Netherlands, July 20–24, 1998, Vol. 1. Brill Academic Publishers, Leiden, The Netherlands. Walossek , D. 1993. The Upper Cambrian Rehbachiella and the phylogeny of Branchiopoda and Crustacea. Fossils and Strata 32:1–202. Walossek , D., I. Hinz-Schallreuter, J.H. Shergold, and K.J. Mü ller. 1993. Three-dimensional preservation of arthropod integument from the Middle Cambrian of Australia. Lethaia 26:7–15. Walossek , D., and K.J. Mü ller. 1998a . Cambrian “Orsten”-type arthropods and the phylogeny of Crustacea. Pages 67–86 in R.A. Fortey and R.H. Thomas, editors. Arthropod relationships. Systematics Association Special Volume Series, Vol. 55. Chapman and Hall, London . Walossek , D., and K.J. Mü ller. 1998b. Early arthropod phylogeny in light of the Cambrian “Orsten” fossils. Pages 185–231 in G.D. Edgecombe, editor. Arthropod fossils and phylogeny. Columbia University Press, New York. Walossek , D., and K.J. Mü ller. 1989. A second type A-nauplius from the Upper Cambrian “Orsten” of Sweden. Lethaia 22:301–306.

Evolution of Crustacean Appendages Walossek , D., and K.J. Mü ller. 1990. Upper Cambrian stem-lineage crustaceans and their bearing upon the monophyletic origin of Crustacea and the position of Agnostus. Lethaia 23:409–427. Waloszek , D. 2003a . The “Orsten” window—a three-dimensionally preserved Upper Cambrian meiofauna and its contribution to our understanding of the evolution of Arthropoda. Paleontological Research 7:71–88. Waloszek , D. 2003b. Cambrian “Orsten”-type preserved arthropods and the phylogeny of Crustacea. Pages 66–84 in A. Legakis, S. Sfenthourakis, R. Polymeni, and M. Thessalou-Legaki, editors. The new panorama of animal evolution. Pensoft Publishers, Sofia, Bulgaria . Waloszek , D., J. Chen, A. Maas, and X. Wang. 2005 . Early Cambrian arthropods—new insights into arthropod head and structural evolution. Arthropod Structure and Development 34:189–205. Waloszek , D., A. Maas, J. Chen, and M. Stein . 2007. Evolution of cephalic feeding structures and the phylogeny of Arthropoda. Palaeogeography Palaeoclimatology Palaeoecology 254:273–287. Watling. L. 1993. Functional morphology of the amphipod mandible. Journal of Natural History 27:837–849. Wehner, R., and W. Gehring. 1995 . Zoologie. Thieme, Stuttgart . Williams, T.A. 2004 . The evolution and development of crustacean limbs: an analysis of limb homologies. Pages 169–193 in G. Scholtz , editor. Evolutionary developmental biology of Crustacea. Crustacean Issues, Vol. 15. Balkema, Lisse, The Netherlands. Wolff, C., and G. Scholtz . 2008. The clonal composition of biramous and uniramous arthropod limbs. Proceedings of the Royal Society of London Series B 275:1023–1028. Zhang , X.-g., D.J. Siveter, D. Waloszek , and A. Maas. 2007. An epipodite-bearing crown-group crustacean from the Lower Cambrian. Nature 449:595–598.

73

3 MECHANISMS OF LIMB PATTERNING IN CRUSTACEANS

Terri A. Williams

Abstract The structural diversity of crustacean limbs is enormous, and their evolution is hotly debated. Attempts have been made to understand the developmental patterning mechanisms that generate distinct types of adult limbs by analyzing genes known to regulate limb development in the arthropod model system, Drosophila . This has led to the discovery of deeply conserved features of limb patterning, although it has not clarified how some of the most basic limb structures—a biramous limb or endites or exites—are patterned. Indeed, based on available data, one hypothesis is that endites and exites may have varied independently during evolution. Analyses of patterning during crustacean limb development are further complicated by the fact that their larval stages can have limbs that are quite distinct from those of adults. This means that any particular body segment may develop two or more structurally distinct limbs during the course of the life cycle—a phenomenon yet to be captured by models of limb patterning. Finally, the diversity of limbs is evident not only in their overall plan but also in the details of setae, joints, and muscles that permit their functional specialization. Like many other areas of study in crustacean limb development, the analysis of how these structures develop has barely begun.

INTRODUCTION The evolutionary radiation of arthropods was driven by limb diversity, and nowhere is this more evident than in crustaceans. The structural diversity of adult limbs is enormous. Indeed, the differences are so large that homologies are not fully resolved among some limbs (Williams 2004). The intricacy and functional breadth of crustacean limbs make them a natural choice for analyzing how development might have been modified to produce such morphological variation. While we lack a systematic comparison of leg morphogenesis in crustaceans based on

74

Functional Morphology and Diversity. Edited by Les Watling and Martin Thiel. © 2013 Oxford University Press. Published 2013 by Oxford University Press.

Mechanisms of Limb Patterning in Crustaceans classical descriptive research, more specialized studies—for example, of cellular morphogenesis or molecular patterning—are emerging. The purpose of this chapter is to provide an overview of the genetic regulation of patterning in crustacean thoracic limb development. The focus is on the morphological diversity of crustaceans, and the failure of models of comparative limb development to explain that diversity is noted. The first section is a brief overview of leg development in crustaceans, emphasizing that most crustaceans have larval stages that include functional limbs. There are various degrees of metamorphosis between stages, and these include transformations of limb morphology. Thus, models of crustacean limb development and patterning should include not just the initial embryonic limb but also the transformations of later developmental stages. Currently, our understanding of which genes function to regulate patterning in crustacean limbs derives directly from comparisons of development in the fruit fly Drosophila melanogaster. Therefore, to summarize what is known about crustacean limb patterning, the next section is an overview of leg patterning in Drosophila and other insects. This is followed by specific cases where these leg-patterning genes have been examined in crustaceans. No general model of developmental patterning has emerged that explains adult limb diversity in crustaceans. However, in this chapter it is hypothesized that limb patterning in crustaceans was controlled ancestrally by a number of distinct regulatory networks that later became more or less interdependent in the highly derived and specialized legs of Drosophila. One novel implication of this hypothesis is that medial and lateral limb structures may have varied independently during evolution. This hypothesis is followed by an exploration of the fact that crustaceans have different larval stages with functionally different limbs. Explaining these phenomena requires explaining how radical transformations of limb structure can precede the development of the adult limb. However, this is not addressed in current models of limb development. That is, the question is not simply how one limb is patterned but how a series of quite distinct limbs is patterned on one segment. Most models of limb patterning explain only a basic coordinate system of positional information, that is, proximal-distal, anterior-posterior, and dorsal-ventral axes, that map out a generic limb field. In the final section of this chapter, limbs are fleshed out as structures made of setae, muscle, and nerves. The development of these structures and the potential for integrating them into current models of limb patterning are considered. Although crustaceans provide the most diverse taxon for studies of limb development and evolution, that diversity has just begun to be sampled and underrepresents the actual limb diversity in crustaceans. This limits our ability to make inferences about which features of limb patterning are ancestral and which patterning mechanisms are variable versus constrained. The emphasis throughout this chapter is on the richness of structural limb diversity that remains virtually unexplored by developmental analysis and yet is the basis for the functional limb diversity that is showcased in this book.

A GENERAL DESCRIPTION OF LEG DEVELOPMENT IN CRUSTACEANS Despite their varied life histories, crustaceans tend to develop in a progressive and sequential manner. That is, body segments form in an anterior-to-posterior sequence regardless of whether they develop in the embryonic or larval stage. Subsequently, limb buds develop as direct outpocketings of the ventral or ventral lateral body wall. In parallel with the sequential segment development in crustaceans, limb development occurs in a progressive fashion, with limb buds gaining complexity over time (Fig. 3.1). Are there common cellular dynamics that accomplish outgrowth of the limb bud from the ventral body wall? We know very little about this in crustaceans. In Drosophila, the arthropod

75

76

Functional Morphology and Diversity

Fig. 3.1. Progressive leg development in the anostracan larva Thamnocephalus platyurus. T. platyurus has a series of similar limbs on the trunk. As they develop, limb buds show a gradual elaboration of morphological features, from less developed posterior limbs to more developed anterior ones (inset). Scale bar, 100 μm.

model for limb development, initial specification of the limb bud differentiates a cluster of cells that do not undergo immediate rapid growth relative to flanking cells but are instead set aside for later proliferation. By contrast, in crustaceans, limb buds emerge directly from the surrounding body wall in a manner analogous to vertebrate limbs, where we know cell dynamics play an important role. In vertebrate limbs, the sites of future limb development are detectable as regions of cell proliferation in the lateral flank. After the emergence of the limb bud from the flank, a ridge of stratified epithelium, the apical ectodermal ridge, extends across the distal tip of the limb bud and functions via signaling to maintain a higher rate of cell division in the distal limb mesenchyme (Sun et al. 2002). However, these types of cell dynamics typically remain undescribed for direct developing arthropod limbs. In one case where the cell dynamics of initial outgrowth was examined, Freeman et al. (1992) found that a combination of cell division and cell shape change accounts for initial outpocketing of the limb bud from the ventral body wall in the branchiopod Artemia. Furthermore, they found that pharmacologically arresting cell proliferation could prevent normal evagination of the limb bud. This is a potentially fruitful but unexplored area; characterizing the cell dynamics underlying limb outgrowth in more species might uncover common mechanisms of outgrowth, analogous to the apical ectodermal ridge in vertebrate limbs. Once the limb bud differentiates from the body wall, it undergoes growth, elongation, and segmentation. Again, in vertebrate limbs, the course of this process is well described at the cellular level: differentiation proceeds in a proximal-to-distal sequence, driven by a distal zone of

Mechanisms of Limb Patterning in Crustaceans proliferating cells (Sun et al. 2002). By contrast, in arthropods, the basic sequence of morphological differentiation of limb segments is not well known for most species. What is known suggests that arthropods do not form their leg segments in a simple proximal-to-distal sequence. Instead, they undergo a process of intercalary growth. In crickets, for example, boundaries formed by developing joints arise sequentially, sometimes proximal or sometimes distal to the previous boundary (Inoue et al. 2002). It would be interesting to know whether segmented walking legs in crustaceans share a developmental sequence in segmentation that might reflect a shared underlying patterning mechanism. This has never been examined systematically in crustaceans, nor have growth and elongation in nonwalking legs, for example, pleopods or phyllopods. How does adult morphology arise? In crustaceans, there are two rather distinct routes to generate adult limb morphology. First, adult legs may arise directly from the gradual development and refinement of the initial limb bud that emerges from the body wall. This happens in, for example, peracarids and anostracans (Fig. 3.2A). In these species, limb buds emerge from the body and gradually increase in size and complexity as they form the adult limb morphology. However, most crustacean taxa undergo varying degrees of metamorphosis during their life cycle (Snodgrass 1956). So, for the majority of species, adult limb morphology develops from modifications of preexisting larval limbs. In these cases, the first limb bud that emerges from the body wall becomes a functioning larval limb. This limb may then be modified one or more times before assuming the adult morphology. Adult limb morphology is produced only after limbs of quite different morphology have developed on the same segment (Fig. 3.2B). Notably, these differences can arise from one molt to the next in quite dramatic fashion without a period of quiescence. The types of metamorphosis are too numerous to catalogue systematically but include the following transformations of limb morphology (typically observed in a single molt): gain or loss of the exopod, gain or loss of limb segments, gain or loss of endites or exites, complete loss of limb and later redevelopment in alternate form, and extreme hypertrophy of one feature of the limb (for examples, see Gurney 1942, Snodgrass 1956). Although not accounted for by models of leg development, explaining these diverse routes of development are crucial for understanding most crustaceans.

LEG PAT TERNING IN INSECTS BASED ON A DROSOPHILA MODEL OF LEG DEVELOPMENT Models of arthropod leg patterning are based on leg development in Drosophila melanogaster because it has been possible in that species to dissect the function of many genes that provide positional information to the developing leg. These genes give cells in the developing leg their spatial fates, for example, proximal or distal and dorsal or ventral. Although a wealth of information has been discovered about limb patterning in Drosophila, it has been challenging to compare this information to other arthropods. In particular, Drosophila’s metamorphic development is highly atypical among arthropods: legs do not grow directly out of the body wall, and indeed, Drosophila has no legs until it reaches the adult stage. However, leg patterning occurs throughout the life cycle, beginning with specification of the leg primordium in embryogenesis, followed by elaboration of the leg positional information during the larval stages, and finally differentiation of leg morphology in the pupal stage. Leg development in Drosophila begins during embryogenesis when a small group of cells on the ventral body wall are specified to become leg cells (Cohen 1990, Cohen et al. 1993). These 10–15 cells invaginate to form a saclike structure of epithelial cells lying beneath the larval epidermis called the leg imaginal disc . Then, during the second and third larval instars, cells within the leg disc proliferate and form a highly folded epithelium (Fristrom and Fristrom

77

78

Functional Morphology and Diversity

Fig. 3.2. (A) Progressive growth of legs on a crustacean that develops without radical metamorphosis in the notostracan Triops longicaudatus (successive panels represent more than a single larval molt). (B) Transformations of a leg in a crustacean that has metamorphic development in the brachyuran Perisesarma fasciatum (after Guerao et al. 2004); first maxilliped in first zoea, megalopa, and first crab larva.

1993). The teardrop-shaped disc contains the leg in compressed fashion, accordioned upon itself. At metamorphosis, it everts and extends to form the adult leg (Fig. 3.3A; Fristrom and Fristrom 1993, Taylor and Adler 2008). The constraints of metamorphosis and the consequent highly modified leg development cause leg patterning to be analyzed in discrete phases: (1) embryonic positioning of the primordial limb field on the body wall, (2) larval patterning of leg axes, and (3) the late pupal transition to metamorphosis. That is, there are periods when it is technically difficult to follow the fate of leg cells directly, for example, after invagination but before appreciable cell proliferation and growth of the disc. This lack of simple temporal continuity creates not only technical barriers but also conceptual difficulties when formulating a general model of limb development applicable to more typical arthropods that lack metamorphosis.

Mechanisms of Limb Patterning in Crustaceans

Fig. 3.3. Establishing the proximal-distal axis in Drosophila leg development. (A) Comparison of a larval leg disc to an adult leg showing the relative positions of proximal (P) and distal (D) leg. During eversion and elongation, the leg disc extends out of the plane of the paper to form the cylindrical structure of the adult leg (disc after Schubiger 1971; leg traced from a wild-type adult leg). (B) Successive time points during leg development in Drosophila showing the intercalation of additional domains of gene expression subdividing the leg disc: Distal-less in dark gray, dachshund in medium gray, overlap of Distal-less and dachshund in light gray, and homothorax (nuclear extradenticle) in black (after photo of expression patterns in Abu-Shaar and Mann 1998).

Early Positioning of Limb Primordia Straddles the Anterior-Posterior Boundary of the Embryonic Segment Leg patterning in Drosophila begins with the initial positioning and specification of the appendage primordia on the ventral body wall of the embryo. Appendage primordia are specified relative to boundaries along the anterior-posterior (AP) and dorsal-ventral (DV) body axes. In fact, the information used to pattern the body axes is sufficient to position the limb primordia without other input and imparts AP positional information to the leg. This can be visualized by boundaries of gene expression1; cells expressing the engrailed gene mark the posterior portion of each segment and subdivide the leg disc. In Drosophila, wingless-expressing cells, which form

79

80

Functional Morphology and Diversity a stripe just anterior to the engrailed-expressing cells, activate cells along the AP boundary to form limb primordia that are visible as a cluster of cells that express Distal-less. Mutants lacking Distal-less develop only proximal leg structures (Cohen and Jü rgens 1989). Comparative studies show that, while the positioning of the primordia along the AP segmental boundary appears conserved in other arthropods, the actual genes that pattern the early limb primordia—specifically, the activation of Distal-less by wingless—are not conserved even in other insect taxa (reviewed in Angelini and Kaufman 2005). Thus, mechanisms of primordia patterning in Drosophila are therefore unlikely to be ancestral for insects or, by extension, for the Pancrustacea. Establishing Proximal/Distal Positional Information in the Leg In Drosophila, fate mapping has shown that the cells that initially express Distal-less in the embryo assume multiple appendage fates (McKay et al. 2009)—wing, leg, or Keilin’s organs (larval sensory structures “homologous” to larval legs). As embryogenesis proceeds, this group of cells becomes subdivided into subsequent fates: one population of cells forms the distal limb structures in the adult (the telopod), another population forms proximal leg structures, and a third population produces the Keilin’s organs. Thus, one of the earliest patterning events in the leg proper is the polarization of proximal and distal leg domains: cells that form proximal leg structures lose the initial Distal-less expression, while cells that become distal leg maintain Distal-less expression (albeit driven by an enhancer different from that of the initial Distal-less expression). The leg cells at this point do not divide, and only during the larval instars will cell division resume and the bulk of patterning that defines the proximal-distal (PD) axis occur. Because of the small number of cells in the disc and their relatively cryptic nature early in larval development, limb patterning in Drosophila is most frequently studied from late-second-instar larvae through third-instar larvae to pupation. During this time, the leg disc grows by cell proliferation and assumes its characteristic folded shape within the larva. The initial specification of proximal and distal accomplished in the embryo becomes elaborated as new domains of PD genes are established and refined (reviewed in Kojima 2004). The elaboration of PD positional information depends on the secreted signaling molecules wingless and decapentaplegic (reviewed in Campbell and Tomlinson 1995, Held 1995, Brook et al. 1996, Williams and Nagy 1996, Blair 1999). wingless and decapentaplegic cooperatively activate target genes that are expressed in discrete domains along the PD axis of the limb (Distal-less, dachshund, and indirectly extradenticle through the gene homothorax; Lecuit and Cohen 1997). These genes are collectively termed the leg gap genes because mutations of these genes form truncated legs with gaps in PD leg morphology. In the second-instar imaginal discs, a boundary is established between extradenticle expression in a proximal domain and Distal-less expression in a distal domain. In the third instar, the Distal-less domain is subsequently subdivided into a Distal-less domain and dachshund domain (Fig. 3.3B). The boundaries of dachshund and Distal-less expression are regulated by low and high levels of wingless and decapentaplegic signaling, respectively, and through repression proximally by homothorax. These three primary domains—homothorax/extradenticle, dachshund, and Distal-less—are maintained by mutual repression. In addition, wingless and decapentaplegic regulate cell proliferation (a role particularly well studied in the wing; see Posakony et al. 1991; reviewed in Serrano and O’Farrell 1997). Similarly, the gap genes also influence leg growth. Loss-of-function mutants in leg gap genes cause truncations of normal leg morphology that span several morphological leg segments. The expression and function of the leg gap genes appear conserved in other insects. Distalless is found in the distal part of legs (Jockusch et al. 2000, 2004, Beermann et al. 2001, Abzhanov and Kaufman 2000, Inoue et al. 2002, Rogers et al. 2002), with dachshund and homothorax in

Mechanisms of Limb Patterning in Crustaceans intermediate and proximal domains, respectively (Prpic et al. 2001, Inoue et al. 2002, Angelini and Kaufman 2004). Notch Signaling Plays a Key Role in Joint Formation in the Drosophila Leg Joints within the leg disc develop during the pupal stage. They arise by invagination of the leg epithelium, driven at least in part by cell shape changes (Mirth and Akam 2002). The epithelium that forms joints can be divided into three PD positions based on gene expression and cell behavior: proximal, mid-distal, and distal. When the joint begins to form, cells of the distal and mid-distal region undergo apical constriction causing indentation of the leg cylinder. Then distal cells on the anterior and posterior sides become columnar so that the mid-distal tissue bends into the leg. Finally, proximal cells extend in a proximodistal direction forming a palisade over the indented tissue. The complex cellular dynamics of joint formation are regulated by Notch. Notch is a transmembrane receptor involved with numerous developmental decisions (reviewed in Artavanis-Tsakonas et al. 1999). Notch signaling is activated by binding to its ligands, Delta and Serrate. In the leg, Notch signaling is localized to the joints and promotes both joint formation and leg growth. The localization of Notch signaling is regulated by the leg gap genes, which act in a combinatorial fashion to regulate the expression of its ligands as well as the modulator fringe (Rauskolb 2001). In early larval stages, Notch is expressed broadly, but by the third larval instar, it is more concentrated at the joint regions (de Celis et al. 1998). Delta and Serrate are expressed as rings in each leg segment in the proximal joint region and signal to the more distal cells (de Celis et al. 1998, Bishop et al. 1999, Rauskolb and Irvine 1999, Rauskolb 2001). Notch is activated at the distal margin of the joints, and this expression correlates with the distal cells known to invaginate to form the joint (Mirth and Akam 2002). Clones of cells produced within the pupal epithelium mutant for either Notch or Delta and Serrate show the same phenotype: if the clones are in the joint regions, normal joints fail to form and the leg is shorter than wild type. If expressed ectopically, Notch causes supernumerary joints and leg outgrowths (de Celis et al. 1998, Bishop et al. 1999, Rauskolb and Irvine 1999). Thus, Notch not only regulates joint formation but, like the other PD patterning genes, also regulates leg growth. Reports from insects other than Drosophila corroborate a role for Notch signaling in leg segmentation. Beermann et al. (2004) showed the expression of Serrate in wild-type beetle Tribolium legs. Tc-Serrate is expressed in a series of rings corresponding to each segment of the leg. Similarly, in analyzing the role of fringe in grasshopper body segmentation, Dearden and Akam (2000) also described fringe expression in the leg. fringe is expressed in rings along the leg corresponding to regions proximal to the infolding joints. Summary of Insect Leg Patterning Based on Drosophila Drosophila legs are initially patterned by the same coordinate system that provides positional values within the segments. After this positioning on the body wall, the leg axis is patterned. However, this axial patterning does not directly regulate morphologically defined structures. For example, genes that control different PD regions of the limb do not map to specific limb segments or branches. Instead, the coordinate system appears to be generic: it provides cells with positional values along PD, DV, and AP axes but does not directly specify adult morphology. Indeed, this generic feature of many patterning networks allows them to be conserved in evolution and to pattern a range of morphologies by modifications of downstream regulation. Thus, although taxon sampling is sparse, it appears that the genes that pattern the PD

81

82

Functional Morphology and Diversity leg axis—Distal-less, dachshund, extradenticle —are conserved in other insects and used to pattern legs from flies to beetles to grasshoppers. How far does this conservation extend in crustaceans?

REVIEW OF PD PAT TERNING IN CRUSTACEAN LIMBS Some features of the generic patterning network discovered in Drosophila are found in crustaceans. This supports the idea that a generic limb-patterning network is used throughout the arthropods, within which different limb morphologies can be specified. However, not all features of the regulatory network described above are conserved and found in most taxa sampled. The two main conserved features are the positioning of limb buds at an AP boundary within the segment and the initiation and early subdivision of the PD axis. The crustacean data are represented in Table 3.1, which includes all the leg-patterning genes known from crustaceans. Positioning of Limb Buds at the AP Segment Boundaries Positioning of the limb bud along the AP boundary within a segment is found across crustaceans (Fig. 3.4; reviewed in Williams and Nagy 2001). As in all arthropods, engrailed expression marks the posterior part of each segment within crustaceans (e.g., Patel et al. 1989, Scholtz et al. 1993, 1994, Queinnec et al. 1999, Abzhanov and Kaufman 2000). In malacostracans, in which the segmentation of the body and the early development of the limb bud can be followed via a predictable cell lineage, the cells that will produce the limb bud straddle the AP boundary within the segment. Double labeling of Distal-less and engrailed in biramous limbs of malacostracans shows that Distal-less is initiated just anterior to the engrailed boundary (Parhyale hawaiiensis, Browne et al. 2005; Orchestia cavimana and Porcellio scaber pleopods, Hejnol and Scholtz 2004; Thamnocephalus platyurus, T.A. Williams, unpublished observation). The signaling genes wingless and decapentaplegic are virtually unsampled in crustaceans. In the one case that wingless has been examined, the phyllopodous limbs of the branchiopod Triops longicaudatus, it appears, just like in insects, to be positioned just anterior to engrailed expression (L.M. Nagy, personal communication). The Initiation and Early Subdivision of the PD Axis The most extensively sampled leg-patterning genes in crustaceans are the leg gap genes. Of those, Distal-less has been most widely sampled (Table 3.1) and has been examined in representative crustacean limbs of quite distinct structure: uniramous, biramous, and phyllopodous. In uniramous limbs of peracarids, Distal-less is expressed on the body wall in a relatively small cluster of cells that eventually form the tip of the outgrowing limb bud (thoracic limbs of Porcellio scaber, Abzhanov and Kaufman 2000, Hejnol and Scholtz 2004; Orchestia cavimana, Hejnol and Scholtz 2004). In biramous thoracic limbs, the cluster of Distal-less expressing cells is proportionately somewhat larger but also forms an unbroken zone of expression on the body wall. This cluster will eventually subdivide to form the endopod and exopod (Mysidopsis bahia, Panganiban et al. 1995). This also occurs in biramous abdominal limbs: in two other peracarids, the amphipod Orchestia cavimana and the isopod Porcellio scaber, biramous abdominal pleopods initially develop Distal-less-expressing limb buds that subsequently subdivide to form endopod and exopod. That is, both branches are part of the distal leg domain (Hejnol and Scholtz 2004). Interestingly, this pattern is also found in phyllopodous limbs both in branchiopods (Williams 1998, 2008, Williams et al. 2002) and in phyllocarids (Williams 1998). In both groups, an initial

Table 3.1 Overview of limb-patterning genes examined within crustaceans. Class/species

Signaling genes wg

Branchiopoda Artemia franciscana Thamnocephalus platyurus Triops longicaudatus Cyclestheria hislopi Leptodora kindtii Daphnia magna Maxillopoda Sacculina carcini Malacostraca Nebalia pugettensis Mysdopsis bahia Porcellio scaber Orchestia cavimana Parhyale hawaiensis Pacifastacus l eniusculus

✓9

dpp

Leg “gap” genes

Dll

exd

✓1, 2 ✓7, 8

✓2 ✓7

✓7, 10 ✓12 ✓12 ✓13

✓7

hth

✓11

Leg segmentation genes

dac

✓11

N

Dl

“Wing” genes

“Trachea” genes

nub

vvl

✓3

✓ 4, 5

✓6

Preliminary

✓

✓14 ✓10 ✓1 ✓15, 16 ✓16 ✓18

Body segmentation gene engrailed

✓15 ✓17

✓15

✓15

✓17

✓15

✓16 ✓16 ✓18

✓3, 5, 19

Gene symbols: wg , wingless; dpp, decapentaplegic; Dll , Distal-less; exd , extradenticle; hth , homothorax; dac , dachshund; N, Notch; Dl , Delta; nub, nubbin; vvl , ventral veinless. References: 1Panganiban et al. 1995; 2Gonzalez-Crespo and Morata 1996; 3Averof and Cohen 1997; 4Mitchell and Crews 2002; 5Franch-Marro et al. 2005 6Manzanares et al. 1996; 7 Williams et al. 2002; 8Williams 2008; 9Nulsen and Nagy 1999; 10Williams 1998; 11Sewell et al. 2008; 12Olesen et al. 2001; 13Shiga et al. 2002; 14Mouchel-Vielh et al. 1998; 15Abzhanov and Kaufman 2000; 16Hejnol and Scholtz 2004; 17 Prpic and Telford 2008; 18Browne et al. 2005; 19Damen et al. 2002.

84

Functional Morphology and Diversity

Fig. 3.4. The size of the limb primordia is highly variable in crustaceans: schematic views of ventral segments in different crustaceans illustrating the relative amount of ventral segment fated to become limbs. In all cases, primordia retain their anterior-posterior positioning regardless of size (see text for examples).

domain of Distal-less expression forms on the ventrolateral body wall. Like the biramous domain, it eventually subdivides to form the endopod and exopod. (However, subsequent Distal-less expression is more widely distributed; see below.) Thus, throughout crustaceans—as throughout all arthropods sampled—Distal-less expression is consistent with the role of patterning the distal region of the leg and promoting PD outgrowth. For the other leg gap genes, extradenticle and dachshund, published data exist only for uniramous and phyllopodous limbs. Where they have been examined in crustaceans, extradenticle is expressed in a proximal domain complementing and exclusive of the Distal-less domain, indicating a conserved role for the early stages of limb patterning.2 In uniramous limbs, extradenticle is expressed in the first two limb segments (Porcellio scaber, Abzhanov and Kaufman 2000). In early phyllopodous limb bud, extradenticle is expressed in the limb bud outside the Distalless domain, but later, when Distal-less is expressed more proximally, extradenticle expression is maintained so that the two are coexpressed (Triops longicaudatus and Thamnocephalus platyurus, Williams et al. 2002). Like extradenticle, dachshund expression in uniramous crustacean limbs is similar to that in Drosophila. It arises between the distal Distal-less and proximal extradenticle domain (Porcellio scaber, Abzhanov and Kaufman 2000). In phyllopodous limbs, one domain of initial dachshund expression is just proximal to the initial Distal-less expression, as might be expected (Sewell et al. 2008). However, the expression is dynamic and becomes more complex

Mechanisms of Limb Patterning in Crustaceans as the limbs develop (see below). Taken as a whole, the patterns of expression in leg gap genes in crustaceans are similar to Drosophila: initially, proximal and distal domains are established, and subsequently, new domains are intercalated that can provide more precise PD information. Notch in Crustaceans Although there are no published reports of expression of Notch pathway genes in crustaceans, there are preliminary data that Notch protein is expressed in reiterated stripes in developing endites in branchiopods and functions in proper endite formation (T.A. Williams, unpublished observation). Notch is expressed differentially in the limb, in the medially repeated lobes (endites) but not the lateral lobes. DAPT, a gamma secretase inhibitor that blocks Notch signaling, can block formation of endites in branchiopod legs. In the most extreme phenotype, all endites (except the most proximal one) fuse to form a single unbranched lobe. Again, the effect within the limb is not uniform, fusing medial lobes (endites) but not lateral lobes. Summary All crustacean limbs currently sampled show conservation in AP limb bud positioning and initial PD patterning. In a sense, it is remarkable that a generic PD coordinate system is used throughout arthropods given their limb diversity. At the same time, the conservation of patterning mechanisms in the face of great morphological diversity makes PD patterning genes surprisingly unhelpful in explaining how limb morphology may have evolved; known mechanisms do not specify whether the limbs will be fundamentally uni- or biramous, for example, nor do they pattern the medial and lateral lobes. In short, patterning genes known from Drosophila cannot as yet explain any of the differences so prevalent in crustacean limbs. In particular, the fact that crustaceans have legs quite distinct from uniramous walking legs gives rise to questions about how biramous or phyllopodous legs are patterned. How Are the Two Primary Limb Branches Patterned? Patterning the primary limb branches is a fundamental and yet unanswered question in crustaceans. Everything we have learned to date points to a single PD patterning axis with a single domain of Distal-less from which both the endopod and exopod arise. That is, there is no evidence to indicate that the formation of the endopod and exopod within the distal domain is controlled by genes known from Drosophila leg patterning. More generally, outgrowths from the limbs are not patterned as new PD growth axes (reviewed in Williams 1998, Nagy and Williams 2001, Williams and Nagy 2001). However, patterning of branches from the main axis is particularly important regarding the formation of the endopod and exopod because biramous limbs are both ancestral and widespread among crustaceans. While loss of Distal-less function in crustaceans should produce loss of both the endopod and exopod, we do not know which genes would prevent splitting while permitting outgrowth. In addition, although we know that limb branches are not patterned as simply reiterated PD growth axes, we do not know to what degree the bifurcation of the endopod and exopod is distinct from the regulatory mechanisms that form other limb outgrowths (exites and endites). Phenomenologically, the formation of the two branches can vary among species. Fig. 3.5 shows some scenarios for morphogenesis in biramous limbs: both limb branches might arise simultaneously from the body wall, or limb branches might form from a subdivision of an already elongating limb bud, or a second branch might arise late in morphogenesis as an outgrowth from the main branch. There is some evidence that all three modes are used in crustaceans.

85

86

Functional Morphology and Diversity

Fig. 3.5. Schematic of possible types of limb morphogenesis leading to biramous limbs. (A) The limb bud arises from the body wall bilobed and continues to extend both branches. (B) The limb bud arises unbranched from the body wall and only after some growth subdivides and forms two branches. (C) The limb bud arises unbranched from the body wall, grows, and then forms a small bud that subsequently grows as the second branch.

What controls patterning in biramous crustacean limbs is unknown, and there are no cross-species comparisons of the development of biramous limbs. However, there have been comparisons within peracarid species of the development of biramous abdominal limbs to their uniramous thoracic counterparts. These comparisons depend on the fact that peracarids establish their segments (and early limb buds) via a repeatable cell lineage. Therefore, it is straightforward to compare the fates of cells that formed the exopod in the biramous abdominal limbs to the same cells on the thorax. In the amphipod Orchestia cavimani, Hejnol and Scholtz (2004) found the cells that are homologous to abdominal exopod cells indeed express Distal-less early on but then subsequently lose Distal-less expression. This suggests that these cells still have an early specification to form an exopod—or at least distal leg structures—but that that specification is lost during subsequent development. By contrast, in the isopod Porcellio scaber, cells in the thorax that are homologous to abdominal exopod cells never express Distal-less (Hejnol and Scholtz 2004). This suggests that these cells never acquire a distal limb fate at all. Thus, in taxa

Mechanisms of Limb Patterning in Crustaceans that independently evolved uniramous from biramous limbs, the mechanisms controlling suppression of the exopod in the thorax appear distinct. In related studies, Wolff and Scholtz (2008) marked cells in the early uniramous thoracic limb bud and the biramous abdominal limb bud in the amphipod Orchestia cavimani, in order to follow the subsequent fate of those cells as the embryos grew. Before limb outgrowth, body segments contain rows of aligned cells approximately 10 cells wide. In the abdomen, 6 of 10 cells contribute to the limb bud and do so without much change of register; for example, the medial cells form the endopod, and the lateral cells form the exopod. In the thorax, cells also grow out in register, although fewer cells of the row form the leg. Notably, the cells that form the exopod in pleopods form the lateral part of the walking thoracic leg. This suggests that, in these amphipods, the formation of an unbranched limb evolved from the suppression of the mechanism that would split the initial domain into endopod and exopod. These data are very interesting and, in the absence of genetic manipulation, allow some speculation on how a single axis forms two axes. Of course, in these cases, one of two branches is lost during evolution, leaving unanswered the original question—how two branches are patterned from a single PD domain. It is also unclear how much these results can be generalized. The lineage results that lead to the conclusion that there is a suppression of the split into a biramous limb are derived from Orchestia cavimani (Wolff and Scholtz 2008). However, the earlier comparison of Distal-less expression within Orchestia cavimani and Porcellio scaber suggests that, even within the peracarids, exopod loss occurs via different mechanisms (Hejnol and Scholtz 2004). Also, malacostracans are unique in having teloblastic growth and cell-lineage-based formation of the germband and limb buds. Whether their mechanisms of limb patterning are similarly derived remains to be seen. Common Aspects of Patterning in Phyllopodous Limbs of Branchiopods Phyllopodous limbs are even more distinct than biramous ones from the kind of uniramous walking leg represented in most models of arthropod limb patterning. Despite the previously mentioned conservation of some aspects of limb patterning, phyllopodous limb buds have complicated temporal and spatial expression of the PD leg-patterning genes. In addition, both branchiopods and phyllocarids have large thoracic limb buds that occupy virtually the entire ventral to ventrolateral body wall. For example, in Artemia the limb bud occupies six of the eight precursor rows that form the AP segmental anlage, whereas the interleg region is only two of eight rows (Freeman et al. 1992). This contrasts with peracarids, where the initial limb buds occupy less than half the segment (e.g., Dohle and Scholtz 1988, Scholtz 1990). These differences in relative size of the limb anlagen, as well as differences in the geometry of the developing limbs, argue against hypotheses of straightforward conservation of limb patterning since such differences would certainly influence the action of, for example, short-range signaling molecules. Additional arguments arise simply from the complex patterns of gene expression found in phyllopodous limbs. As mentioned above, expression of the leg signaling gene wingless has been examined in the branchiopod Triops longicaudatus and is expressed initially in segmental stripes along the body anterior to the engrailed stripes. However, as the limbs develop on the segments, each wingless stripe breaks up, eventually coming to occupy only a portion of each lobe of the phyllopod (Nulsen and Nagy 1999). The wingless expression varies by lobe: on the most medial lobe (the gnathobase), expression is on the lateral margin; on the other endites as well as the endopod and exopod, the expression is on the medial margin; and on the epipod, the expression outlines the entire lobe. Similarly, expression of Distal-less and extradenticle in branchiopods is initially comparable to Drosophila: Distal-less is expressed in the distal limb, in both the developing endopod and exopod, exclusive of and surrounded by extradenticle expression. However,

87

88

Functional Morphology and Diversity wingless dachshund Distal-less

epi gn

exo

E2 E3 endo E5

E4

Fig. 3.6. Schematic of gene expression in the early limb bud of the branchiopod Triops longicaudatus. With the exception of the gnathobase, the endites show identical patterns of expression. All other lobes are unique. Abbreviations: E2–E5, endites 2–5; endo, endopod; epi, epipod; exo, exopod; gn, gnathobase.

as development proceeds, Distal-less expression spreads into the more proximal lobes—with particularly extensive expression in the endites—and this later expression of Distal-less overlaps with extradenticle. When this later Distal-less expression is traced through limb development, it coincides exactly with seta-forming cells of the limb (Williams 2008). Early expression of dachshund in the Triops limb bud is consistent with an intermediate dachshund region between the Distal-less- only expression domain and the extradenticle domain of the proximal leg. However, expression occupies only a small region of cells in a large limb bud, and the topology that would correspond to a uniramous limb can be only roughly inferred. In addition, dachshund is expressed in a proximal stripe and, later, in a series of medially reiterated stripes in each endite (Sewell et al. 2008). Indeed, one striking feature of many limb-patterning genes in phyllopodous limbs is that they show reiterated domains of expression in the endites. In Triops, the branchiopod that has been examined most extensively, the endites between the gnathobase and endopod show identical expression patterns for wingless, Distal-less, extradenticle, dachshund, and Notch (Fig. 3.6). This is true in another branchiopod, Thamnocephalus platyurus, although the dachshund and Notch data are preliminary. Below I argue that a medial patterning system controls medially reiterated structures in crustacean limbs. Summary In Drosophila a generic coordinate system provides positional information for developing limbs, and aspects of this generic system appear conserved even in limbs of highly different morphology, that is, initial PD patterning in phyllopodous limbs. However, this fact alone highlights how little these positional coordinates inform us about morphology later in development: phyllopodous limbs are so structurally different from uniramous limbs that sharing the same axial PD

Mechanisms of Limb Patterning in Crustaceans patterning must relegate that patterning to only the most indirect control of adult morphology. Within this context, I propose a modified view of limb patterning in crustaceans and describe how that view accounts for some patterns of diversity.

LIMITS OF PD PAT TERNING MODEL IN EXPLAINING THE DIVERSIT Y OF LIMB MORPHOLOGIES AND AN ALTERNATIVE VIEW OF LIMB PAT TERNING One point emerges clearly from the analysis of Drosophila leg-patterning genes in crustaceans: interactions of those genes alone cannot provide a satisfactory model to explain the diversity of adult limb morphologies in crustaceans. One particular point is worth reiterating since it is central to crustacean limb morphology. We do not know how the two main branches of a biramous limb are patterned. This is important because the biramous limb is most likely ancestral (see chapter 2). Beyond that, biramous thoracic limbs are very common among crustaceans, although they happen not to occur in the few crustaceans typically used for studies of limb patterning. Beyond the question of biramous limbs, crustaceans have a variety of endites and exites not accounted for by the model based on patterning in Drosophila. For example, Boxshall (2004) distinguishes as many as 9–10 distinct nonhomologous exites among crustaceans. In addition, on medial limb margins, most anterior feeding thoracic limbs have well-developed endites used for sorting and breaking up food. In branchiopods, the phyllopodous limbs are defined by their well-developed endites and exites. These patterns among crustacean limbs lead to an alternative view of crustacean limb patterning based on certain contrasts between Drosophila and crustaceans (Fig. 3.7). Model for Ancestral Limb Patterning First, the amount of ventral and ventrolateral body wall devoted to the limb bud and the timing of adult limb development vary. In Drosophila, the leg develops from a very small set of cells patterned in the embryo that then invaginate and develop segregated from the larval epidermis (see above). Through further proliferation and patterning, these cells produce the adult leg morphology. By contrast, no crustaceans set aside part of the body wall epidermis for limb formation. Limbs develop directly from the body wall in coordination with adjacent tissue. In some cases the amount of body wall devoted to the adult limb is relatively small, but in some cases it is not. In branchiopods, virtually the entire ventrolateral body wall becomes limb. This is likely to represent the ancestral case for pancrustaceans since stem lineage larval forms appear to have developing limbs that are large relative to the body wall. In parallel with these differences in the timing and extent of ventral body wall involvement in leg development, crustaceans show an enormous variety of limb morphologies compared to the single unbranched leg in Drosophila. Laterally, exite morphology is highly variable. Medial limb morphology is diverse as well but often is organized around a series of medially reiterated structures: endites, leg segments, or some combination of those. These considerations lead me to develop a hypothetical model for how limb-patterning genes were deployed ancestrally in Pancrustacea and modified in extant taxa (Fig. 3.7). I hypothesize that, ancestrally, the limb-patterning field was broad and occupied a large area of the ventral to lateral body wall. Within that patterning field, a number of patterning networks operated but were only loosely coupled with one another. An overall PD patterning axis promoted limb outgrowth and controlled the formation of both the endopod and exopod. This is supported by the fact that all crustaceans have limb buds with a region of exclusive Distalless expression that gives rise to the exopod and endopod. In concert, a medial patterning

89

90

Functional Morphology and Diversity Drosophila

field of limb patterining on body wall

adult limb morphology

crustaceans

segregation early in development from the rest of the ventral body wall

continuous functional integrity of the ventral body wall

unbranched leg

high diversity of morphologies, e.g., biramous, phyllopodous

HYPOTHESIS L limb pattering systems

P

N

D

L

P

N

D

PD and N patterning are strictly superimposed; lateral genes (used to specify the wing) maintain the same topological relationship but are not in the limb field.

Patterning axes are only loosely coupled. PD patterns main axis of leg (1 or 2 branches); N patterns medially reiterated structures (lobes or segments); L patterns lateral limb lobes.

Fits rapid life cycle highly specialized for ecological niche

Flexibility of patterning permits high limb diversity

DV homologies are comparable along PD axis

Medial/lateral homologies can vary independently along PD axis

consequences

Fig. 3.7. Comparison of a number of features of Drosophila leg development and crustacean leg development illustrating the hypothesis that patterning systems now coupled in Drosophila may have been only loosely coupled ancestrally. The rectangles represent a half segment, for example, from the midline outward. Gray shading represents the size of the early limb primordia relative to that half segment. N, L, and PD represent gene patterning networks: Notch, lateral, and proximodistal, respectively. DV, dorsal-ventral axis.

system based on Notch signaling produced repeated structures. Those structures could be either endites or segments. This is supported by preliminary data of Notch expression and function in branchiopods. Laterally, another set of genes (nubbin, apterous) operated to produce exites (Averof and Cohen 1997). These sets of loosely coupled patterning networks would have high f lexibility to produce the various limbs found in extant groups. In Drosophila , those loosely coupled networks are now tightly coupled: the PD and Notch networks are linked, and

Mechanisms of Limb Patterning in Crustaceans the lateral genes are expressed in cells that undergo an early migration away from the rest of the limb anlagen. In uniramous limbs, Notch regulates leg segments; in the ancestral case, it regulated any medially repeated structures. In the derived, uniramous limb of Drosophila , Notch and PD patterning are linked, the limb bud is a reduced part of the body wall, and much more of the limb bud is restricted to the initial distal region demarcated by the PD patterning system. This contrasts with the much more extensive limb buds found in branchiopods and in stem group crustaceans. Consequences and Predictions This hypothesis suggests a rethinking of how development of limbs may have evolved and forces a somewhat atypical view of the developmental basis of limb homologies. In particular, it predicts that homologs can vary independently on the medial versus lateral margins along the PD axis; that is, exite morphology varies independently from endite morphology. This is consistent with the fact that lateral limb branches are highly variable (Boxshall 2004). This hypothesis implies that any particular slice along the PD leg axis cannot be considered as an integrated homolog the way we conceive of a PD series of limb elements in the vertebrate limb. It also implies that Notch regulates all medially repeated structures, that within any limb medial structures are serial homologs, and that between limbs certain dissimilar medial structures, lobes, and segments are direct homologs. (Despite these elements being direct homologs, it may not be possible to draw one-to-one homologies between elements, a common problem with reiterated serial homologs [Bateson 1894, Wagner 1989, Van Valen 1993].) This hypothesis also leads to specific predictions in taxa outside of crustaceans. For example, in insects, the gnathal appendages have repeated medial lobes (endites). I predict that these lobes will be regulated by Notch signaling as are leg joints. This has not been examined in Drosophila because of their highly reduced mouthparts.

THE IMPACT OF DIVERSE LARVAL STAGES ON MODEL S OF LIMB DEVELOPMENT In the preceding sections, the focus was on adult thoracic limb diversity from a standard perspective in evolutionary developmental biology: how is development modified to produce one adult leg morphology versus another? Although this is normally how comparisons between species are framed, this narrow focus is justified only in the case of crustaceans that develop their adult limbs directly. However, the majority of crustacean taxa undergo some metamorphosis. In these cases, a particular thoracic segment can, for example, produce a series of limbs with distinct morphologies during successive molts. As described above, adult limb morphology may be achieved only after limbs of quite different morphology have developed on the same segment. Why is this significant? Consider the differences in limb development in direct versus indirect developers that are illustrated schematically in Fig. 3.8A. For direct developers, each species produces its own particular adult morphology in a single developmental pathway from primordium to adult limb. This type of life history lends itself to the question commonly posed by current research: How do different morphologies develop? By contrast, crustaceans with distinct larval stages develop the adult limb morphology via a sequence of potentially distinct developmental pathways. In these cases, patterning of the adult limb is built upon prior functional limbs that are transformed in a variety of ways. For example, development to the adult limb may occur gradually or as a marked modification of a preexisting larval limb, or even as a wholesale replacement after a larval limb is lost.

91

92

Functional Morphology and Diversity

Fig. 3.8. Conceptual differences in modeling limb development in direct developers versus those with multiple functional larval stages, as are found in many crustaceans. (A) For direct developers, rows represent a single thoracic segment in each of three different species. The adult thoracic limb morphology is different in each species (as represented by different shading patterns). The typical question posed by current research is how these different morphologies develop. However, in crustaceans with functional larvae, the three rows of different species have more complex pathways to reach the adult thoracic limb morphology. Larval stages are represented in which limbs may be only slightly different from the adult (row 1), quite distinct from the adult (row 2), and even be lost (no color) and regained (row 3). (B) The impact of multiple functional larval stages is similar for considerations of patterning segment identity along the body axis. For direct developers, rows 1–3 represent three different species with different patterns of functional clusters of limbs (represented by different shading patterns along a series of segments). The typical question posed by current research is how these different patterns of segment identity develop. However, in crustaceans with functional larvae, the developmental pathway leading to the adult pattern is more complex since functional larvae can have distinct clusters of limb identities. In both cases, these changes in morphology over time need to be accounted for in models of limb development.

The series of regulatory control mechanisms that might pattern such diverse routes to adult limb morphology is wholly unaccounted for in models of limb patterning. At the least, analysis of adult limb patterning should include every larval stage and the expression and function of genes in each stage. As yet, patterning of different limb morphologies found in a complex larval sequence remains unexplored. An important and related question is whether, on any one particular segment, the diversity in limb structure is somehow constrained during

Mechanisms of Limb Patterning in Crustaceans development. That is, are only some morphological transformations possible from one molt to the next? These differences between direct and indirect developers also influence our models of how segment identity develops. Comparative data exist for the patterning of segment identities that define body tagmata (reviewed in Deutsch and Mouchel-Vielh 2003). Genetic regulation of different body regions is provided by homeotic (Hox) genes, which show differential expression and control segment identity along the AP body axis. Hox genes are found in all the major crustacean taxa, and differences in their expression control the fate of anterior thoracic segments, for example, determining what segments bear maxillipeds (Averof and Patel 1997, Pavlopoulos et al. 2009). However, the same phenomena that complicate the patterning of adult limb morphology on any one segment can obviously take place anywhere along the body axis. Therefore, the patterns of segment identity within tagmata can change during development (schematized in Fig. 3.8B). Whereas the standard question is to ask how the tagmata in species A are patterned differently from the tagmata in species B, for most crustaceans larval stages with different groupings of limb morphologies present a more complex sequence of tagmata that presumably would require differential regulation at each stage. The presence of larval forms implies that groupings of similar segments seen in the adult may not have been stable throughout the life cycle. In larval forms, patterns of segment identity are much more complex since larval limbs can be structurally distinct from adults. I do not suggest that the underlying tagmata are not stable, merely that the clustering of similar limbs—that is, the apparent segment identity—can change during development thus complicating our models of how segment identity is specified. Despite the absence of patterning data that bear on these questions, I discuss and illustrate them at length because the sheer variability in limb morphology and life history patterns within crustaceans is generally not incorporated into models of arthropod limb patterning. This diversity is key to understanding the radiation of crustaceans, but our models do not account for it.

ANALYZING THE DEVELOPMENT AND DIFFERENTIATION OF LIMB STRUCTURES Models of limb patterning mainly explain how spatial pattern arises during development, typically through the formation of Cartesian axes: proximal-distal, anterior-posterior, and dorsal-ventral. In these models, actual limb structures are a proxy for position; for example, claws at the limb tip become most distal positional values. Although this has been a very useful perspective in illuminating how certain generic regulatory pathways can be used in different situations, it tends to obscure morphological structures. After all, it is specific structures that function in the particular habits and life history of each species. Considered in their own right, the structures that compose the leg display a number of intriguing developmental features. For example, some setae show regularities in positioning that transcend species, such as their position adjacent to joints. Also, many of the genes that provide axial positional information in the leg also participate directly in patterning morphological structures. It seems quite plausible that patterning of leg axes and morphological structures is more integrated than expected; we are slightly misled by Drosophila , where the timing of patterning and differentiation is atypical. In most crustaceans, patterning and differentiation are, at least phenomenologically, quite closely integrated—giving rise to the question of whether they are mechanistically integrated as well. If this is accurate, understanding the development of limb structures could elucidate novel morphological subunits within limbs. For example, fields of setae or setal joint complexes that are linked developmentally would constrain the possibilities of limb evolution.

93

94

Functional Morphology and Diversity In addition to exploring this hypothesis of novel morphological subunits, a more straightforward and pragmatic reason exists for studying descriptive limb morphogenesis. To interpret patterns of gene expression or phenotypes produced by perturbations of gene function, we must know the details of tissue morphogenesis. As the model system, Drosophila has had much of the descriptive groundwork done: the timing of differentiation is known, and cell types are often well characterized. By contrast, that groundwork is lacking in crustaceans. However, to correctly interpret details of gene expression and function, we need to know what tissues these differentiating cells will become. This is particularly important because many patterning genes have multiple functions, and dissecting out those functions requires knowing the fate of cells expressing the gene. To consider whether developmentally novel subunits exist in crustaceans, I discuss development and patterning of muscles and setae within crustacean limbs. I again use Drosophila as a point of reference, although it is important to note that the presence of the pupal stage has obscured analysis that might integrate aspects of limb patterning with setal and muscle development in that species. Various aspects of limb development in Drosophila have been compressed to fit the demands of radical metamorphosis. In particular, the timing of differentiation into the adult relative to patterning of the leg is not characteristic of most crustaceans. This is best grasped schematically, as in Fig. 3.9, where, if we compare events along a normalized timeline, it is obvious that the lack of larval limbs and the radical metamorphosis in Drosophila create an atypical distribution of the timing of patterning versus differentiation compared to most crustaceans—indeed, most arthropods. Cell differentiation occurs along with ongoing positional patterning in crustaceans, as illustrated by the development of the limbs in branchiopod crustaceans. In branchiopods, the earliest differentiation of the limb bud from the body wall is followed by differentiation of both setae and muscles (Williams and Muller 1996, Williams 2007a). The possible integration of patterning and differentiation remains an open question in crustaceans. LIMB DEVELOPMENT

Drosophila

embryo

larva

pupa adult

PATTERNING DIFFERENTIATION

crustacean

embryo

larva

larva

juvenile

adult

PATTERNING DIFFERENTIATION

Fig. 3.9. A schematic timeline of limb development showing the relative timing of patterning and differentiation in Drosophila and crustaceans. In crustaceans, as is likely ancestral, patterning and differentiation are highly overlapping during development. In Drosophila, the constraints of a radical metamorphosis segregate the two events to a marked degree.

Mechanisms of Limb Patterning in Crustaceans Development of Muscle in Crustacean Limbs Muscle in Drosophila and other insects arise from precursor cells that differentiate into muscle founders. These founder cells act as positional templates, attracting other myotubes to fuse with them to form functional muscle. However, in Drosophila , a radical metamorphosis divides muscle formation into two phases: larval and adult. Larval muscles develop directly during embryogenesis. Some adult muscle precursors are specified during embryogenesis, but they become fully developed only during the pupal metamorphosis (Roy and VijayRaghavan 1999). In Drosophila legs, muscles originate from a population of precursors (5–10 myoblasts) associated with the embryonic leg disc primordia. These cells proliferate during the larval period to form about 500 myoblasts associated with the disc epithelium. During late larval and early pupal development, some of these myoblasts begin expressing dumbfounded , an immunoglobulin that attracts other myotubes to fuse with the founder cells and thus serves as a marker of founders. As pupation proceeds, internal tendons form closely associated with founder cells, myotubes fuse with the founders, and attachment sites form on the leg epithelium (Soler et al. 2004). How do crustacean muscles develop? This question is largely unexplored, although one recent paper offers the first evidence that founder cells are present in crustaceans as well as insects. Kreissl et al. (2008) generated a monoclonal antibody against heavy-chain myosin in isopod crustaceans and then traced myosin-expressing cells during development. They found that muscle development proceeded in a fashion similar to that in insects. Initially, only single cells expressed the antigen, and these were arranged in a pattern that was a precursor to the adult pattern. Subsequently, these cells became multinucleate syncytia, although it was not possible to distinguish whether this occurred via fusion (Kreissl et al. 2008). Development of Setae in Crustacean Limbs Crustacean setae are enormously diverse (as described in chapter 6). If external morphology, accessible through light and scanning electron microscopy studies, remains underdescribed, the internal morphology and development of setae are even less explored. Although setae are assumed to be homologous among arthropods, we have no general model for noninsect setae for comparison. In insects, sensilla that form bristles (setae) and other sense organs develop through a well-described lineage mechanism from a single precursor epidermal cell (reviewed in Hartenstein 2005). Stereotyped divisions of the precursor cell produce both sensory neurons and accessory cells that ensheath the neuron or form cuticular outgrowths. Modifications of this lineage via apoptosis and/or repeated divisions in certain lineages can produce all the sensilla types found in insects (Lai and Orgogozo 2004), including setae without sensory function (via apoptosis of the lineage leading to the neuron). Furthermore, whereas insect sensilla can be grouped into two or three main categories based on adult morphology (Hartenstein 2005), crustacean sensilla appear to have a much broader range of underlying cell numbers (Hallberg and Hansson 1999). How do crustacean setae develop? To my knowledge, only two studies describe events of the cellular morphogenesis of crustacean setae. Based on transmission electron microscopy, Guse (1983) described the development of the aesthetascs, tubular sensory setae, on the antennule of the mysid Neomysis integer. The aesthetascs have 60–80 sensory cells and eight accessory cells that ensheath the neurons. Sensory cells are differentiated before the molt, during which the shaft is formed. The shaft begins to develop during apolysis with a retraction of the epidermis from the old cuticle. The shaft develops by forming a cylindrical invagination of the epidermis

95

96

Functional Morphology and Diversity with the midpoint of the shaft being the deepest part of the invagination. The eight ensheathing cells that secrete the shaft cuticle are arranged in a telescoping manner from proximal to distal such that the cuticle of the distal shaft is laid down by one ensheathing cells. with more proximal ensheathing cells laying down more proximal cuticle. In the second study, setal morphogenesis was described for the anostracan Thamnocephalus platyurus using light microscopy: differentiating setal cells could be seen within the developing thoracic limb bud (Williams 2007b). The thoracic limbs in T. platyurus are highly setose, although almost none of the setae are sensory. The only sensory setae present are mechanoreceptors on each endite. In parallel to the anamorphic development of the larva in anostracans, limb buds and setae on the limb buds develop gradually. Even small limb buds just emerging from the body wall form the lobes of the adult limb and begin to develop setae with very small shafts. The cellular composition of the nonsensory setae is uniform: all have six accessory cells, discernable as three pairs whose size and positions are distinct. These cells are clearly identified in the developing limb bud by the hypertrophy and alignment of their nuclei. Although the pairs of accessory cells have distinct nuclei and positions relative to the setae, their functions remain unknown. Even these two studies show that very different numbers of cells are involved in setal production. Neither illuminate the lineage of setal forming cells. Indeed, there are no lineage-tracing studies of cells that generate crustacean setae, and we do not know whether a lineage-based development comparable to insects occurs. Comparisons of setal development within crustaceans would be very useful in their own right, and they would help address some of the bigger questions of the evolution of setae in arthropods. For example, although nonsensory bristles in insects can be derived from sensory ones by apoptosis of the neuronal lineage, is there reason to suppose that sensory setae actually evolved first? That is, might not there have been a selective pressure for structural setae just as there was for sensory accessory cells? In general, the origins of arthropod setae remain unexplored. Genes Important in Limb Patterning Also Pattern Setae Another reason to consider the development of crustacean setae is that a number of leg-patterning genes contribute to patterning bristles or sense organs. This is not surprising since many patterning genes have multiple functions. Nonetheless, it is interesting to consider whether dual patterning might be a consequence, at least ancestrally, of coordination within the leg of axial patterning and setae. For example, two of the main leg “gap” genes in Drosophila—Distal-less and dachshund—additionally play a role of regulating sensory structures. Distal-less plays a role in the development of certain sensory structures in Drosophila. Distal-less-negative clones induced late in development during the third larval period in Drosophila are often associated with anomalous bristle morphology (although it is not reported that cells specifically express Distal-less; Campbell and Tomlinson 1998). Distal-less is required in cells that form bracts, which are cuticular elaborations adjacent to the mechanosensory bristles of the leg (Held 2002). In addition, Distal-less specifies Keilin’s organs, the larval leg sense organs that consist of three bristlelike external sensilla (Cohen and Jü rgens 1989). Indeed, the earliest ideas of the multifunctionality of Distal-less involved co-opting an ancestral function in the nervous system for use in body wall extensions (Panganiban et al. 1997, Panganiban 2000, Mittmann and Scholtz 2001, Williams et al. 2002). Mutations in dachshund, another leg “gap” gene, are known to cause fusions of leg segments. However, they also change the number and distribution of bristles (Mardon et al. 1994). Similarly, Notch signaling, which helps position joints along the PD leg axis, is also used in the early dete mination of sense organs (reviewed in Hartenstein 2005). In addition, pox-neuro

Mechanisms of Limb Patterning in Crustaceans and BarH, two genes that play a role in forming tarsal joints, can also transform sense organs (Awasaki and Kimura 2001). Thus, numerous genes that play a role in limb patterning also play a role in forming sensory structures. Are Evolutionarily Relevant Substructures within Limbs Formed by Developmentally Integrating Multiple Morphological Elements? Given this kind of dual function in genes that pattern leg axes and sensory structures, it seems possible that these two types of patterning were more closely linked ancestrally than they are in Drosophila, with its highly specialized metamorphic development. Such linkage might be found in arthropods like crustaceans that develop limbs directly. Certain patterns in crustacean leg morphology could support this idea. For example, one notable regularity within limbs is the position of setae at the distal margin of the limb podomere. There is a strong functional reason for this position given that such setae can transmit information about the relative position of the two limb joints. This function might be ancestral and would provide an apt selective scenario for joints and sensory structures to be developmentally linked. Another morphological regularity often found in crustacean limbs is the one-to-one correspondence between the annulation and projecting setae in limbs that are multiannulate. Given that mutations in genes that regulate joint formation can also change setal patterns, it seems plausible that the regulation of joints and adjacent setae is mechanistically linked. Robust models of setal patterning in crustaceans are a first step toward evaluating this hypothesis. Although the examples of structural linkage I describe are hypothetical, they point to the fact that understanding such developmental linkages could provide new insights into homologies within crustacean legs, as well as the pathways of limb diversification.

FUTURE DIRECTIONS Comparative limb development is a wide open field in crustacean evolutionary morphology. As shown at many points in this chapter, crustacean limbs are highly diverse both in their basic morphology and in the transformations they undergo during the life cycle. This diversity distinguishes them from most other arthropod groups and is part of what makes their study so fascinating. Our developmental models still cannot explain variation among limbs of different species. This is critical because crustacean limbs have diversified by varying their branches and lobes, including the biramous branching of endopod and exopod and the medial and lateral lobes that serve to specialize and differentiate limbs functionally. While it is remarkable that crustacean (and arthropod) limbs have a conserved PD patterning module, we still need to understand how the branches and lobes of limbs are patterned. We will be aided in this pursuit by techniques that are becoming easier to use in nonmodel organisms, for example, gene sequencing and RNA interference. In addition, the search for other genes, not used in patterning Drosophila legs, is facilitated by the prevalence of genomic techniques. These advances will allow both functional studies and much broader taxon sampling. Beyond the question of how different limbs are patterned remains the question of how series of limbs are patterned, particularly since the metamorphic lifestyle of so many crustaceans means that a given segment can develop morphologically diverse limbs during ontogeny. None of our models address such metamorphic change, and yet this ability has permitted the specialization of larval lifestyles radically distinct from the adult. Finally, the field of comparative limb patterning has mainly focused on the generic Cartesian axes used to pattern an outgrowth from the body wall. While this has shown some surprising

97

98

Functional Morphology and Diversity conservation, we have not yet begun linking that patterning to the details of the morphology that make up the limbs. Indeed, for crustaceans, we lack even basic models as to how fundamental limb structures, such as muscles and setae, develop. Furthermore, exploring this interface between patterning limb axes and patterning limb morphology may uncover functional links, revealing biologically relevant units of development. If so, these developmentally operational units in the limb could help resolve some of still tangled issues of homology between highly divergent limbs.

ACKNOWLEDGMENTS My thanks to M. Thiel and L. Watling for their comments and editing of this chapter and to L. Nagy and K. Dunlap for critical readings. Ívan Hinojosa generously made the final Fig. 3.8 from my color original.

NOTES 1 Typical nomenclature is to write gene names and their abbreviations in italics and to write the protein product of a gene in all capitals. However, to simplify the presentation, I have avoided abbreviations and distinguishing between RNA and protein expression. 2 The function of extradenticle depends on its being transported to the nucleus. This occurs in the presence of homothorax. Strictly speaking, when I refer to extradenticle, I am referring to this nuclear extradenticle, in contrast to other extradenticle expression that may be present but is nonnuclear.

REFERENCES Abu-Shaar, M., and R.S. Mann. 1998. Generation of multiple antagonistic domains along the proximodistal axis during Drosophila leg development. Development 125:3821–3830. Abzhanov, A., and T.C. Kaufman. 2000. Homologs of Drosophila appendage genes in the patterning of arthropod limbs. Developmental Biology 227:673–689. Angelini, D.R., and T.C. Kaufman. 2004 . Functional analyses in the hemipteran Oncopeltus fasciatus reveal conserved and derived aspects of appendage patterning in insects. Developmental Biology 271:306–321. Angelini, D.R., and T.C. Kaufman. 2005 . Insect appendages and comparative ontogenetics. Developmental Biology 286:57–77. Artavanis-Tsakonas, S., M.D. Rand, and R.J. Lake. 1999. Notch signaling: Cell fate control and signal integration in development. Science 284:770–776. Averof, M., and S. Cohen. 1997. Evolutionary origin of insect wings from ancestral gills. Nature 385:627–630. Averof, M., and N.H. Patel. 1997. Crustacean appendage evolution associated with changes in Hox gene expression. Nature 388:682–686. Awasaki, T., and K. Kimura. 2001. Multiple function of poxn gene in larval PNS development and in adult appendage formation of Drosophila. Development Genes and Evolution 211:20–29 Bateson, W. 1894 . Materials for the study of variation. Macmillan, London . Beermann, A., M. Aranda, and R. Schroder. 2004 . The Sp8 zinc-finger transcription factor is involved in allometric growth of the limbs in the beetle Tribolium castaneum. Development 131:733–742. Beermann, A., D.G. Jay, R.W. Beeman, M. Hulskamp, D. Tautz , and G. Jurgens. 2001. The Short antennae gene of Tribolium is required for limb development and encodes the orthologue of the Drosophila Distal-less protein. Development 128:287–297. Bishop, S.A., T. Klein, A.M. Arias, and J.P. Couso. 1999. Composite signaling from Serrate and Delta establishes leg segments in Drosophila through Notch. Development 126:2993–3003.

Mechanisms of Limb Patterning in Crustaceans Blair, S.S. 1999. Drosophila imaginal disc development: Patterning the adult fly. in V. E.A. Russo, D. Cove, L. Edgar, R. Jaenisch, and F. Salamini, editors. Development, genetics, epigenetics and environmental regulation. Springer, Heidelberg. Boxshall, G. A. 2004 . The evolution of arthropod limbs. Biological Reviews 79:253–300. Brook , W.J., F.J. Diza-Benjumea, and S.M. Cohen. 1996. Organizing spatial pattern in limb development. Annual Review of Cell and Developmental Biology 12:161–180. Browne, W.E., A.L. Price, M. Gerberding , and N.H. Patel. 2005 . Stages of embryonic development in the amphipod crustacean, Parhyale hawaiensis. Genesis 42:124–149. Campbell, G., and A. Tomlinson. 1995 . Initiation of proximodistal axis in insect legs. Development 121:619–628. Campbell, G., and A. Tomlinson. 1998. The roles of the homeobox genes aristaless and Distal-less in patterning the legs and wings of Drosophila. Development 125:4483–4493. Cohen, B., A.A. Simcox , and S.M. Cohen. 1993. Allocation of the imaginal primordia in the Drosophila embryo. Development 117:597–608. Cohen, S.M. 1990. Specification of limb development in the Drosophila embryo by positional cues from the segmentation genes. Nature 343:173–177. Cohen, S. M., and G. Jü rgens. 1989. Proximal-distal pattern formation in Drosophila: Graded requirement for Distal-less gene activity during limb development. Roux’s Archive for Developmental Biology 198:157–169. Damen, W.G., T. Saridaki, and M. Averof. 2002. Diverse adaptations of an ancestral gill: A common evolutionary origin for wings, breathing organs, and spinnerets. Current Biology 12:1711–1716. Dearden, P., and M. Akam. 2000. A role for Fringe in segment morphogenesis but not segment formation in the grasshopper, Schistocerca gregaria. Development Genes and Evolution 210:329–336. de Celis, J.F., D.M. Tyler, J. de Celis, and S.J. Bray. 1998. Notch signaling mediates segmentation of the Drosophila leg. Development 125:4617–4626. Dohle, W., and G. Scholtz. 1988. Clonal analysis of the Crustacean segment—the discordance between genealogical and segmental borders. Development 104:147–160. Deutsch, J.S., and E. Mouchel-Vielh. 2003. Hox genes and the crustacean body plan. BioEssays 25:878–887. Franch-Marro, X., N. Mart í n, M. Averof, and J. Casanova. 2005 . Association of tracheal placodes with leg primordia in Drosophila and implications for the origin of insect tracheal systems. Development 133:785–790. Freeman, J.A., L.B. Cheshire, and T.H. MacRae. 1992. Epithelial morphogenesis in developing Artemia: The role of cell replication, cell shape change, and the cytoskeleton. Developmental Biology 152:279–292. Fristrom, D., and J.W. Fristrom. 1993. The metamorphic development of the adult epidermis. Pages 843–897 in M. Bate and A. Mart ı nez-Arias, editors. The development of Drosophila melanogaster. Cold Spring Harbor Laboratory Press, New York . Gonzalez-Crespo, S., and G. Morata. 1996. Genetic evidence for the subdivision of the arthropod limb into coxopodite and telopodite. Development 122:3921–3928. Guerao, G., K. Anger, U. Nettelmann, and C.D. Schubart. 2004 . Complete larval and early juvenile development of the mangrove crab Perisesarma fasciatum (Crustacea: Brachyura: Sesarmidae) from Singapore, with a larval comparison of Parasesarma and Perisesarma . Journal of Plankton Research 26:1389–1408. Gurney, R. 1942. Larvae of decapod Crustacea. Ray Society, London . Guse. G.W. 1983. Ultrastructure, development, and moulting of the aesthetascs of Neomysis integer and Idotea baltica (Crustacea, Malacostraca). Zoomorphology 103:121–133. Hallberg , E., and B.S. Hansson. 1999. Arthropod sensilla: Morphology and phylogenetic considerations. Microscopy Research and Technique 47:428–439. Hartenstein, V. 2005 . Development of insect sensilla. Pages 379–419 in L. Gilbert, K. Iatrou, and S. Gill, editors. Comprehensive molecular insect science, Vol. 1, Reproduction and development. Elsevier, Oxford . Hejnol, A., and G. Scholtz. 2004 . Clonal analysis of Distal-less and engrailed expression patterns during early morphogenesis of uniramous and biramous crustacean limbs. Development Genes and Evolution 214:473–485.

99

100

Functional Morphology and Diversity Held, L.I. 1995 . Axes, boundaries and coordinates: The ABCs of fly leg development. BioEssays 17:721–732. Held, L.I. 2002. Bristles induce bracts via the EGFR pathway on Drosophila legs. Mechanisms of Development 117:225–234. Inoue, Y., T. Mito, K. Miyawaki, K. Matsushima, Y. Shinmyo, T.A. Heanue, G. Mardon, H. Ohuchi, and S. Noji. 2002. Correlation of expression patterns of homothorax, dachshund, and Distal-less with the proximodistal segmentation of the cricket leg bud. Mechanisms of Development 113:141–148. Jockusch, E.L., C. Nulsen, S.J. Newfeld, and L.M. Nagy. 2000. Leg development in flies versus grasshoppers: Differences in dpp expression do not lead to differences in the expression of downstream components of the leg patterning pathway. Development 127:1617–1626. Jockusch, E.L., T.A. Williams, and L.M. Nagy. 2004 . The evolution of patterning of serially homologous appendages in insects. Development Genes and Evolution 214:324–338. Kojima, T. 2004 . The mechanism of Drosophila leg development along the proximodistal axis. Development Growth and Differentiation 46:115–129. Kreissl, S., A. Uber, and S. Harzsch. 2008. Muscle precursor cells in the developing limbs of two isopods (Crustacea, Peracarida): An immunohistochemical study using a novel monoclonal antibody against myosin heavy chain. Development Genes and Evolution 218:253–265. Lai, E.C., and V. Orgogozo. 2004 . A hidden program in Drosophila peripheral neurogenesis revealed: Fundamental principles underlying sensory organ diversity. Developmental Biology 269:1–17. Lecuit, T., and S.M. Cohen. 1997. Proximal-distal axis formation in the Drosophila leg. Nature 388:139–145. Manzanares, M., T.A. Williams, R. Marco, and R. Garesse. 1996. Segmentation in the crustacean Artemia as revealed by engrailed protein distribution. Roux’s Archive for Developmental Biology 205:424–431. Mardon, G., N.M. Solomon, and G.M. Rubin. 1994 . dachshund encodes a nuclear protein required for normal eye and leg development in Drosophila. Development 120:3473–3486. McKay, D.J., C. Estella, and R.S. Mann. 2009. The origins of the Drosophila leg revealed by the cis-regulatory architecture of the Distalless gene. Development 136:61–71. Mirth, C., and M. Akam. 2002. Joint development in the Drosophila leg: Cell movements and cell populations. Developmental Biology 246:391–406. Mitchell, B., and S.T. Crews. 2002. Expression of the Artemia trachealess gene in the salt gland and epipod. Evolution and Development 4:344–353. Mittmann, B., and G. Scholtz. 2001. Distal-less expression in embryos of Limulus polyphemus Chelicerata, Xiphosura and Lepisma saccharina Insecta, Zygentoma suggests a role in the development of mechanoreceptors, chemoreceptors, and the CNS. Development Genes and Evolution 211:232–243. Mouchel-Vielh, E., C. Rigolot, J. Gibert, and J.S. Deutsch. 1998. Molecules and the body plan: The Hox genes of cirripedes Crustacea. Molecular Phylogenetics and Evolution 9:382–389. Nagy, L.M., and T.A. Williams. 2001. Comparative limb development as a tool for understanding the evolutionary diversification of limbs in arthropods: Challenging the modularity paradigm . Pages in G.P. Wagner, editor. The character concept in evolutionary biology. Academic Press, New York. Nulsen, C., and L.M. Nagy. 1999. The role of wingless in the development of multi-branched crustacean limbs. Development Genes and Evolution 209:340–348. Olesen, J., S. Richter, and G. Scholtz. 2001. The evolutionary transformation of phyllopodous to stenopodous limbs in the Branchiopoda Crustacea—is there a common mechanism for early limb development in arthropods? International Journal of Developmental Biology 45:869 –876. Panganiban, G. 2000. Distal-less function during Drosophila appendage and sense organ development. Developmental Dynamics 218:554–562. Panganiban, G., S.M. Irvine, C. Lowe, H. Roehl, L.S. Corley, B. Sherbon, J.K. Grenier, J.F. Fallon, J. Kimble, M. Walker, G.A. Wray, B.J. Swalla, M.Q. Martindale, and S.B. Carroll. 1997. The origin and evolution of animal appendages. Proceedings of the National Academy of Sciences of the USA 94:5162–5166. Panganiban, G., A. Sebring , L. Nagy, and S. Carroll. 1995 . The development of crustacean limbs and the evolution of arthropods. Science 270:1363–1366.

Mechanisms of Limb Patterning in Crustaceans Patel, N.H., T.B. Kornberg, and C.S. Goodman. 1989. Expression of engrailed during segmentation in grasshopper and crayfish. Development 107:2 01–212. Pavlopoulos, A., Z. Kontarakis, D.M. Liubicich, J.M. Serano, M. Akam, N.H. Patel, and M. Averof. 2009. Probing the evolution of appendage specialization by Hox gene misexpression in an emerging model crustacean. Proceedings of the National Academy of Sciences of the USA 106:13897–13902. Posakony, L.G., L.A. Raftery, and W.M. Gelbart. 1991. Wing formation in Drosophila melanogaster requires decapentaplegic gene function along the anterior–posterior compartment boundary. Mechanisms of Development 33:69 –82. Prpic , N.-M., and M.J. Telford. 2008. Expression of homothorax and in the legs of the crustacean Parhyale evidence for a reversal of gene in the pancrustacean lineage. Development Genes and Evolution 218:333–339. Prpic , N.-M., B. Wigand, W.G.M. Damen, and M. Klinger. 2001. Expression of dachshund in wild-type and Distal-less mutant Tribolium corroborates serial homologies in insect appendages. Development Genes and Evolution 211:467–477. Queinnec , E., E. Mouchel-Vielh, M. Guimonneau, J.-M. Gibert, Y. Turquier, and J.S. Deutsch. 1999. Cloning and expression of the engrailed.a gene of the barnacle Sacculina carcini. Development Genes and Evolution 209:180–205. Rauskolb, C. 2001. The establishment of segmentation in the Drosophila leg. Development 128:4511–4521. Rauskolb, C., and K.D. Irvine. 1999. Notch-mediated segmentation and growth control of the Drosophila leg. Developmental Biology 210:339 –350. Rogers, B.T., M.D. Peterson, and T.C. Kaufman. 2002. The development and evolution of insect mouthparts as revealed by the expression patterns of gnathocephalic genes. Evolution and Development 4:96 –110. Roy, S., and K. VijayRaghavan. 1999. Muscle pattern diversification in Drosophila: The story of imaginal myogenesis. BioEssays 21:486 –498. Scholtz , G. 1990. The formation, differentiation and segmentation of the post-naupliar germ band of the amphipod Gammarus pulex L. (Crustacea, Malacostraca, Peracarida). Proceedings of the Royal Society of London Series B 239:163–211. Scholtz , G., W. Dohle, R.E. Sandeman, and S. Richter. 1993. Expression of engrailed can be lost and regained in cells of one clone in crustacean embryos. International Journal of Developmental Biology 37:299 –304. Scholtz , G., N.H. Patel, and W. Dohle. 1994 . Serially homologous engrailed stripes are generated via different cell lineages in the germ band of amphipod crustaceans (Malacostraca, Peracarida). International Journal of Developmental Biology 38:471–478. Schubiger, G. 1971. Regeneration, duplication, and transdetermination in fragments of the leg disc of Drosophila melanogaster. Developmental Biology 26:277–295. Serrano, N., and P.H. O’Farrell. 1997 Limb morphogenesis: Connections between patterning and growth. Current Biology 7:R186–R195. Sewell, W., T.A. Williams, J. Cooley, M. Terry, R. Ho, and L.M. Nagy. 2008. Evidence for a novel role for dachshund in patterning the proximal leg. Development Genes and Evolution 218:293–305. Shiga, Y., R. Yasumoto, H. Yamagata, and S. Hayashi. 2002. Evolving role of Antennapedia protein in arthropod limb patterning. Development 129:3555–3561. Soler, C., M. Daczewska, J.P. Da Ponte, B. Dastugue, and K. Jagla. 2004 . Coordinated development of muscles and tendons of the Drosophila leg. Development 131:6041–6051. Snodgrass, R.E. 1956. Crustacean metamorphoses. Smithsonian Miscellaneous Collections 131:1–78. Sun, X., F.V. Mariani, and G.R. Martin. 2002. Functions of FGF signaling from the apical ectodermal ridge in limb development. Nature 418:501–508 Taylor, J., and P.N. Adler. 2008. Cell rearrangement and cell division during the tissue level morphogenesis of evaginating Drosophila imaginal discs. Developmental Biology 313:739 –751. Van Valen, L.M. 1993. Serial homology: The crests and cusps of mammalian teeth. Acta Palaeontologica Polonica 38:145–158. Wagner, G.P. 1989. The origin of morphological characters and the biological basis of homology. Evolution 43:1157–1171.

101

102

Functional Morphology and Diversity Williams, T.A. 1998. Distalless expression in crustaceans and the patterning of branched limbs. Development Genes and Evolution 207:427–434. Williams, T.A. 2004 . The evolution and development of crustacean limbs: An analysis of limb homologies. Pages 169–193 in G. Scholtz , editor. Evolutionary developmental biology of Crustacea. Crustacean Issues, Vol. 15. Balkema, Lisse, The Netherlands. Williams, T.A. 2007a . Limb morphogenesis in the branchiopod crustacean, Thamnocephalus platyurus, and the evolution of proximal limb lobes within Anostraca. Journal of Zoological Systematics and Evolutionary Research 45:191–201. Williams, T.A. 2007b. Development of setal cells in the crustacean, Thamnocephalus platyurus. Arthropod Structure and Development 36:63–76. Williams, T.A. 2008. Early Distal-less expression in a developing crustacean limb bud becomes restricted to setal forming cells. Evolution and Development 10:114–120. Williams, T.A., and G.B. Muller. 1996. Limb development in a primitive crustacean, Triops longicaudatus: Subdivision of the early limb bud gives rise to multibranched limbs. Development Genes and Evolution 206:161–168. Williams, T.A., and L.M. Nagy. 1996. Comparative limb development in insects and crustaceans. Seminars in Cell and Developmental Biology 7:615–628. Williams, T.A., and L.M. Nagy. 2001. Developmental modularity and the evolutionary diversification of arthropod limbs. Journal of Experimental Zoology, Molecular and Developmental Evolution 291:241–257. Williams, T.A., C. Nulsen, and L.M. Nagy. 2002. A complex role for Distal-less in crustacean appendage development. Developmental Biology 241:302–312. Wolff, C., and G. Scholtz. 2008. The clonal composition of biramous and uniramous arthropod limbs. Proceedings of the Royal Society of London Series B 275:1023–1028.

4 THE CRUSTACEAN CARAPACE: MORPHOLOGY, FUNCTION, DEVELOPMENT, AND PHYLOGENETIC HISTORY

Jørgen Olesen

Abstract A carapace (a shield extending from the head region and enveloping a smaller or larger part of the body) is a characteristic feature of many crustaceans. This chapter reviews functional, ontogenetic, and evolutionary aspects of the crustacean carapace. Carapace morphology in Crustacea shows much variation, which is reflected in the many functions present in the various subgroups. Among the more widespread functions of carapaces are that in many taxa, they provide hydrodynamic advantages, offer protection, or form a feeding chamber, a respiration chamber, or a brooding chamber. Special attention is devoted to the Branchiopoda and Malacostraca, which both show a large variation in carapace morphology and ontogeny. The influential textbook by Calman (1909) on crustacean morphology and systematics suggested that a carapace was present primitively in both Malacostraca and Crustacea. This assumption was long unchallenged, but a few decades ago attempts were made to invalidate/reject Calman’s carapace hypothesis. Here it is argued that the best starting point may still be to assume homology, at least within Malacostraca. Whether a carapace is homologous between major crustacean taxa is more uncertain due to a general large morphological disparity, but most major taxa have members with a “classical Calman type” of carapace extending from the rear of the head region either as adults or as larvae. The information from the Cambrian “Orsten” crustaceans is ambiguous on this question. Some taxa, such as Rehbachiella and Walossekia, have a classical type of carapace, while others, such as Skara and the even older Yicaris, lack a carapace entirely.

INTRODUCTION A carapace, broadly defined as a shield extending from the head region and enveloping a smaller or larger part of the body, is a characteristic feature of many crustaceans, for example, Functional Morphology and Diversity. Edited by Les Watling and Martin Thiel. © 2013 Oxford University Press. Published 2013 by Oxford University Press.

103

104

Functional Morphology and Diversity

B

D

A

E C

F

1)

2)

3)

Fig. 4.1. Examples of carapace-bearing nonmalacostracan Crustacea. (A) Triops cancriformis (Branchiopoda: Notostraca) (from Gruner 1993). (B) Cyzicus sp. (Branchiopoda: Spinicaudata) mating (from Gravier and Mathias 1930). (C) Bythotrephes longimanus (Branchiopoda: Cladocera) (from Lilljeborg 1901). (D) Chonopeltis inermis (Branchiura) (from Fryer 1956). (E) Euphilomedes aspera (Ostracoda) (from Müller 1894). (A–E all taken from Gruner 1993, with permission from Gustav Fischer Verlag). (F) Metamorphosis of Lepas (Cirripedia): 1, cypris larva; 2, attached larva; 3, young Lepas still surrounded by loosened cyprid carapace (modified from Korschelt and Heider 1890). Arrowheads indicate the carapace.

branchiopods, ostracods, and decapods (Figs. 4.1, 4.2). The term carapace is traditionally used for various types of shields or outgrowths covering the body of many crustaceans, but its use is not restricted to crustaceans. The word is also used for the shields of a variety of noncrustacean arthropod fossils and even for the dorsum of the fused head and thorax in spiders or horseshoe crabs, as well as the well-known shell of a turtle. Crustacea is a large taxon (>60,000 species; see Martin and Davis 2006) with an almost unchallenged variation in morphological appearance spanning from the cave-dwelling, wormlike remipedians to the large claw-bearing lobsters. An equal variation is seen in the ways Crustacea have developed shields (carapaces) to cover the body. Sometimes the shield is fused with the body, as in crabs and lobsters (Fig. 4.2); in other species it is a free shield attached only anteriorly, as in leptostracans (Fig. 4.2A) or clam shrimps (Spinicaudata, Laevicaudata) (Fig. 4.1B); in others it is modified to a dorsal brood pouch, as in raptorial cladocerans (Fig. 4.1C). The large variation in carapace morphology in Crustacea is reflected in the many known (or assumed) functions of carapaces in the various crustacean subgroups. Among the most important functions, not all of which apply to all taxa, are protection, respiration chamber, filtration chamber, brood chamber, and hydrodynamic streamlining.

The Crustacean Carapace

A B

D

C

F

E

G

H

Fig. 4.2. Examples of carapace-bearing malacostracans. (A) Nebalia geoffroyi (Leptostraca) (from Claus 1888). (B) Squilla mantis (Stomatopoda) (from Giesbrecht 1916). (C) Spelaeogriphus lepidops (Spelaeogriphacea) (from Gordon 1960). (D) Euphausia superba (Euphausiacea) (from Gruner 1993). (E) Mysis relicta (Mysida) (Tattersal and Tattersal 1951). (F) Callinectes sapidus (Brachyura) (from Rathbun 1930). (G) Nannastacus unguiculatus (Cumacea) (from Gruner 1993). (H) Penaeus setiferus (Penaeidae) (from Pérez Farfante 1988). (B-H all taken from Gruner 1993, with permission from Gustav Fischer Verlag). Arrowheads indicate the carapace.

Calman (1909), in his classical and influential treatment of crustacean morphology and systematics, suggested that the carapace is a primitive character in Crustacea. In the 1980s and 1990s, the homologies and evolution of the carapace were discussed intensively for various crustacean groups (Dahl 1983, 1991, Newman and Knight 1984, Walossek 1993, Fryer 1996, Watling 1999), but since then this subject has rarely been touched. Considering the variation in morphology, function, and development of carapace structures within Crustacea, it is not surprising that the evolution of this structure has generated much discussion. One of the key questions in the earlier arguments by Dahl (1991) and Newman and Knight (1984) was whether all crustacean carapaces are homologous or, phrased differently, whether a carapace is ancestral in Crustacea. This question is still difficult to answer, but at least the discovery of well-preserved, fossil microcrustaceans from the Cambrian, some of which have classical Calman-type carapaces extending from the rear of the cephalic region (e.g., Mü ller and Walossek 1988, Walossek 1993), provides some hints of early evolution of the carapace. The wide occurrence of carapace structures in representatives of nearly all higher crustacean taxa (see Figs. 4.1, 4.2), and the fact that so many different functions can be attributed to the carapace in different taxa (Table 4.1) suggests that a carapace has been important in

105

Table 4.1. Overview of carapace morphology and function in recent Crustacea.

Branchiopoda (Figs. 4.1A–C, 4.4, 4.5A, 4.7A–E, 4.8–4.14)

Branchiura (Fig. 4.1D)

Thecostraca (Fig. 4.1F)

Ostracoda (Figs. 4.1E, 4.7F)

General morphology in adults Much variation. Notostraca: carapace dorsal shield continuous with head; Spinicaudata, Laevicaudata, and Cyclestherida: carapace bivalved, capable of enclosing entire body (or most); Cladocera: most taxa with bivalve carapace covering body but not head; raptorial cladocerans (Onychopoda, Haplopoda): carapace as dorsal brood pouch without free valves covering body. All branchiurans with flat, shieldlike carapace (e.g., Møller 2009) with head integrated into anterior part and with free posterior lobes overhanging varying parts of body; carapace evenly rounded anteriorly in some taxa (Dolops and some Argulus) and in other taxa divided into lobes (Chonopeltis).

Carapace of head shield type in cyprid or cypridiform stages of all thecostracans (e.g., Høeg et al. 2004); Facetotecta: carapace more or less maintains naupliar shape (not bivalved) overhanging body (Høeg et al. 2004); Ascothoracida and Cirripedia: carapace bivalved with dorsal hinge line (Høeg et al. 2009); adult cirripedes (barnacles, etc.): no carapace, but carapace valves of cyprid larvae are ontogenetic precursors to fleshy mantle with calcified shell plates (e.g., Calman 1909, Glenner and Høeg 1993). Carapace most often calciferous and bivalved with valves held together dorsally by ligament, often also by hinge (Maddocks 1992); a few recent taxa (Swanson 1989) have a univalved carapace.

Important functions in adults General protection of body and limbs (most taxa); streamlining during swimming (some cladocerans and Notostraca); facilitates burrowing (Notostraca); part of respiration/feeding current system (“Conchostraca” and Cladocera); protection of eggs/embryos (Notostraca) or as dorsal brood chamber (“Conchostraca” and Cladocera); housing long, curled tubules of maxillary glands. Smooth, flattened carapace together with generally low profile of all branchiurans reduces frictional drag when attached to hosts; general protection of body and limbs; houses widely branched digestive system; carapace lobes ventral surface with so-called respiratory areas involved in osmoregulation (Haase 1975, Boxshall and Jaume 2009); houses spermatophore glands in Dolops (Fryer 1960). Hydrodynamics (e.g., Crisp 1955, Walker and Lester 2000, Walker 2004); general protection of body and limbs; surface with sensory organs such as “lattice organs” (Jensen et al. 1994, Rybakov et al. 2003).

General protection of limbs and body; respiration (e.g., Abe and Vannier 1995); spaces between internal lamellae of carapace valves are continuations of body cavity and may house organs related to reproduction, digestion, nervous system (Maddocks 1992).

Leptostraca (Figs. 4.2A, 4.16A,B)

Large, bivalved, but with no dorsal hinge; dorsum of carapace continues anteriorly into small articulated rostral plate; in most taxa carapace covers phyllopodous thoracopods laterally (e.g., Nebalia), but in some taxa (e.g., Paranebalia and Nebaliopsis) thoracopods extend between valves ventrally.

Euphausiacea (Fig. 4.2D)

Carapace fused with thorax dorsally so that cephalon and thorax form one piece (cephalothorax); a short spine extends in the midline anteriorly; laterally carapace folds relatively short not covering gills of thoracopods. Carapace fused with thorax dorsally so that cephalon and thorax form one piece (cephalothorax); lateral folds of carapace form gill chambers; much taxon-specific variation in general carapace morphology.

Decapoda (Figs. 4.2F,H, 4.3, 4.5B–D, 4.15C, 4.16E,F)

Stomatopoda (Figs. 4.2B, 4.15B, 4.16C,D)

Mysidacea (Figs. 4.2E, 4.6A,B)

Carapace of cephalothorax type incorporating head and one or two thorax segments (number uncertain since this body region is compressed); cephalothorax has free lateral margins; anteriorly is median rostral plate articulated to cephalothorax; some taxa with distinct longitudinal keel in the midline of the cephalothorax and with additional longitudinal lateral ridges. Carapace typically large and with free folds posteriorly and ventrally covering most of thorax and the basal parts of the thoracopods; carapace in Mysida dorsally fused with first three thoracic segments (in Stygiomysis with first four) (Gruner 1993).

General protection of body and limbs; form sides of filtration chamber (see Cannon 1927); respiration chamber, but gas exchange over inner surface uncertain (see chapter 14 in this volume); hydrodynamics; rostral plate acts as ram during burrowing (Vannier et al. 1997); embryos carried ventrally by thoracopods between carapace valves. Hydrodynamics; inner side possesses very thin cuticle so probably respiratory (Zimmer and Gruner 1956) (see chapter 14); carrying compound organs dorsally (function unknown) (Mauchline and Nemoto 1977). General protection due to heavy calcification; hydrodynamics; provides robust attachment sites for limb musculature; more room for internal organs; carapace fused with thorax (cephalothorax) may facilitate backward swimming as performed during “escape reaction”; sides of carapace (branchiostegites) form gill chambers closed to varying degree in various taxa; branchiostegites involved in respiration known only for few species (see chapter 14); sound production (Henninger and Watson 2005). Sound production by carapace vibration (Patek and Caldwell 2006); general protection of mouth parts and bases of maxillipeds; female carapace grabbed by male during copulation (e.g., Dingle and Caldwell 1972); carapace length shorter than burrow width (probably to enable possibility of turning around) (Atkinson et al. 1997). Mysida: carapace forms a respiratory chamber with gas exchange taking place on the inner surface (Mayrat et al. 2006, Wirkner and Richter 2007); respiratory current is produced by lamellar epipod of first thoracopod (Cannon and Manton 1927); Lophogastrida: carapace previously considered nonrespiratory (but see Wirkner and Richter 2007). (Continued)

Table 4.1. (Continued)

Cumacea (Figs. 4.2G, 4.6D–F)

Tanaidacea (Fig. 4.6G,H)

Thermosbaenacea (Figs. 4.6C, 4.7G)

Speleogriphacea (Fig. 4.2C) Mictacea

General morphology in adults Carapace of cephalothorax type incorporating dorsally anterior three or four thoracic segments (sometimes also segments 5 and 6). The lateral lobes of carapace form chambers closed posteriorly (comparable to gill chambers in brachyurans) (Zimmer 1941). Carapace of cephalothorax type incorporating dorsally two (rarely three) thoracic segments; lateral carapace folds cover mouth appendages laterally, forming semiclosed lateral respiratory chambers; posteriorly with a pair of respiratory pores at edge of carapace. In Thermosbaena mirabilis short carapace extends from head to posterior border of fourth thoracic segment; dorsally fused with first thoracic segment (Barker 1962); lateral folds of carapace extend anteriorly and cover lateral bases of mouth parts and first and second antennae. In Spelaeogriphus lepidops short carapace, which extends from head posteriorly, consists of pair of lateral lobes overhanging anterior somites. Carapace not developed posteriorly but small lateral carapace folds cover bases of maxillae mx1, mx2, and mxp (head fused to first thoracic somite) (Bowman et al. 1985).

Important functions in adults Respiration/feeding chambers; water current enters ventrally and leaves anteriorly in Diastylis (but see text) generated by movable and ventilatory epipod of maxilliped (Calman 1909, Oelze 1931, Zimmer 1932; Dennell 1937). Respiration chambers with inner side being involved in gas exchange (e.g., Tanais cavolinii) (Lauterbach 1970, Johnson and Attramadal 1982).

Respiratory chamber; inner side of carapace with epithelium containing system of vascular lacunae connected to perivisceral and pericardial sinuses (Siewing 1958); epipodite of the maxilliped beat within branchial chambers to draw inhalant respiratory current (Barker 1962); ovigerous females carry embryos under enlarged carapace. At inside is “oval patch” that may be respiratory (Gordon 1957, Grindley and Hessler 1971). Lateral carapace lobes thought to be respiratory but maxillipedal epipod associated with respiratory function, as in other taxa, lacking carapace (Bowman et al. 1985).

The Crustacean Carapace crustacean evolution. The evolutionary success of crustaceans (or arthropods in general) is probably very much linked to the plasticity of segmented limbs and the possibilities this has given (e.g., Walossek 1993, Boxshall 2004; see also chapters 2 and 7 in this volume), but the carapace, with all the functional possibilities such a structure gives, seems to have been equally important. This chapter highlights the evolutionary significance of carapace structures within Crustacea, both from a functional perspective and from a comparative morphological perspective. The following aspects are reviewed: (1) morphology and function of the carapace in selected crustaceans; (2) case studies on the ontogeny of the carapace in two crustacean key taxa, Branchiopoda and Malacostraca, both of which exhibit much variation in carapace morphology and function of both larvae and adults, and both of which played a central role when Calman formulated his crustacean carapace hypothesis; (3) critical summary of recent discussions of Calman’s (1909) inf luential carapace hypothesis; and (4) Cambrian (“Orsten”) evidence of carapace structures.

MORPHOLOGY AND FUNCTION OF THE CRUSTACEAN CARAPACE The following is a brief overview of some of the main functions of the crustacean carapace, but there are other functions not mentioned below (but see Table 4.1). The literature is vast, so this summary is by no means exhaustive. Hydrodynamics Carapace morphology and hydrodynamics are linked in many ways in Crustacea. Most obviously, a carapace often provides general streamlining of the body surface and reduces drag since it covers limbs and other projecting body parts. The carapace in Cambrian crustaceans such as Rehbachiella (see Fig. 4.8, below) and Bredocaris probably had such a streamlining function, since in these taxa the carapace is a simple posterior extension of the naupliar shield. Indeed, hydrodynamic advantages may have been driving early carapace evolution (but were probably also coupled with feeding advantages). Multiple other functions have evolved (feeding, respiration, brooding), but hydrodynamic aspects remain important in many taxa. Streamlining (reduction of drag) is important in two ways: (1) it allows for association (attached or moving) with the substratum without being pulled off by water currents, and (2) it increases swimming speed or makes it less costly. Here I present a few examples of crustacean carapaces where hydrodynamic properties clearly play a significant role. In the parasitic Branchiura (carp lice), streamlining as an adaptation to ectoparasitism on fish has been taken to an extreme (Fig. 4.1D). The flattened carapace and the generally low body profile effectively reduce drag when attached to the fish host. A general flattening of the body is seen in many other crustacean taxa as a modification to their parasitic lifestyle (e.g., Copepoda, Isopoda, Amphipoda) but without involving a carapace. Another example of a carapace with clear hydrodynamic advantages is that of the cyprid larvae in cirripede crustaceans. The cyprid is a larval type found only in cirripedes. It is a nonfeeding larva with a suite of specializations for the purpose of locating a suitable substratum before irreversibly settling and molting to a juvenile cirripede (barnacles, gooseneck barnacles, or parasitic barnacles) (e.g., Høeg et al. 2004, Høeg and Møller 2006) (Fig. 4.1F). Among the specializations of the cyprid larva is a bivalved carapace capable of enclosing the body almost entirely between the valves. The hydrodynamic advantages of this carapace shape for swimming cyprid larvae were explored by Walker (2004), who found that cyprids of Heterosaccus lunatus were significantly faster than the naupliar stages earlier in development, which reflects the efficiency of the fusiform shape of the cyprid.

109

110

Functional Morphology and Diversity In general, cyprid larvae swim faster than other similarly sized invertebrate larvae (Walker et al. 1987). During the phase where the cyprid explores the substratum and where eventually attachment takes place, a streamlined carapace is clearly important since it reduces drag from strong currents and tides and allows the cyprids to rapidly scan multiple sites. Also in branchiopods, such as notostracans, where the relatively flattened carapace forms one piece with the head, streamlining during swimming is important (Fig. 4.1A), but this morphology may also be related to its occasional burrowing behavior (Fryer 1988). A well-known locomotory adaptation in malacostracan crustaceans is the “escape reaction,” which involves repeated tail-flips performed by the often heavily muscularized pleon. In some taxa a tail-flip results in a 180° turn along the body axis (e.g., stomatopods or thermosbaenaceans; see Olesen et al. 2006), while in heavily calcified decapods such as crayfish, a series of repeated flips results in backward swimming. Jacklyn and Ritz (1986) compared the swimming of scyllarid lobsters with that of a panulirid lobster with respect to how much the shape of the carapace contributes to maneuverability during backward swimming. They found that the flattened carapace in scyllarid lobsters, in contrast to panulirid lobsters, is formed in such a way that it produces a hydrodynamic lift when the large flattened scales of the antennae are lowered (Fig. 4.3). They also saw that the independent movement of the right and left antennal scales, which are at the rear during backward swimming, may alter the water flow and change the distribution of the lift and thereby control rolling or “pitching” during swimming. In the panulirid lobster also examined by Jacklyn and Ritz (1986), the tail-flips were used only to travel in straight lines and over a short distance during the escape response, which they explained as the result of the production of a negligible amount of lift during each tail-flip and by the lack of antennae shaped or positioned to control lift. Another type of swimming among decapods is that found in portunid crabs, which use their fifth pair of pereopods as swimming paddles (Fig. 4.2F). Portunids can swim in all directions, slowly when going backward or forward but rapidly when moving sideways (Lockhead 1961). Blake (1985) examined the hydrodynamic properties of the carapace of Callinectes sapidus. He found that f luid resistance (drag) was least when the side of the carapace was oriented at right angles to the f low, which was interpreted as an adaptation to sideways high-speed swimming. Blake (1985) also found that the carapace of Callinectes in general is adapted for minimum resistance and to generate hydrodynamic lift at low speeds during forward swimming. Feeding In a number of well-known examples among crustaceans, the carapace plays a direct role in the feeding process. In branchiopods (except some raptorial cladocerans), the beating of the trunk limbs sets up a powerful current of water that is drawn in medially between the limbs and then expelled laterally, a current that is an important part of the feeding process and sometimes also in locomotion (e.g., Anostraca and Notostraca). In diplostracan branchiopods, where a large and often bivalved carapace is present, this current is mostly used for feeding only. In laevicaudatan branchiopods, which have a large and globular carapace (see Figs. 4.7A, 4.12C, below), the carapace plays a role in setting up the feeding current. The exopods of the trunk limbs fit neatly against the inner wall of the carapace, thereby sealing the interlimb spaces laterally and creating a vacuum in these spaces, with the result that water is drawn medially into the food groove when the limbs are beating metachronically (Fryer and Boxshall 2009). Laevicaudatan branchiopods are not true filtrators, but the basic principles of how the feeding current is set up are the same as in most other branchiopods. In cladocerans such as Daphnia, a feeding current is generated in a largely similar way (see Fryer 1991), and also here the carapace plays a role in guiding the currents

The Crustacean Carapace A

1)

2)

3)

B

LL+

d

Fig. 4.3. Hydrodynamics of the scyllarid Ibacus peronii. (A) Three different types of backward swimming (1–3). The propulsion is generated with repeated tail-flips. The carapace and the antennal scales together form an effective aerofoil profile resulting in lift. The swimming direction is controlled by changing the angles of the antennal scales. (B) By changing the articulation angle of the flattened antennal scales, the distribution of the lift can be altered so that pitching and rolling movements are created. L– and L+ indicate a downward- and upward-directed lift, respectively, when one antennae (right) is raised and the other (left) is lowered; d indicates the perpendicular distance from rolling axis to the point of application of the resultant lift forces. From Jacklyn and Ritz (1986), with permission from Elsevier.

(Fig. 4.4A,B). A number of cladoceran branchiopods have exploited niches where carapace specializations and function go hand in hand. For example, the daphniid genera Scapholeberis and Megafenestra have a modified ventral carapace in such a way that they are capable of suspending themselves in an inverted position from the surface film of water (e.g., Fryer 1991) (Fig. 4.4C,D). In general, modifications of the ventral carapace margins have been important within Cladocera

111

112

Functional Morphology and Diversity C A B D water entering carapace chamber G

Exhalant current

E

F

H

Fig. 4.4. The carapace as a feeding chamber and modifications of ventral carapace margins in cladoceran Branchiopoda. (A and B) Daphnia galeata (Daphniidae), with carapace drawn transparent to show feeding appendages (A) and schematic drawing showing where feeding/respiration currents enter and leave the carapace chamber (B). (C and D) Scapholeberis mucronata (Daphniidae) using modifications of ventral carapace margins to suspend itself in an inverted position from the surface film of water (C) and showing ventral carapace modifications (D). (E and F) Peracantha truncata (Chydoridae) crawling over the surface when balancing on the ventral margins. (G and H) Graptoleberis testudinaria (Chydoridae) seen from below (G) and behind (H) as it glides over a surface. Note the row of long setae on the ventral carapace margins, which seals the carapace chamber when gliding. A–D from Fryer (1991), and E–H from Fryer (1968), with permission from the Royal Society of London.

(e.g., Fryer 1968, 1991, Smirnov and Kotov 2009), but especially within Chydoridae this has been an evolutionary trend. Fryer (1968) described a number of cases where the morphology of the ventral carapace margins are important for crawling while food is collected in various ways from the substratum by the trunk limbs (e.g., Fig. 4.4E–H). In the least specialized forms, such as Alonopsis elongata, the setae on the ventral carapace margin are in contact with the substratum on which it balances when it crawls forward by means of the first trunk limbs. Specializations for crawling have been further exploited by Alonella exigua, which has more ventral modifications (e.g., more ventral setae, wider ventral flange). It is capable of largely sealing the carapace chamber and then pumping water from it, which maintains a pressure difference between the water inside and outside, enabling it to cling to and crawl over surfaces like a fly on a ceiling (Fryer 1968). Graptoleberis testudinaria has established an almost entirely water-tight seal with a very wide ventral flange and specialized setae (Fryer 1968) (Fig. 4.4G,H), enabling it to slide over surfaces like a gastropod mollusc by setting up a pressure difference between the water inside and outside the carapace chamber. Species from the genus Nebalia (Leptostraca) spend much of their life burrowed in mud. They have a large, bivalved carapace that constitutes the lateral sides of a large filtration chamber. Cannon (1927) reported that the phyllopodous trunk limbs produce a water current, which

The Crustacean Carapace enters the filtration chamber anteriorly and leaves it posteroventrally, from which particles are being filtered out by the setae on the limbs. According to Cannon (1927), the anterior freemoving part of the carapace, the rostrum, can be depressed and thereby appears to control the current entering the filtration chamber. Vannier et al. (1997) suggested that the rostrum acts as a ram during burrowing, preventing large particles from entering the filtration chamber from anterior (see Fig. 10.1D in chapter 10 in this volume). Respiration In a number of examples within the Crustacea, the carapace forms a respiration chamber in which gas exchange takes place, commonly over gills situated within the chamber (decapods), but sometimes directly over the inner side of the carapace (branchiopods, ostracods, peracarid malacostracans). There are even cases where the carapace chamber functions partially as a lung, as it contains air, which is used for respiration while the animal is on land (e.g., grapsid or ocypodid crabs). For branchiopods, it was previously assumed that lateral, saclike limb structures, the epipods (or “gills”), were the main organs for respiration (e.g., Gicklhorn 1925), but later studies of the epithelia (e.g., Kikuchi 1983) have shown that osmoregulation seems to be the main function of the epipods (see Maas et al. 2009). It remains open to discussion whether branchiopod epipods also have a respiratory function (see chapter 14 in this volume). Instead, the inner side of the carapace is important for gas exchange. For Daphnia magna, recent studies show that the feeding current is important for uptake of oxygen from the ambient medium and that gas exchange occurs mainly within the filtering chamber (Pirow et al. 1999a) (Fig. 4.5A). A follow-up study by the same authors showed that the inner wall of the carapace is the major site of respiratory gas exchange (Pirow et al. 1999b). In contrast to earlier studies, where assumptions about respiratory functions were mainly based on studies of tissue composition, Pirow et al. (1999b) developed an elegant method to directly image hemoglobin oxygen saturation in transparent animals, which allowed them to localize the specific areas on the body surface where oxygen uptake and release take place. In Daphnia magna the highest values of hemoglobin oxygen saturation occurred near the posterior margin of the carapace and in the rostral part of the head (which is outside the carapace), suggesting that these areas are the most important for oxygen uptake (Pirow et al. 1999b). In small animals living in the range of low Reynolds numbers, where viscous forces are dominant (Koehl and Strickler 1981), the boundary layer is an obstacle for diffusive gas transport. The thickness of the boundary layer depends on the velocity of the surrounding water relative to the surface of the organism, and Pirow et al. (1999b) pointed out that reduced boundary layers should occur inside the filtering chamber, where water is being pumped by the beating phyllopodous limbs. They found that the enlarged carapace valves not only form the lateral boundaries of the filtering chamber but are also employed for oxygen uptake due to the thin nature of the inner wall of the carapace. In Daphnia magna, oxygen uptake does not only occur through the inner wall of the carapace—the rostral part of the head had high oxygen values, which may provide additional oxygen for sensory and central nervous system structures located in the head. This additional respiratory surface may be of advantage during hypoxia if the carapace-based respiratory system fails to supply enough oxygen to the head regions (Pirow et al. 1999b). In ostracods, as with many other smaller crustaceans, it is generally assumed that gas exchange takes place over the general surface of the body. But notable exceptions are found in larger ostracods, for example, in Vargula hilgendorfii, where oxygen uptake is assumed to occur preferentially through the inner (posterior) surface of the carapace where the hemolymph sinuses are best developed and are in direct contact with seawater (Abe and Vannier 1995). In Branchiura, the so-called respiratory areas on the ventral sides of the carapace lobes have traditionally been

113

114

Functional Morphology and Diversity B

A

Midgut gland

Heart

Pericarical sinus Carapace

Carapace Gills

Gill chamber

Artery to leg Ventral nerve cord

D

C

–∼ 60˚ AIR AIR 5

4

3

2

1

Fig. 4.5. The carapace as a feeding and respiration chamber in various Crustacea. (A) Respiratory function of carapace in Daphnia magna (Cladocera). Arrows left of specimen show feeding/respiratory current. Arrows inside specimen show main routes of blood flow. Fluorescence microscopy has revealed that the main sites of gas exchange are at the inner wall of the carapace near the posterior margin and at the rostrum (from Pirow et al. 1999b, with permission from the Company of Biologists, Ltd.). (B) Cross section of generalized decapod showing position of gill chambers, which are formed by lateral flanges of the carapace (gill covers). Drawing combined from various sources. (C) Procambarus clarkii (Astacidea) with the gill covers removed showing outer layer of gills. 1–5 represent pereopods 1–5 (from Bauer 1998, with permission from Wiley and Sons). (D) Branchial chambers modified for air breathing and water circulation in the semaphore crab Heloecius cordiformis (Ocypodidae). The chambers are divided in an upper part containing air and a lower part containing water (where the gills are). Forward tilting allows water to be tipped forward and pumped out of the branchial chambers anteriorly and air simultaneously to be sucked in posteriorly (from Maitland 1990b, with permission from Springer).

assumed to be involved in gas exchange, but physiological/anatomical experiments by Haase (1975) suggested that these areas have an osmoregulatory function. Among malacostracans, there are multiple examples of the carapace forming a respiratory chamber, often with water being actively pulled into it by beating limbs specialized for this purpose. The respiratory system of decapod crustaceans is well known—the lateral lobes of the carapace (branchiostegites) form a chamber on each side of the body, housing gills of varying morphology attached to the thoracopods basally or to the body laterally (see Hong 1988 for details and overview) (Fig. 4.5B,C). A specialized part of maxilla 2, the scaphognathite or gill bailer, beats and performs rapid movements independent from the movements of the remaining part of the limb, thereby dragging a ventilatory water current through the carapace chamber. The gills are normally irrigated with a posterior-anterior flow, and the water is expelled from the gill chambers through openings near the mouth. The precise site where water enters the chambers varies among taxa, but in brachyurans it is generally through discrete openings near

The Crustacean Carapace the bases of the pereopods, while in other taxa it may enter at varying positions along the ventral carapace margin (e.g., Dyer and Uglow 1978, Batang and Suzuki 1999). Reversal of flow direction within the gill chambers has been recognized as a gill-cleaning strategy in certain decapods (Bauer 1989, 1998) and may be particularly important for removing sediment from the gills in burrowing species. In one example, Metapenaeus macleayi, the respiratory current enters a tube in the sediment formed by the antennal scales and the antennules before it flows into the gill chambers (Ruello 1973). It is well known that many terrestrial or semiterrestrial brachyuran decapods retain water in the branchial chambers of their carapace while active on land and, at the same time, have a part of the chamber modified as a lung, relying on oxygen uptake directly from the air (Farrelly and Greenaway 1994). Aspects of water retaining behavior are particularly well studied in the semaphore crab, Heloecius cordiformis (Ocypodidae). A series of papers by Maitland (1990a, 1990b, 1992a, 1992b) showed that the water-retaining capacity of the carapace is intimately linked to both respiration and feeding (Fig. 4.5D). One of the key findings is that H. cordiformis is an obligate air breather while active on land. Air-based gas exchange takes place above the gills in air-filled cavities lined with a vascular epithelium that function as lungs. While active on land, the gill chambers are still partly filled with water thus irrigating the gills, but in experiments where the branchial water has been removed, activity levels and oxygen percentages were unaffected, suggesting that gill respiration is of less importance (for more details, see Maitland 1990a, 1990b, 1992a, 1992b). While on land, H. cordiformis sequentially depresses and elevates its carapace in a pumplike manner (as in many other terrestrial brachyurans). The specific function of this behavior is complicated, but Maitland (1992a) suggests that the carapace pump may enable Heloecius to functionally partition lung ventilation from water circulation, thereby alleviating the potential conflict between air breathing and water circulation. Another important function of the branchial water in H. cordiformis is for feeding. Water from the branchial chambers is pumped out onto the mouthparts and is used in the separation of edible material from mud and sand (Maitland 1990b). The carapace also serves as a respiratory chamber in some peracarids. One example is the Mysida, where no epipods acting as gills are present, but respiration takes place over the inner wall of a relatively long carapace (e.g., Mayrat et al. 2006, Wirkner and Richter 2007). Cannon and Manton (1927) studied the feeding and swimming of Hemimysis lamornae and found that alongside with feeding and swimming currents, a special respiratory current is drawn in under the posterior edge of the carapace and pushed out anteriorly at the sides of the maxillae by the beating, lamellar epipods of the first thoracic limbs (see also Laverack et al. 1977) (Fig. 4.6A). Lophogastrids, which are sometimes grouped with Mysida in Mysidacea (e.g., Richter and Scholtz 2001), are in many ways similar to mysids, but the carapace has been considered nonrespiratory (e.g., by Dahl 1991). This notion has recently been challenged by Wirkner and Richter (2007). As in mysids, a respiration current inside the carapace is generated by a beating epipodite of the first pair of thoracopods and apparently also by the exopods of the second maxillae (Childress 1971), but in contrast to mysids, gills (epipods) are present on thoracopods 2–7 (see Boxshall and Jaume 2009), which probably takes care of most of the gas exchange (Fig. 4.6B). In the Thermosbaenacea, Tanaidacea, Cumacea, and Speleogriphacea, a short carapace forms a small respiratory chamber (Grindley and Hessler 1971), inside of which a branchial/ ventilatory first thoracopod (maxilliped) epipod generates a water current. The thermosbaenacean Thermosbaena mirabilis was studied in detail by Barker (1962), and notes were also made on the respiratory system based on the study of living specimens. As in mysids, a respiratory current is drawn through the posterior opening of the short carapace (Fig. 4.6C) by the beating of the maxilliped epipod. Barker (1962) suggested that the epipod functions as a gill since it

115

116

Functional Morphology and Diversity Inhalant respiratory current

A

Carapace

MYSIDA

B

LOPHOGASTRIDA Carapace

SCAPHOGNATHITE # 2

LATERAL GILL

LEG BASE

THERMOSBAENACEA

Exhalant respiratory current

SCAPHOGNATHITE # 1 VENTRAL GILL

Carapace Inhalant D respiratory current

C

Carapace

CUMACEA

Carapace

E

th 2 th 3 th 4 th 5 mxp ep II

III

IV

Antennule Mandible

Exhalant current

endopod Inhalant current

F

epipod

Maxilliped

G

TANAIDACEA

Inhalant respiratory current

Respiratory pores

H

Fig. 4.6. Respiratory carapaces and ventilation currents in peracarid malacostracans. (A) Lateral view of thorax of Hemimysis lamornae (Mysida) showing the course of the respiratory currents (arrows). The current is drawn in under the hind edge of the carapace, the inner wall of which is respiratory (no gills) (from Cannon and Manton 1927, with permission from the Royal Society of Edinburgh). (B) Schematic lateral view of Gnathophausia ingens (Lophogastrida) showing the course of the respiratory current under the carapace, where gills take care of gas exchange. The respiratory current is generated by two different “scaphognathites”: one is the exopod of the second maxilla; the other is the epipod of the first trunk limb (from Childress 1971, with permission from the Biological Bulletin/Marine Biological Laboratory, Woods Hole, MA). (C) Respiratory carapace in Thermosbaena mirabilis (Thermosbaenacea), carapace partly removed (outline indicated by dashed line). Arrows show direction of respiration current (from Barker 1962, with permission from the Company of Biologists, Ltd.). (D) Diastylis species (Cumacea) buried in sand, showing the inhalant and exhalant feeding and respiratory/feeding currents (arrows) in buried animals (from Dennell 1937, based on Zimmer, with permission from the Royal Society of Edinburgh). (E) Anterior part of body of Diastylis from lateral showing carapace with large right respiratory maxillipedal epipods inside. Water current enters ventrally (from Oelze 1931). (F) Maxilliped of Diastylis with respiratory epipod (from Calman 1909). (G and H) Tanais cavolinii (Tanaidacea) (from Johnson and Attramadal 1982, with permission from Taylor and Francis, Ltd.). (G) Schematic drawing showing water flow around the body (arrows). Respiratory water enters through dorsal respiratory pores into respiratory carapace. (H) Dorsal view of carapace showing position of respiratory pores (arrows).

contains vascular lacunae and offers a relatively large surface. However, the main site of respiratory exchange is the inner side of the carapace wall, where the cuticle covers a system of vascular lacunae in the carapace epithelium (Siewing 1958, Barker 1962). In the Cumacea, the epipods of the first maxilliped generate a respiratory current as in the other mentioned taxa, but by contrast, this current also serves as a feeding/filtration current. Zimmer (1932) found, in nonburied specimens of Diastylis rathkei , that the inhalant current

The Crustacean Carapace enters the respiratory chambers between the bases of the third maxillipeds, which seem in accordance with Oelze (1931) (Fig. 4.6E). In buried specimens, which had been undisturbed for a period of time, Zimmer (1932) found that the inhalant current forms a small funnel in the sediment between the bases of the third maxillipeds and the body. The exhalant current leaves anterior to the site of intake, guided by the pseudorostrum (Zimmer 1932). Dennell (1937), based on Zimmer (1932), provided an illustration of a buried specimen of Diastylis species with arrows close to the anterior margin of the carapace indicating the respiratory currents (Fig. 4.6D). This illustration has since been reproduced in many text books. However, the anterior position of the inhalant current indicated by Dennell is rather different from that shown in a figure by Oelze (1931), where the current was found to enter more ventrally through small slits (compare Fig. 4.6D,E). Possibly the two different representations can be reconciled. The ventral current entrance shown by Oelze (1931) (Fig. 4.6E) may be the anatomically true one, while the entrance shown in the overview figure by Dennell (1937) shows where the inhalant current approximately enters when the animal is buried and the ventral side is covered by sediment. The ventilatory epipod of the first maxilliped in Diastylis is rather large and has a region that is folded into branchial lamellae (Calman 1909, Oelze 1931) (Fig. 4.6E,F), which must function in gas exchange. The inner side of the carapace must be involved in gas exchange, as in a number of other peracarids, since both valves are filled with hemolymph channels (Oelze 1931, Siewing 1952). In Tanaidacea the carapace forms a respiratory chamber on each side of the body as well. Calman (1909) reported that the respiratory system of Apseudidae resembled that of cumaceans, for example, with respect to the presence of a ventilatory maxillipedal epipod. Calman (1909) also mentioned that the lateral folds of the carapace in tanaidaceans are traversed by a network of blood channels and suggested that these form the chief organs of respiration, possibly assisted by the epipodites of the maxillipeds, all of which was later confirmed by Lauterbach (1970) for Tanais cavolinii. Johnson and Attramadal (1982) made detailed observations on the mechanics of the carapace-based respiratory system in T. cavolinii. The tube-building lifestyle of T. cavolinii prevents simple observation of its behavior, so to overcome this problem, the animals were offered capillary glass tubes of varying diameters, which were readily accepted (Fig. 4.6G). Observations showed that the basics of the respiratory system of T. cavolinii is comparable to that of thermosbaenaceans and cumaceans (see above) but that certain characteristics can be explained as adaptations to tube-dwelling behavior. According to Johnson and Attramadal (1982), the ventilatory epipods of the maxillipeds, under normal conditions, generate a respiratory current in each branchial chamber, which enters the chamber through two respiratory pores at the posterior edge of the carapace valves (see Fig. 4.6H) and leaves through ventral pores close to the base of the maxillipeds. Occasionally, the respiration current is reversed (linked with reversal of pleopod pumping) so that the water enters the branchial chamber ventrally and is expelled through the dorsal pores. Johnson and Attramadal (1982) assume that this is to protect the dorsal respiratory pores and the branchial chamber from being clogged by foreign particles in the suspension. Spelaeogriphus lepidops, the first discovered species of the Spelaeogriphacea (Gordon 1957) (Fig. 4.2C), has a respiratory carapace, but the respiratory chamber is less closed than that of tanaidaceans and cumaceans and consists basically of lateral carapace lobes overhanging the anterior somites. Based on Gordon (1957, 1960) and Grindley and Hessler (1971), who contrary to Gordon had the opportunity to study live material, it can be concluded that S. lepidops apparently has two different respiratory systems, which is unusual for Crustacea: one based on the thoracopodal exopods (which will not be treated here) and one involving the carapace. Both Gordon (1957) and Grindley and Hessler (1971) assumed that the cup-shaped maxillipedal epipod probably is respiratory and that the “oval patch” (Gordon 1957) on the inside of the carapace

117

118

Functional Morphology and Diversity might well be a respiratory surface, but unfortunately it was not possible for Grindley and Hessler (1971) to observe the specific functioning of the epipodal gill of the maxilliped in the living specimens studied by them. Within Mictacea, which is another small peracarid order with only few species and which show much resemblance to thermosbaenaceans and spelaeogriphaceans, it has been reported that Mictocaris halope has an inflated, thin-walled elliptical area dorsal to the carapace fold apparently functioning in respiratory exchange, but no respiratory maxillipedal epipod is present (Bowman et al. 1985). Brooding Chamber A variety of crustaceans have brood care, and in some taxa the offspring are protected by the carapace during part of their development. Branchiopod crustaceans (clam shrimps and water fleas) are probably best known for this habit, but also in certain ostracods, ascothoracidan thecostracans, and thermosbaenacean malacostracans, the carapace offers protection during development. In spinicaudatan and laevicaudatan (Fig. 4.7A) branchiopods (clam shrimps), the eggs (embryos) are retained in a pair of clusters under the carapace valves, kept in position by various supporting structures of the female’s appendages. In both these taxa the eggs are kept under the carapace but only during a short part of their development, and they are released as free-swimming nauplii or nauplia-like larvae (Olesen and Grygier 2003, 2004, Olesen 2005). Cyclestheria hislopi, a clam shrimp with many similarities to Spinicaudata but now recognized as a sister group to the Cladocera (water f leas) (e.g., Martin and Davis 2001, Richter et al. 2007), has taken it a step further and retains the offspring under the carapace attached to dorsal filaments of the trunk limb exopods until they are released as small adultlike juveniles (“direct development”; Olesen 1999) (Fig. 4.7B; see also Fig. 4.11C, below). In the taxonomically diverse Cladocera, development is also direct and takes place dorsally under the carapace (a single exception are the winter eggs of Leptodora kindtii , from which free-swimming metanauplii hatch), but different from C. hislopi, the embryos are not attached to the female via filaments. In most cladocerans, the inside of the brood chamber is directly connected with the surrounding media, but in a few taxa (Moinidae, Onychopoda, Penilia), the developing embryos are nourished by a placenta-like organ, a so-called nä hrboden, inside the brood chamber (e.g., Claus 1877, Potts and Durning 1980, Egloff et al. 1997, Dumont and Negrea 2002). Branchiopods are well known for producing diapause eggs, or resting eggs, as a part of their life cycle, which appear when conditions are unfavorable (winter or drought), and in the diverse Anomopoda the carapace is used in various ways as a container for the shed resting eggs (ephippia), a habit that is unique among branchiopods (Fryer 1996). Probably familiar to most students of zoology is the characteristic ephippium of Daphnia (Fig. 4.7C), which is a modified part of the carapace housing two resting eggs and is the most specialized of such structures among anomopods. In many anomopods, such as macrothricids and chydorids, the ephippium is less elaborate than that of Daphnia and basically consists of the entire, unmodified carapace that after molting separates from the individual and comes to persist as an independent envelope for a number of eggs. Even the simplest ephippia can withstand drought (Fryer 1996). The ephippia of Daphnia with its resting eggs inside probably ensures that the population can be continued after the winter or after drought but probably also plays a role in dispersal. In some taxa, such as Streblocerus serricaudatus (Fig. 4.7D,E), the ephippium is attached to vegetation and so reduces the chance of dispersal, which suggests that it is often advantageous to ensure persistence of a population instead of undertaking the risks involved in dispersing (Fryer 1996).

The Crustacean Carapace A

C

D

Eggs mating

sexual cycle Ephippium sexual egg

Ephippium (carapace) Carapace

hatching after diapause

Embryonised larvae

B

parthenogenetic cycle parthenogenetic daughter

haploid egg formation

E

parthenogenetic son Carapace Carapace

F

Carapace

G

Embryo

H

Carapace

Eye egg Antennule

Pleopod 1

Fig. 4.7. Carapace as brood chamber in various crustaceans. (A) Lynceus brachyurus (Branchiopoda, Laevicaudata) with a cluster of eggs in a dorsal brood chamber between body and carapace valves (from Sars 1896). (B) Cyclestheria hislopi (Branchiopoda, Cyclestherida) with well-developed embryos in dorsal brood chamber attached to limb filaments (from Sars 1887). (C) Life cycle of Daphnia (Branchiopoda, Cladocera), where a part of the carapace is modified to a characteristic ephippium housing the resting eggs (from Ebert 2005, used with permission from the author). (D) Ephippium of Streblocerus serricaudatus (Branchiopoda, Cladocera) consisting of carapace attached to leaf (from Fryer 1972, with permission from Wiley and Sons). (E) Embryos of S. serricaudatus near the point of hatching from a resting egg (from Fryer 1972, with permission from Wiley and Sons). (F) Vestalenula cornelia (Ostracoda) with right valve removed (from Smith et al. 2006, with permission from the Royal Society of London). (G) Thermosbaena mirabilis (Thermosbaenacea) with dorsal brood chamber containing embryos (from Barker 1962, with permission from the Company of Biologists, Ltd.). (H) Amphionides reynaudii (Amphionidacea), which has been proposed to have a ventral brood chamber between carapace valves with elongate first pleopods forming the roof of the chamber (but brooding never directly observed) (from Williamson 1973, with permission from Brill).

For Ostracoda, Maddocks (1992) reports that many burrowing and swimming forms brood their eggs and young instars within the posterior part of the carapace (Fig. 4.7F). In many taxa, however, eggs are laid individually or in clutches on plants or other substrata before development (Maddocks 1992). Brooding was recently also reported for a Silurian ostracod where a specimen with eggs, perhaps even juveniles, has been described, demonstrating that brooding is a very conserved strategy within ostracods (Siveter et al. 2007). Within Thecostraca, many species of Ascothoracida brood their offspring inside the carapace that is sometimes much enlarged (Kolbasov et al. 2008). Grygier and Fratt (1984) reported that, for Ascothorax gigas, a parasite in the bursae of the brittlestar Ophionotus victoriae, the number of offspring varies between a few hundred and well more than a thousand per female.

119

120

Functional Morphology and Diversity They identified five distinct larval stages, which, according to the authors, may not represent all the instars that A. gigas passes through before release. Within the Malacostraca, only the Thermosbaenacea brood the embryos dorsally under a swollen carapace. Barker (1962) examined Thermosbaena mirabilis and found that the carapace in breeding females is greatly enlarged by successive molts to form a brood pouch, which carries an average of 10 embryos constantly agitated by the inhalant respiratory current (Fig. 4.7G). Barker (1962) suggested that this current may also play a role in sucking the embryos into the brood pouch after they have appeared from the posteriorly directed vaginae. The embryos emerge from the brood pouch as miniature adults. For the aberrant Amphionides reynaudii (Malacostraca, Amphionidacea), Williamson (1973) suggested that the enlarged female carapace forms a brood chamber where the long, anteriorly directed first pleopods form the roof of the chamber (Fig. 4.7H). However, specimens with eggs or embryos within this putative brood pouch are yet to be discovered.

ONTOGENY, EVOLUTION, AND MORPHOLOGICAL DIVERSIT Y OF THE CARAPACE IN SELECTED CRUSTACEA The Branchiopod Carapace The Branchiopoda is an ideal case study on carapace evolution since this taxon is fairly well defined but still offers some variation in the morphology and function of the carapace. Branchiopod carapaces include a broad range of types: a more or less flattened dorsal shield (Notostraca, Fig. 4.1A; see also Fig. 4.10C, below), bivalved carapaces sometimes capable of enclosing the whole body (clam shrimps and cladocerans, Figs. 4.1B, 4.4, 4.5A, 4.7A–E; see also Figs. 4.9, 4.11C, 4.12C, below), and dorsally attached brood pouches showing only little resemblance to the carapace in other branchiopods (raptorial cladocerans, Fig. 4.1C; see also Figs. 4.13C, 4.14C, below). As suggested by Walossek (1993) and appreciated by Fryer (1996) in his review of the branchiopod carapace, the well-preserved Cambrian microfossil Rehbachiella kinnekullensis Mü ller, 1983 provides important information on early carapace development in Branchiopoda. Rehbachiella has been suggested by Walossek (1993) to be an early branchiopod, and while there is room for discussion of its precise phylogenetic position, a branchiopod affinity seems most convincing. The carapace of Rehbachiella and its growth mode are taken here as an indication of how the ancestral branchiopod carapace looked and are therefore summarized in the following. Walossek (1993) showed very clearly that the carapace of Rehbachiella originates as a simple extension of the original naupliar shield. In the earliest stage, the naupliar shield covers an area dorsally that corresponds to the naupliar appendages: the first and second antennae and the mandibles (Fig. 4.8A). A few stages later, the naupliar shield includes also the first maxilla segment (Fig. 4.8B) but not the second maxilla segment. Later the shield starts to grow into what can now be termed a carapace since it has a free posterior fold starting to overgrow the second maxilla segment and more segments posteriorly (Fig. 4.8C). Free folds also develop laterally and start overgrowing the limbs. At a later stage, when the posterior margin of the carapace has become freed, the dorsum of the second maxilla has become fused with the carapace (Fig. 4.8D), so that the carapace fold at this and later stages appears topographically to be an extension of the posterior margin of the second maxilla segment, while it can be argued that it is more correct to consider it as developing from the first maxilla segment. At the latest known stage, the carapace of Rehbachiella is attached in the “cephalic” region (which includes the maxilla 2 segment) and has developed free folds posteriorly and laterally overhanging about seven somites posteriorly and the proximal parts of the limbs (including cephalic appendages).

The Crustacean Carapace B

Maxilla 2

Antennule

Maxilla 1 Mandible

Carapace

A

C

Antenna 2 Antennule

Antenna 2 Mandible

Antennule

D Antenna 2

Mandible

Maxilla 2 Maxilla 1

Carapace

Mandible Antennule Antenna 2 Maxilla 1

Maxilla 2

Fig. 4.8. Carapace development in Rehbachiella kinnekullensis, a Cambrian crustacean that most likely is a close relative to the Branchiopoda, from Walossek’s (1993) series A (A–C) and series B (D), not to same scale. Figures used with permission from Blackwell Publishing.

Given the presumed close relationship between Rehbachiella and other branchiopods, it seems safe to assume that the carapace types in other branchiopods have evolved from a Rehbachiella type of carapace. The phylogeny of Branchiopoda is relatively well understood (Richter et al. 2007, Olesen 2009) (Fig. 4.9), from which a number of conclusions can be drawn. The first off-split within Branchiopoda is the anostracan lineage (Sarsostraca), which consists of the recent Anostraca and the Devonian fossil Lepidocaris rhyniensis, both of which lack a carapace. Based on parsimony, a carapace may have been lost in this lineage. It is not yet certain which taxon constitutes the next branch within Branchiopoda. Traditionally (based on morphology), Notostraca has been considered the sister group to a monophyletic Diplostraca (= clam shrimps and water fleas), but most molecular work has suggested the Notostraca as an ingroup of Diplostraca (Stenderup et al. 2006, Richter et al. 2007) (Fig. 4.9, dashed line with question mark). This uncertainty holds importance for the idea of carapace evolution in the Notostraca. If Notostraca really is a diplostracan ingroup, then it may be assumed that the notostracan ancestor had some kind of bivalved carapace as seen in Spinicaudata, Cyclestherida, and Laevicaudata. In this context, it is interesting that the carapace of Triops cancriformis has a clear paired anlage in early larvae (Møller et al. 2003) (Fig. 4.10A), very similar to that seen in the larvae or embryos of Spinicaudata, Cyclestherida, and Haplopoda (Olesen 1999, Olesen and

121

Functional Morphology and Diversity (6) (10)

(4)

(12)

(5)

(8) (3) (7)

(1)

(13)

(9)

(11)

(2) )

Re hb ac hie lla  1) An os tra ca 2 ) Le pid oc ari s 3 Ca ) str ac oll is No 4) tos tra ca 5 Ka ) za kh art hr La a 6 ev ) ica ud ata 7) Sp ini ca ud ata Cy 8) cle sth eri a 9) Ct en op od a 10 An ) om op od a 11 Le ) pto do ra 1 2) On yc ho po da 1 3

122

Gymnomera ? Sarsostraca

Cladocera Cladoceromorpha

Calmanostraca

Diplostraca

Phyllopoda Branchiopoda (s. str.) Branchiopoda (s. lat.)

Fig. 4.9. Phylogeny of the Branchiopoda showing variation in the morphology of the branchiopod carapace (in gray). Dashed line with question mark indicates alternative position of the notostracan clade (Calmanostraca). From Olesen (2007), with permission from the Crustacean Society.

Grygier 2003, 2004, Olesen et al. 2003). Despite the similar developmental origin in the mentioned taxa, the subsequent development leads to adult carapace types that are quite different: in Notostraca, the paired larval anlagen are precursors of a large univalved dorsal plate continuous with the head in the adults (Fig. 4.1A, 4.10C), whereas in the Spinicaudata and Cyclestherida the paired anlagen develop into the bivalved carapace, where each anlage corresponds to the left or right valve (Fig. 4.11). The paired anlagen of the univalved carapace in Notostraca may be an indication of a bivalved origin independently of whether or not Notostraca is a diplostracan ingroup. The longitudinal dorsal keel seen on the carapace of Triops, for example (e.g., Fig. 4.1A), may be a reminiscence of an earlier simple articulation like the ones seen in recent Spinicaudata. The appearance of a univalved carapace in Notostraca from a bivalved carapace of some type may have evolved alongside a shift in feeding strategy from filtration, as seen in Anostraca and Spinicaudata, to a benthic lifestyle as omnivorous scavengers and opportunistic predators. The Kazacharthra, a Jurassic extinct sister taxon to Notostraca (McKenzie and Chen 1999, Olesen 2009), also has a carapace that apparently is a dorsal plate, but not enough details are known to be considered in this context. Walossek (1993) noted that the carapace in diplostracan branchiopods (clam shrimps and cladocerans) is “disconnected” from the head in the sense that the carapace in these taxa does not appear as a simple posterior and lateral growth of the margins of a head shield as it does in Rehbachiella. Consequently, he termed the diplostracan carapace a secondary shield

The Crustacean Carapace

Carapace

Carapace anlagen

Mandible Antenna 1

Antenna 1 A

100 μm

Antenna 2

Antenna 2 100 μm

B

Carapace

1 mm

C

Fig. 4.10. Carapace development in Triops (Notostraca): (A and B) stages I and II (from Møller et al. 2003, with permission from Wiley and Sons, Ltd.) and (C) adult (from Olesen 2009, with permission from Senckenberg Gesellschaft f ü r Naturforschung).

Carapace

A

Carapace

50 μm Carapace

B

50 μm

C

500 μm

Embryonised larva

Fig. 4.11. Carapace development in Cyclestheria hislopi (Cyclestherida): lateral (A) and dorsal (B) view of embryonized larva and (C) adult female with left carapace valve removed. From Olesen (1999), with permission from Wiley and Sons, Ltd.

123

124

Functional Morphology and Diversity (e.g., Cyclestheria embryos in Fig. 4.11A,B). I agree with Walossek (1993) that this is a significant evolutionary novelty in branchiopod carapace formation and therefore a synapomorphy at some level. As explored elsewhere (Richter et al. 2007, Olesen 2009), there is some evidence that Notostraca qualifies to be included in this group, since the carapace anlage in early larvae of Triops (Fig. 4.10A) is disconnected from the head region in a way similar to that of Cyclestheria hislopi, for example (Fig. 4.11A). But setting this aside, the lack of a (probably lost) carapace in the anostracan lineage leaves uncertainty as to how early in branchiopod evolution this disconnection evolved between an anterior head shield and a posterior free carapace fold. It could have been as early as in the branchiopod ancestor. In any case, the disconnection has taken place, and it is therefore relevant to consider how. The ontogeny of Lynceus may provide a clue since it exhibits both types of carapaces but in two different parts of its life cycle. In the larvae, the carapace is an extension of the head shield as in Rehbachiella but with a unique morphology (Olesen 2005) (Fig. 4.12A,B). In juveniles and adults, the carapace is large, bivalved, noncontinuous with the “head,” and clearly of the “secondary shield” type as outlined by Walossek (1993) (Figs. 4.7A, 4.12C). The shift in carapace morphology takes place from one instar to the next. This abrupt change in morphology between two stages may in some way reflect what took place in evolution. Hence, Lynceus exhibits at the same time, but in different stages, a plesiomorphic head shield type of carapace (in larvae) and an apomorphic “disconnected” secondary shield type of carapace (juveniles and adults).

Carapace

A

Carapace

50 μm Carapace

B

Antenna 1 50 μm

Labrum

C

500 μm

Fig. 4.12. Carapace development in Lynceus (Laevicaudata): (A and B) dorsal (A) and lateral (B) view of larva of L. brachyurus (from Olesen 2005, with permission from Wiley and Sons, Ltd.) and (C) frontal view of adult of L. tatei (from Olesen 2009, with permission from Senckenberg Gesellschaft f ü r Naturforschung).

The Crustacean Carapace Cladoceran carapace origin and evolution pose other interesting questions concerning carapace evolution. Cladocerans are often thought to have evolved neotenically from clam shrimp ancestors, with free-living clam shrimp larvae as the starting point (e.g., Schminke 1981). However, since it is now known that Cyclestheria hislopi (Cyclestherida), which has embryonized larvae (see Olesen 1999), is the sister taxon to Cladocera, free-living clam shrimp larvae cannot have been the starting point for an eventual neotenic origin of cladocerans. If neoteny (more generally, heterochrony) has been involved, then the embryos of Cyclestheria hislopi would have instead been the starting point. It is striking to note that all it takes to “make” a cladoceran carapace from an Cyclestheria-like ancestor is to stop the carapace development of Cyclestheria before it starts overgrowing the head region (e.g., at stage VII; see Olesen 1999). Most cladocerans have a bivalved carapace where left and right side valves cover the body and the trunk limbs (the head is free) (e.g., Figs. 4.4, 4.5A, 4.7C). But in the raptorial/predatory cladocerans, Onychopoda and Haplopoda, the carapace has been modified further and constitutes only a dorsal brood pouch, with no free valves covering the body and trunk limbs. This has been taken to an extreme in the large, predatory Leptodora kindtii (Haplopoda), where the carapace in adults is a saclike structure placed very far behind on the body (Fig. 4.13C). In this respect, L. kindtii is different from other Crustacea, where the carapace in adults often is a posterior extension of the cephalic region or at least is placed very close to this. Interestingly, in the early ontogeny of Leptodora, the carapace appears as a narrow dorsal swelling behind the mandibular region (Fig. 4.14A), which is far more anterior than the position of the carapace/ brood pouch in the adult female would indicate. During development, a posterior displacement of the carapace takes place that can be followed step by step by studying the embryonized larvae in the female’s brood pouch. After having appeared as a narrow, dorsal anterior fold, the carapace becomes a dorsal flap (Fig. 4.14B,C) that gradually grows in size while it apparently moves posteriorly until it constitutes a large, dorsal brood pouch as found in adults (Samter 1895, Olesen et al. 2003) (Fig. 4.13). Olesen et al. (2003) suggested the anterior part of the carapace has gradually fused to the dorsal part of the thorax during ontogeny, leaving only the posterior parts of the valves free. The posterior displacement of the carapace in the ontogeny probably reflects what took place during evolution of the characteristic brood pouch in Leptodora. A comparable ontogeny has been described for certain malacostracans where the carapace also “fuses” with the thoracomeres (see “Ontogeny of the Carapace in Some Malacostracans with Free Larvae,” below). The Malacostracan Carapace Morphological Diversity of the Malacostracan Carapace A carapace is present in many malacostracans and may have played a major role in the success of this highly diverse taxon in combination with other features of the caridoid facies, such as the muscular pleon (Hessler 1983, Hessler and Watling 1999). Not all taxa have a carapace, but those that do exhibit a variation (Fig. 4.2) that is fascinating not only from a functional perspective (see summary of carapace functions above) but also from an evolutionary-morphological perspective. The diverse Malacostraca with their variation in carapace sizes, cephalothorax sizes, and development modes have provided much room for discussing carapace homologies, challenged in complexity perhaps only by an even more convoluted discussion on limb homologies. However, the carapace/cephalothorax structures exhibited by malacostracans indeed provide a toolbox for considering very important aspects of malacostracan evolution.

125

126

Functional Morphology and Diversity

A B

C

Fig. 4.13. Carapace development in Leptodora kindtii (Cladocera): three different developmental stages (A–C) showing that the free part of the carapace (brood pouch, arrows) are migrating in posterior direction during development. From Samter (1895), with permission from Elsevier.

Carapace anlage

Carapace

Antenna 1

Antenna 2 Mandible Trunk limb 1

A

50 μm

100 μm

B

Carapace Antenna 2

Antenna 1 Labrum C

100 μm

Trunk limb 1

Fig. 4.14. Carapace development in Leptodora kindtii (Cladocera). (A) Lateral view of head with early carapace lobe behind the head. (B) Lateral view of head and thorax showing the early carapace as a small lobe having migrated slightly more posterior than in A. (C) Juvenile with a small carapace lobe. From Olesen et al. (2003), with permission from Wiley and Sons, Ltd.

The Crustacean Carapace Taxa such as amphipods and isopods lack a carapace entirely. Only a few taxa, Leptostraca and Mysidacea, have as adults what could be characterized a classical Calman type of carapace enveloping a larger part of the thoracic region (Figs. 4.2A,E, 4.6A,B; see also Fig. 4.16A,B, below). In both decapods and euphausiids, the head and thorax regions are combined into an unsegmented cephalothorax (Fig. 4.2D,F,H), which is also the case for stomatopods (Fig. 4.2B) but involving a smaller part of the thorax. Other malacostracans, such as cumaceans, have a shorter carapace fused with the first three thoracic segments but with free lateral lobes enclosing the appendages that serve for respiration and feeding (Figs. 4.2G, 4.6D–F). Thermosbaenaceans (at least Thermosbaena mirabilis) are unique among malacostracans since females carry the developing embryos dorsally beneath the carapace (Fig. 4.7G). Evolution of the Malacostracan Carapace One of the key questions concerning carapace evolution in the Malacostraca is whether a Calman type of carapace was present in the common malacostracan ancestor, and the lack of it, or various modifications, is therefore secondary. Calman suggested that a free carapace enveloping the thoracic region was a primitive attribute of the Malacostraca along with a number of other shrimplike characteristics (his “caridoid facies”) (Fig. 4.15A). This suggestion was based on the straightforward observation that a carapace of that type occurs in what he called the more “primitive members” of the main taxa into which he divided the Malacostraca (e.g., Mysidacea within Peracarida). This view has been adopted by later authors such as Hessler (1983) and Newman and Knight (1984). Richter and Scholtz (2001) also interpreted the lack of a carapace as a derived character within the Malacostraca. On the contrary, Dahl (1983, 1991) preferred a carapaceless Anaspides-like type of “caridoid facies” (= morphology of common ancestor) for the Malacostraca. Also, Watling (1999), based on an alternative phylogenetic hypothesis, suggested that the “caridoid facies” with its Calman-type carapace was not a part of the ground pattern for the Malacostraca but rather a specialization in a caridoid clade consisting of Lophogastrida, Mysida, Euphausiacea, and Decapoda. Any considerations of character evolution, such as carapace evolution, should preferably be done on the background of a well-supported phylogeny. For the Malacostraca, despite the attempts mentioned above, it has been surprisingly difficult to obtain a robust result, and the existing molecular and morphological evidence seems insufficient (Jenner et al. 2009, Richter et al. 2009). However, if Leptostraca, which also have a large carapace enveloping the thoracic region (Figs. 4.2A, 4.16A,B) is indeed the sister group to the remaining malacostracans, as perceived by many authors (e.g., Siewing 1963, Richter and Scholtz 2001, Meland and Willassen 2007, Wirkner and Richter 2009, Regier et al. 2010), then it seems simplest to assume its presence in the malacostracan ancestor, retained largely unchanged in the Mysidacea, modified or even lost in other taxa, exactly as was implied by Calman (1909). Ontogeny of the Carapace in Some Malacostracans with Free Larvae Within the Malacostraca, it seems that dendrobranchiate decapods, euphausiids, and stomatopods are particularly useful for considering early carapace evolution since, in these taxa, the carapace during early larval development is a free fold attached to the posterior margin of the head region (e.g., Figs. 4.15B,C, 4.16C–F). In adults of these taxa, the free carapace has been transformed into a cephalothorax involving a varying number of thoracic somites (e.g., Fig. 4.2B,D,F,H). The presence of a naupliar sequence in the development of euphausiids and dendrobranchiate decapods is most often considered primitive for Malacostraca. Scholtz (2002),

127

128

Functional Morphology and Diversity Carapace A

Carapace B

Carapace C

Fig. 4.15. Calman’s (1909) hypothetical malacostracan ancestor and two malacostracan free-living larvae with the carapace as posterior extension of the head shield not involving thorax segments. (A) Calman’s (1909) illustration of a generalized type of Malacostraca showing a morphology that he called “caridoid fascies” (shrimplike). He suggested that a large carapace overhanging the thorax was ancestral to Malacostraca. (B) Antizoea of Lysiosquilla eusebia (Stomatopoda) (from Gurney 1942, with permission from the Ray Society/Natural History Museum, London). (C) Protozoea of Penaeopsis species (from Gurney 1942, with permission from the Ray Society/Natural History Museum, London).

on the contrary, suggested secondary reappearance of free-living nauplii within Malacostraca, based partly on parsimony using the phylogeny of Richter and Scholtz (2001). However, since the phylogeny of Malacostraca is still highly uncertain ( Jenner et al. 2009, Richter et al. 2009), which weakens parsimony arguments, and since the early larvae (naupliar sequences) of for example, both euphausiids and dendrobranchiates in many respects are indeed very similar to larvae of nonmalacostracans, it is here assumed that an anamorphic development, probably in many respects similar to that of dendrobranchiate decapods, is ancestral for Malacostraca. Considering dendrobranchiate development, it is really striking that the early carapace of protozoea larvae of Penaeopsis (Gurney 1942, Fig. 4.15C) or Penaeus monodon (Fig. 4.16E,F), for example, is very similar to that of the “Orsten” crustaceans Rehbachiella (Fig. 4.8) or Bredocaris. The carapace in all of these taxa is an extension of the posterior margin of the naupliar shield. And, as in Rehbachiella, it seems in Penaeopsis as though the segment of the second maxilla is not

The Crustacean Carapace

Carapace

100 μm

A

100 μm

B Carapace

Carapace

200 μm

C E

Carapace

F

200 μm

D Carapace

50 μm

50 μm

Fig. 4.16. Carapace development in various malacostracans. (A and B) Nebalia longicornis (Leptostraca) (from Olesen and Walossek 2000, with permission from Springer): lateral view of late juvenile (A) and posterior view showing dorsal attachment of carapace (B). (C and D) Lateral (C) and posterior (D) view of antizoea of Squilla species (Stomatopoda). (E and F) Protozoea of Penaeus monodon (Penaeidae) (material provided by G. Scholtz; data in Biffis et al. 2009): early protozoea with free posterior carapace fold (E) and later protozoea where carapace has started to fuse to thorax (F). The arrow points at the cuticle between the carapace and the thorax, in the process of shifting backward.

incorporated into the carapace in early stages (Fig. 4.15C), highlighting that a carapace “originating” from a second maxilla segment is not crucial when considering carapace homologies among different taxa (see discussion below). It has long been known that the free carapace in larvae of dendrobranchiate decapods develops into the well-known decapod type of cephalothorax, but the exact way this happens ontogenetically has been difficult to elucidate. According to Newman and Knight (1984), who looked at the development of several dendrobranchiates and one euphausiid, it takes place as follows. In the first protozoea, the carapace is represented by a backward and somewhat lateral extension of the shield, which covers a portion of the thorax. Then, during ontogeny, which is achieved by successive ecdyses, the thoracic and carapace cuticles are shed, and the underlying new carapace cuticle shifts farther back on the thorax as molting progresses, until the posterior margin of the

129

130

Functional Morphology and Diversity last thoracic segment is reached. Also, as stated by Newman and Knight (1984), in effect the cuticle underlying the carapace and overlying the thorax is withdrawn, and in the process, the tissue of the two structures melds or fuses together. Thus, the dorsal surface of the cephalothorax in the adult is formed by the previous surface of the carapace in the larva. Figure 4.16F shows a protozoea stage of Penaeus monodon caught in the middle of transformation of the carapace from a free shield to a cephalothorax (the arrow points at the cuticle between the carapace and the thorax, in the process of shifting backward). Casanova (1991, 1993) and Casanova et al. (2002) use another terminology for describing carapace formation in a number of malacostracans, but there seems to be a basic congruence with the description of Newman and Knight (1984). Casanova et al. (2002) described the fusion between the carapace and the thorax in Penaeus indicus as involving a “splitting” of the thorax tergites. It is well known that there is a fundamental connection between development and evolution of living organisms. Any change in morphology must necessarily involve a change of something in development, for example, a change in the rate or timing of the ontogeny of certain structures (= heterochrony). With respect to the decapod carapace, it is tempting to suggest that the way in which the cephalothorax is formed ontogenetically in, for example, dendrobranchiate shrimps, as described by Newman and Knight (1984) and Casanova et al. (2002), roughly ref lects the way in which the decapod cephalothorax approximately appeared originally during evolution, that is, as a gradual withdrawal of cuticle between the carapace and thorax.

ON CARAPACE HOMOLOGIES AND EARLY CARAPACE EVOLUTION Calman’s Crustacean Carapace Hypothesis: Critical Survey of Recent Discussions The term carapace has been used for Crustacea prior to Calman (1909), but since he gave a brief and precise summary of the occurrence of the carapace within the Crustacea, his work has often been used as a starting point when discussing its evolution. Because his summary has been referred to so often in later works and has occasionally been rebutted (e.g., Dahl 1991), a brief outline of what Calman actually wrote is given here. Calman (1909) suggested that since a dorsal shield or carapace (he used both terms) occurs in the most diverse groups of Crustacea, it is probably a primitive attribute of the taxon. He also said that it originates as a fold of the integument from the posterior margin of the cephalic region. Then he outlined some of the various forms a carapace can take, such as in notostracan branchiopods, where the carapace loosely envelopes more or less of the trunk, or as in other branchiopods (clams shrimps) and ostracods (mussel shrimps), where it forms a bivalved shell completely enclosing body and limbs. Finally, he mentioned some of the extreme cases, such as adult cirripedes, where it forms a fleshy mantle usually strengthened by shelly plates, and some malacostracans, where the carapace has coalesced with the tergites of some or all thoracic somites, though it may project freely at the sides, overhanging, as in Decapoda, the branchial chambers. This short summary of Calman was excellent for the time and encapsulates a number of the questions still discussed today. For example, the question of whether or not Crustacea originally had a carapace was treated by Hessler and Newman (1975), who depicted two different possible “urcrustaceans,” one with a carapace and one without. We have still not reached an entirely convincing solution to this question, which, given the age, diversity, and morphological disparity of Crustacea, is not too surprising (see chapter 1 in this volume). Even Hexapoda (insects and allies) may be an ingroup of Crustacea (e.g., Regier et al. 2010), which highlights the great variation in crustacean body plans. Since a crustacean/arthropod carapace is an adaptation

The Crustacean Carapace to an aquatic lifestyle, no such structure is needed in the terrestrial/aerial hexapods and must have been reduced if hexapods eventually turn out to have originated from carapace-bearing crustaceans, or it may never have been present in the lineage leading to Hexapoda. Dahl (1983, 1991) spent much effort attempting to weaken/invalidate Calman’s idea of the carapace as a primitive (and therefore homologous) attribute of the Crustacea. The evolutionary status of the carapace for Crustacea as a whole is still uncertain, but at least some of the main arguments used by Dahl were problematic. The essence of Dahl’s view seems to be that if structures (e.g., a carapace) have dissimilar ontogenesis in different taxa, then they cannot be homologous. He argued that since carapace structures originate ontogenetically in so many different ways in Crustacea and, in Dahl’s (1991) view, seemingly never as a dorsal fold derived from the cephalon, then this invalidates the “carapace hypothesis” of Calman (1909). Indeed, the early ontogeny of the carapace in Crustacea is very diverse, and certainly a straightforward ontogeny from the rear of the cephalon is rare, but is nevertheless seen in various malacostracan free-living larvae (see above and Figs. 4.15B,C, 4.16C–F), in branchiurans, in various thecostracans, in “Orsten” crustaceans such as Rehbachiella (Fig. 4.8), and in branchiopods such as Triops (Fig. 4.10). Dahl (1991) distinguished a number of different carapace types named after the part of the body that gives rise to the early carapace fold, and even though he said that the terms used for describing these were “purely descriptive and without evolutionary or phylogenetic implications,” this was nevertheless probably his most important argument for considering Calman’s “carapace hypothesis” as invalid. He named six types of early carapace ontogenies that differed with respect to the specific location of the early carapace fold(s) (dorsal, lateral, etc.) and with respect to how much of the body is involved in forming the early carapace lobe (e.g., how many segments) (Dahl 1991). Watling (1999) summarized Dahl’s arguments and highlighted that early carapace formation in the short-carapace peracarids, such as Cumacea, takes place as lateral outpouchings (so-called branchiostegal folds). However, while it obviously speaks in favor of homology if the carapace in various taxa can be shown ontogenetically to have the same origin, the opposite is not necessarily the case: structures, such as a carapace, may still be homologous despite having dissimilar ontogenetic origins (for a summary of the connection between ontogeny and evolution, see de Beer 1958, 149–153). It is well known that a given ontogenetic sequence in itself is also subject to evolutionary modifications, so the dissimilar ontogeny of various carapaces in Crustacea does not necessarily indicate convergent evolution; they could with equal right be explained as modifications of the ontogenetic sequence. Much attention has been devoted to whether or not the “origin” of the carapace is maxillary. Calman (1909) was cautious enough to specify the “origin” of the carapace as being in the “cephalic region ,” while later authors, including Dahl (1991), have taken this as Calman (1909) specifically suggesting that the carapace folds originated from the rear of the second maxilla segment. It is difficult to trace at what stage Calman’s carapace hypothesis became modified from Calman’s broad statement to a more specific definition as the one referred to by Dahl (1983, 198), who writes that “according to the classical concept the carapace is a fold growing out from the maxillary segment.” But it seems that much of the discussion and confusion of carapace homologies within Crustacea could have been avoided if the focus had not been so much on whether the carapace is maxillary. As a matter of fact, there is no reason for considering an ontogenetic origin of the carapace from the second maxilla segment as the holy grail when it comes to carapace homologies within Crustacea. As outlined above, a dissimilar ontogeny does not necessarily invalidate homology. The development of the carapace in the Cambrian Rehbachiella kinnekullensis is here of special interest since, as outlined above, the carapace in different phases of the development integrates a varying proportion of the cephalic region. For example, at a certain stage, the carapace extends from the rear of the first

131

132

Functional Morphology and Diversity maxilla region (Fig. 4.8C) and later from the rear of the second maxilla (Fig. 4.8D) (Walossek 1993). In the same line, as highlighted by Dahl (1991), taxa such as Notostraca and Leptostraca (see Fig. 4.16A,B) have carapaces that do not extend entirely freely from the cephalic region but actually are fused to the dorsum of one or more thoracic segments. However, an integration of a few extra segments into the carapace does not necessarily indicate nonhomology to the carapace of taxa where this is not the case. Contribution to Carapace Discussion by Well-Preserved Cambrian “Orsten” Fossils Important evidence on early carapace evolution within the Crustacea is provided by the studies on Cambrian microfossils such as Rehbachiella kinnekullensis and Bredocaris admirabilis (see Mü ller and Walossek 1988, Walossek 1993). The early carapace development in both taxa is rather similar and can be described as a simple growth of the posterior margin of the naupliar shield present earlier in development, resulting in free posterior shield margins covering a smaller or larger part of the trunk (Fig. 4.8). Actually, as already recognized by Fryer (1996), such a simple type of carapace growing from the rear of the head region as seen in Rehbachiella is indistinguishable from what was suggested primitively for Crustacea by Calman (1909). Walossek (1993, 198) used the term “cephalic shield” for Rehbachiella but was aware of the similarity to the “carapace” as this structure is understood by other authors since, for example, he stated that the ‘free carapace’ of Newman and Knight (1984) is more or less synonymous with the cephalic shield of Rehbachiella and other taxa. As mentioned above, one of the much-discussed dogmas concerning carapace homologies has been the assumed origin from the second maxilla segment (“cephalic region” in Calman’s terms). Walossek (1993) has clarified that this is a simplified discussion since in both Rehbachiella and Bredocaris the exact segment from which the free carapace margins extend depends on which developmental stage is considered (Fig. 4.8). Of course, one should be careful assuming that the mode of carapace development in Rehbachiella is ancestral just because Rehbachiella is old. Regardless, from an evolutionary point of view, the carapace seen in Rehbachiella would seem as an ideal starting point for carapaces in crustaceans such as malacostracans, branchiopods, branchiurans, and thecostracans. However, the information provided by Cambrian “Orsten” fossils is ambiguous since some taxa lack a carapace. Skara (see Mü ller and Walossek 1985), for example, is lacking a carapace that could be an adaptation to an interstitial lifestyle. The even older Yicaris also lacks a carapace and has a shield restricted to only the head region (Zhang et al. 2007).

CONCLUSIONS The Crustacea exhibit a large variation in carapace structures with a wealth of functions intimately linked to the lifestyle of the taxon. Some well-demonstrated functions are reduction of drag during swimming, crawling, or during temporarily resting on the substratum; as a feeding chamber; as a respiration chamber; and as a brooding chamber. It is clear that the plasticity of the crustacean carapace has been an important component in crustacean evolution. It has often been discussed whether a carapace extending from the rear of the cephalon and enveloping a smaller or larger part of the body was present in the common ancestor to the Malacostraca or even in the common ancestor to the Crustacea. The occurrence of a carapace of this type within the Malacostraca, either as larvae or as adults, seems to suggest, in accordance with Calman (1909), that such a structure was present in the Malacostraca originally, from which it follows that it has been modified or lost in a number of malacostracan subtaxa.

The Crustacean Carapace Whether a carapace was present in the crustacean ancestor is more uncertain because of a larger morphological gap between these taxa, but also similarities in carapace ontogeny in certain taxa (malacostracans, branchiurans, branchiopods, and “Orsten” fossils) may suggest a common origin. The fossil record is ambiguous on this question since some taxa have a carapace (Rehbachiella, Bredocaris, Walossekia) and others lack it (Skara, Yicaris). The carapace in the Cambrian fossil Rehbachiella kinnekullensis probably was close in morphology and ontogeny to such an eventual crustacean ancestor. Differences in carapace ontogeny between different crustacean taxa have sometimes been considered as evidence for parallel evolution of different crustacean carapace types (e.g., Dahl 1991). However, in this chapter it is argued that differences in ontogenetic origin of the carapace do not necessarily indicate nonhomology since an ontogenetic sequence in itself is also subject to evolution.

ACKNOWLEDGMENTS I thank the editors of this book, Martin Thiel and Les Watling, for inviting me to contribute this chapter. Gerhard Scholtz kindly provided specimens of Penaeus monodon used in Fig. 4.16. This treatment greatly benefited from discussions with Jens Høeg, Stefan Richter, Dieter Waloszek, and Les Watling at various occasions over the last years. Martin Thiel, Les Watling, and Stefan Richter all read the chapter and gave useful comments. This work was supported by the Danish Research Council (grant 09–066003).

REFERENCES Abe, K., and J. Vannier. 1995 . Functional morphology and significance of the circulatory system of Ostracoda, exemplified by Vargula hilgendorfii (Myodocopida). Marine Biology 124:51–58. Atkinson, R.J.A., C. Froglia, E. Arneri, and B. Antolini. 1997. Observations on the burrows and burrowing behaviour of Squilla mantis (L.) (Crustacea: Stomatopoda). Marine Ecology 18:337–359. Barker, D. 1962. A study of Thermosbaena mirabilis (Malacostraca, Peracarida) and its reproduction. Quarterly Journal of Microscopical Science 103:261–286. Batang , Z.B., and H. Suzuki. 1999. Gill-cleaning mechanisms of the mud lobster Thalassina anomala (Decapoda: Thalassinidea: Thalassinidae). Journal of Crustacean Biology 19:671–683. Bauer, R.T. 1989. Decapod crustacean grooming: Functional morphology, adaptive value, and phylogenetic significance. Pages 49–73 in B.E. Felgenhauer , L. Watling , and A.B. Thistle, editors. Functional morphology and grooming in Crustacea. Crustacean Issues, Vol. 6. Balkema, Rotterdam. Bauer, R.T. 1998. Gill-cleaning mechanisms of the crayfish Procambarus clarkii (Astacidea: Cambaridae): experimental testing of setobranch function. Invertebrate Biology 117:129 –143. Biffis, C., F. Alwes, and G. Scholtz. 2009. Cleavage and gastrulation of the dendrobranchiate shrimp Penaeus monodon (Crustacea, Malacostraca, Decapoda). Arthropod Structure and Development 38:527–540. Blake, R.W. 1985 . Crab carapace hydrodynamics. Journal of Zoology (London) 207:407–423. Bowman, T.E., S.P. Garner, R.R. Hessler, T.M. Iliffe, and H.L. Sanders. 1985 . Mictacea, a new order of Crustacea Peracarida. Journal of Crustacean Biology 5:74–78. Boxshall, G.A. 2004 . The evolution of arthropod limbs. Biological Reviews. 79:253–300. Boxshall, G.A., and D. Jaume. 2009. Exopodites, epipodites and gills in crustaceans. Arthropod Systematics and Phylogeny 67:229 –254. Calman, W.T. 1909. Crustacea. Pages 1–346 in E.R. Lankester, editor. A treatise on Zoology, Part 7, Fascicle 3. Adam and Charles Black , London . Cannon, H.G. 1927. On the feeding mechanism of Nebalia bipes. Transactions of the Royal Society of Edinburgh 55(2):355–369.

133

134

Functional Morphology and Diversity Cannon, H.G., and S.M. Manton. 1927. On the feeding mechanism of a mysid crustacean, Hemimysis lamornae . Transactions of the Royal Society of Edinburgh 55:219 –253. Casanova, B. 1991. Origine protocéphalique antennaire de la carapace chez les Leptostracés, Mysidacés et Eucaridés (Crustacés). Comptes Rendus hebdomadaires des Séances de l’Académie des Sciences Paris, Série III 312:461–468. Casanova, B. 1993. L’origin protocéphalique de la carapace chez les thermosbaenacés, tanaidacés, cumacés et stomatopodes. The protocephalic origin of the carapace in thermosbaenaceans, tanaidaceans, cumaceans, and stomatopods. Crustaceana 65:144–150. Casanova, B., L. De Jong , and X. Moreau. 2002. Carapace and mandibles ontogeny in the Dendrobranchiata (Decapoda), Euphausiacea, and Mysidacea (Crustacea): A phylogenetic interest. Canadian Journal of Zoology 80:296 –306. Childress, J.J. 1971. Respiratory adaptations to the oxygen minimum layer in the bathypelagic mysid Gnathophausia ingens. Biological Bulletin 141:109–121. Claus, C. 1877. Zur Kenntnis des Baues und der Organisation der Polyphemiden. Denkschrift Kaiserlichen Academie der Wissenschaften, Vienna 37:137–160. Claus, C. 1888. Ü ber den Organismus der Nebaliiden und die systematische Stellung der Leptostraken. Arbeiten aus dem Zoologischen Institut der Universität Wien 8:1–149. Crisp, D.J. 1955 . The behaviour of barnacle cyprids in relation to water movement over a surface. Journal of Experimental Biology 32:569 –590. Dahl, E. 1983. Malacostracan phylogeny and evolution. Pages 189–212 in F.R. Schram, editor. Crustacean phylogeny. Crustacean Issues, Vol. 1. Balkema, Rotterdam. Dahl, E. 1991. Crustacea Phyllopoda and Malacostraca: A reappraisal of cephalic and thoracic shield and fold systems and their evolutionary significance. Philosophical Transactions of the Royal Society London Series B 334:1–26. de Beer, G.R. 1958. Embryos and ancestors. Clarenden Press, Oxford . Dennell, R. 1937. On the feeding mechanism of Apseudes talpa, and the evolution of the Peracaridan feeding mechanisms. Transactions of the Royal Society of Edinburgh 59:57–78. Dingle, H., and R.L. Caldwell. 1972. Reproductive and maternal behavior of the mantis shrimp Gonodactylus bredini Manning (Crustacea: Stomatopoda). Biological Bulletin 142:417–426. Dumont, H.J., and S.V. Negrea 2002. Introduction to the class Branchiopoda. Guides to the Identification of the Microinvertebrates of the Continental Waters of the World, Vol. 19. Backhuys, Leiden . Dyer, M.F., and R.F. Uglow. 1978. Gill chamber ventilation and scaphognathite movements in Crangon crangon (L.). Journal of Experimental Marine Biology and Ecology 31:195–207. Ebert, D. 2005 . Ecology, epidemiology, and evolution of parasitism in Daphnia [excerpt]. U.S. National Library of Medicine, National Center for Biotechnology Information, Bethesda, MD. Available: http://www.ncbi.nlm.nih.gov/books/NBK2036/ (accessed May 16, 2012). Egloff, D.A., P.W. Fofonoff, and T. Onbé 1997. Reproductive biology of marine cladocerans. Advances in Marine Biology 31:79 –167. Farrelly, C.A., and P. Greenaway. 1994 . Gas exchange through the lungs and gills in air-breathing crabs. Journal of Experimental Biology 187:113–130. Fryer, G. 1956. A report of the parasitic Copepoda and Branchiura of the fishes of Lake Nyasa. Proceedings of the Zoological Society 127:293–344. Fryer, G. 1960. The spermatophores of Dolops ranarum (Crustacea, Branchiura): Their structure, formation, and transfer. Quarterly Journal of Microscopical Science 101:407–432. Fryer, G. 1968. Evolution and adaptive radiation in the Chydoridae (Crustacea, Cladocera): A study in comparative functional morphology and ecology. Philosophical Transactions of the Royal Society of London Series B 254:221–385. Fryer, G. 1972. Observations on the ephippia of certain macrothricid cladocerans. Zoological Journal of the Linnean Society 51:79 –96. Fryer, G. 1988. Studies on the functional morphology and biology of the Notostraca (Crustacea: Branchiopoda). Philosophical Transactions of the Royal Society London Series B 321:27–124. Fryer, G. 1991. Functional morphology and the adaptive radiation of the Daphniidae (Branchiopoda: Anomopoda). Philosophical Transactions of the Royal Society of London Series B 331:1–99.

The Crustacean Carapace Fryer, G. 1996. The carapace of the branchiopod Crustacea. Philosophical Transactions of the Royal Society of London Series B 351:1703–1712. Fryer, G., and G.A. Boxshall. 2009. The feeding mechanisms of Lynceus (Crustacea: Branchiopoda: Laevicaudata), with special reference to L. simiaefacies Harding. Zoological Journal of the Linnean Society 155:513–541. Gicklhorn, J. 1925 . Ü ber spezifische und lokale Reduktion von Silber- und anderen Metallsalzen in den Kiemensäckchen von Daphnia. Lotos 73:83–96. Giesbrecht, W. 1910. Stomatopoden. Fauna Flora Golf Neapel 33:1–239. Glenner, H., and J.T. Høeg. 1993. Scanning electron microscopy of metamorphosis in four species of barnacles (Cirripedia Thoracica Balanomorpha). Marine Biology 117:431–439. Gordon, I. 1957. On Spelaeogriphus, a new cavernicolous crustacean from South Africa. Bulletin of the British Museum (Natural History) 5:31–47. Gordon, I. 1960. On a Stygiomysis from the West Indies, with a note on Spelaeogriphus (Crustacea, Peracarida). Bulletin of the British Museum (Natural History) Zoology 6:285–324. Gravier, C., and P. Mathias. 1930. Sur la reproduction d’un Crustacé Phyllopode du groupe des Conchostracés (Cyzicus cycladoides (Joly)). Comptes rendus de l’Académie des sciences 191:183–185. Grindley, J.P., and R.R. Hessler. 1971. The respiratory mechanism of Spelaeogriphus and its phylogenetic significance. Crustaceana 20:141–144. Gruner, H.E. 1993. Arthropoda (ohne Insecta). A. Kaestner, editor. Lehrbuch der speziellen Zoologie. Gustav Fischer, Stuttgart . Grygier, M.J., and D.B. Fratt. 1984 . The ascothoracid crustacean Ascothorax gigas: Redescription, larval development, and notes on its infestation of the Antarctic ophiuroid Ophionotus victoriae. Biology of the Anarctic Seas XVI, Antarctic Research Series 41:43–58. Gurney, R. 1942. The larvae of the decapod Crustacea. Ray Society, London . Haase, W. 1975 . Ultrastruktur und Funktion der Carapaxfelder von Argulus foliaceus L. (Crustacea: Branchiura). Zeitschrift f ü r Morphologie der Tiere 81:161–189. Henninger, H.P., and W.H.I. Watson. 2005 . Mechanisms underlying the production of carapace vibrations and associated waterborne sounds in the American lobster, Homarus americanus. Journal of Experimental Biology 208:3421–3429. Hessler, R.R. 1983. A defense of the caridoid facies; wherein the early evolution of the Eumalacostraca is discussed. Pages 145–164 in F.R. Schram, editor. Crustacean phylogeny. Crustacean Issues, Vol. 1. Balkema, Rotterdam. Hessler, R.R., and W.A. Newman. 1975 . A trilobitomorph origin for the Crustacea. Fossils and Strata 4:437–459. Hessler, R.R., and L. Watling. 1999. Les Péracarides: un groupe controversé. Pages 1–11 in J. Forest, editor. Traité de Zoologie. Anatomie, Systématique, Biologie, Tome VII, Fascicule IIIA. Crustacés Péracarides, Vol. 19. Memoires de l’Institut Oceanographique Fondation Albert I, Monaco. Høeg , J.T., N.C. Lagersson, and H. Glenner. 2004 . The complete cypris larva and its significance in thecostracan phylogeny. Pages 197–215 in G. Scholtz , editor. Evolutionary developmental biology. Balkema, Lisse, The Netherlands. Høeg , J.T., and O.S. Møller. 2006. When similar beginnings leads to different ends: Constraints and diversity in cirripede larval development. Invertebrate Reproduction and Development 49:125–142. Høeg , J.T., M. Pérez-Losada, H. Glenner, G.A. Kolbasov, and K.A. Crandall. 2009. Evolution of morphology, ontogeny and life cycles within the Crustacea Thecostraca. Arthropod Systematics and Phylogeny 67:199 –217. Hong , S.Y. 1988. Development of epipods and gills in some pagurids and brachyurans. Journal of Natural History 22:1005–1040. Jacklyn, P.M., and D.A. Ritz. 1986. Hydrodynamics of swimming in scyllarid lobsters. Journal of Experimental Marine Biology and Ecology 101:85–99. Jenner, R.A., C.N. Dhubhghaill, M.P. Ferla, and M.A. Wills. 2009. Eumalacostracan phylogeny and total evidence: Limitations of the usual suspects. BMC Evolutionary Biology 9:21. Jensen, P.G., J. Moyse, J.T. Høeg , and H. Al-Yahya . 1994 . Comparative SEM studies of lattice organs: Putative sensory structures on the carapace of larvae from Ascothoracica and Cirripedia (Crustacea Maxillopoda Thecostraca). Acta Zoologica 75:125–142.

135

136

Functional Morphology and Diversity Johnson, S.B., and Y.G. Attramadal. 1982. A functional-morphological model of Tanais cavolinii Milne-Edwards (Crustacea, Tanaidacea) adapted to a tubicolous life-strategy. Sarsia 61:29 –42. Kikuchi, S. 1983. The fine structure of the gill epithelium of a fresh-water flea, Daphnia magna (Crustacea: Phyllopoda) and changes associated with acclimation to various salinities. Cell and Tissue Research 229:253–268. Koehl, M.A.R., and J.R. Strickler. 1981. Copepod feeding currents: Food capture at low Reynolds number. Limnology and Oceanography 26:1062–1073. Kolbasov, G.A., M.J. Grygier, J.T. Høeg , and W. Klepal. 2008. External morphology of the two cypridiform ascothoracid-larva instars of Dendrogaster: The evolutionary significance of the two-step metamorphosis and comparison of lattice organs between larvae and adult males (Crustacea, Thecostraca, Ascothoracida). Zoologischer Anzeiger 247:159 –183. Korschelt, E., and K. Heider. 1890. Lehrbuch der Vergleichenden Entwicklungsgeschichte der Wirbellosen Thiere. Gustav Fischer, Jena . Lauterbach, K.-E. 1970. Der Cephalothorax von Tanais cavolinii Milne Edwards (Crustacea— Malacostraca). Ein Beitrag zur vergleichenden Anatomie und Phylogenie der Tanaidacea. Zoologische Jahrbücher Abteilung f ü r Anatomie und Ontogenie der Tiere 87:94–204. Laverack , M.S., D.M. Neil, and R.M. Robertson. 1977. Metachronal exopodite beating in the mysid Praunus flexuosus: A quantitative analysis. Proceedings of the Royal Society of London Series B 198:139 –154. Lilljeborg , W. 1901. Cladocera Sueciae. Nova acta Regiae Societatis Scientiarum Upsaliensis Series 3, 19:1–701. Lockhead, J.H. 1961. Locomotion. Pages 313–364 in T.H. Waterman, editor. The physiology of Crustacea: Sense organs, integration, and behavior. Academic Press, New York. Maas, A., C. Haug , J.T. Haug, J. Olesen, X. Zhang , and D. Waloszek. 2009. Early crustacean evolution and the appearance of epipodites and gills. Arthropod Systematics and Phylogeny 67:255–273. Maddocks, R.F. 1992. Ostracoda. Pages 415–441 in F.W. Harrison and A.G. Humes, editors. Microscopic anatomy of invertebrates, Vol. 9, Crustacea. Wiley-Liss, New York. Maitland, D.P. 1990a . Aerial respiration in the semaphore crab, Heloecius cordiformis, with or without branchial water. Comparative Biochemistry and Physiology 95A:267–274. Maitland , D.P. 1990b. Carapace and branchial water circulation, and water-related behaviours in the semaphore crab Heloecius cordiformis (Decapoda: Brachyura: Ocypodidae). Marine Biology 105:275–286. Maitland, D.P. 1992a . Carapace movements aid air breathing in the semaphore crab, Heloecius cordiformis (Decapoda: Brachyura: Ocypodidae). Journal of Comparative Physiology B 162:375–382. Maitland, D.P. 1992b. Carapace movements associated with ventilation and irrigation of the branchial chambers in the semaphore crab, Heloecius cordiformis (Decapoda: Brachyura: Ocypodidae). Journal of Comparative Physiology B 162:365–374. Martin, J.W., and G.E. Davis. 2001. An updated classification of the recent Crustacea. Natural History Museum of Los Angeles County, Science Series 39:1–124. Martin, J.W., and G.E. Davis. 2006. Historical trends in crustacean systematics. Crustaceana 79:1347–1368. Mauchline, J., and T. Nemoto. 1977. Integumental sensilla of diagnostic value in euphausiids. Journal of the Oceanographical Society of Japan 33:283–289. Mayrat , A., B.R. McMahon, and K. Tanaka. 2006. The circulatory system. Pages 3–84 in J. Forest , J.C. von Vaupel Klein, and F.R. Schram, editors. Treatise on zoology—anatomy, taxonomy, biology The Crustacea revised and updated from the Traité de Zoologie, Vol. 2. Koninklijke Brill, Leiden . McKenzie, K.G., and P.-J. Chen. 1999. Kazacharthra. Pages 443–458 in E. Savazzi, editor. Functional morphology of the invertebrate skeleton. John Wiley and Sons, Chichester. Meland, K., and E. Willassen. 2007. The disunity of “Mysidacea” (Crustacea). Molecular Phylogenetics and Evolution 44:1083–1104. Møller, O.S. 2009. Branchiura (Crustacea)—survey of historical literature and taxonomy. Arthropod Systematics and Phylogeny 67:41–55.

The Crustacean Carapace Møller, O.S., J. Olesen, and J.T. Høeg. 2003. SEM studies on the early development of Triops cancriformis (Bosc) (Crustacea: Branchiopoda: Notostraca). Acta Zoologica 84:267–284. Mü ller, G.W. 1894 . Ostracoda . Fauna und Flora Golf Neapel 21:1–404. Mü ller, K.J., and D. Walossek. 1985 . Skaracarida, a new order of Crustacea from the upper Cambrian of V ä stergötland, Sweden. Fossils and Strata 17:1–65. Mü ller, K.J., and D. Walossek. 1988. External morphology of the Upper Cambrian maxillopod Bredocaris admirabilis. Fossils and Strata 23:1–70. Newman, W.D., and M.D. Knight. 1984 . The carapace and crustacean evolution—a rebuttal. Journal of Crustacean Biology 4:682–687. Oelze, A. 1931. Beiträge zur Anatomie von Diastylis rathkei Kr. Zoologische Jahrbücher. Abteilung f ü r Anatomie und Ontogenie der Tiere 54:235–294. Olesen, J. 1999. Larval and post-larval development of the branchiopod clam shrimp Cyclestheria hislopi (Baird, 1859) (Crustacea, Branchiopoda, Conchostraca, Spinicaudata). Acta Zoologica 80:163–184. Olesen, J. 2005 . Larval development of Lynceus brachyurus (Crustacea, Branchiopoda, Laevicaudata): Redescription of unusual crustacean nauplii, with special attention to the moult between last nauplius and first juvenile. Journal of Morphology 264:131–148. Olesen, J. 2007. Monophyly and phylogeny of Branchiopoda, with focus on morphology and homologies of branchiopod phyllopodous limbs. Journal of Crustacean Biology 27:165–183. Olesen, J. 2009. Phylogeny of Branchiopoda (Crustacea)—character evolution and contribution of uniquely preserved fossils. Arthropod Systematics and Phylogeny 67:3–39. Olesen, J., and M.J. Grygier. 2003. Larval development of Japanese “conchostracans”: Part 1, larval development of Eulimnadia braueriana (Crustacea, Branchiopoda, Spinicaudata, Limnadiidae) compared to that of other limnadiids. Acta Zoologica 84:41–61. Olesen, J., and M.J. Grygier. 2004 . Larval development of Japanese “conchostracans”: Part 2, larval development of Caenestheriella gifuensis (Crustacea, Branchiopoda, Spinicaudata, Cyzicidae), with notes on homologies and evolution of certain naupliar appendages within the Branchiopoda. Arthropod Structure and Development 33:453–469. Olesen, J., S.T. Parnas, and J.F. Petersen. 2006. Tail flip and escape response of Tethysbaena argentarii (Crustacea: Thermosbaenacea). Journal of Crustacean Biology 26:429 –432. Olesen, J., S. Richter, and G. Scholtz. 2003. On the ontogeny of Leptodora kindtii (Crustacea, Branchiopoda, Cladocera), with notes on the phylogeny of the Cladocera. Journal of Morphology 256:235–259. Olesen, J., and D. Walossek. 2000. Limb ontogeny and trunk segmentation in Nebalia species. Zoomorphology 120:47–64. Patek , S.N., and R.L. Caldwell. 2006. The stomatopod rumble: Low frequency sound production in Hemisquilla californiensis. Marine and Freshwater Behaviour and Physiology 39:99 –111. Pérez Farfante, I. 1988. Illustrated key to penaeoid shrimps of commerce in the Americas. NOAA Technical Report NMFS 64. U.S. Department of Commerce, Washington, DC. Pirow, R., F. Wollinger, and R.J. Paul. 1999a . The importance of the feeding current for oxygen uptake in the water flea Daphnia magna. Journal of Experimental Biology 202:553–562. Pirow, R., F. Wollinger, and R.J. Paul. 1999b. The sites of respiratory gas exchange in the planktonic crustacean Daphnia magna: An in vivo study employing blood haemoglobin as an internal oxygen probe. Journal of Experimental Biology 202:3089 –3099. Potts, W.T.W., and C.T. Durning. 1980. Physiological evolution in the branchiopods. Comparative Biochemistry and Physiology 67B:475–484. Rathbun, M.J. 1930. The cancroid crabs of America of families Euryliidae, Portunidae, Atelecyclidae, Cancridae, and Xanthidae. Bulletin of the American Museum of Natural History 152:1–593. Regier, J.C., J.W. Shultz , A. Zwick, A. Hussey, B. Ball, R. Wetzer, J.W. Martin, and C.W. Cunningham 2010. Arthropod relationships revealed by phylogenomic analysis of nuclear protein-coding sequences. Nature 463:1079 –1084. Richter, S., O.S. Møller, and C.S. Wirkner. 2009. Advances in crustacean phylogenetics. Arthropod Systematics and Phylogeny 67:275–286. Richter, S., J. Olesen, and W.C. Wheeler. 2007. Phylogeny of Branchiopoda (Crustacea) based on a combined analysis of morphological data and six molecular loci. Cladistics 23:301–336.

137

138

Functional Morphology and Diversity Richter, S., and G. Scholtz. 2001. Phylogenetic analysis of the Malacostraca (Crustacea). Journal of Zoological Systematics and Evolutionary Research 39:113–36. Ruello, N.V. 1973. Burrowing, feeding, and spatial distribution of the school prawn Metapenaeus macleayz (Haswell) in the hunter river region, Australia. Journal of Experimental Marine Biology and Ecology 13:189 –206. Rybakov, A.V., J.T. Høeg , P.G. Jensen, and G.A. Kolbasov. 2003. The chemoreceptive lattice organs in cypris larvae develop from naupliar setae (Thecostraca: Cirripedia, Ascothoracida and Facetotecta). Zoologischer Anzeiger 242:1–20. Samter, M. 1895 . Die Verä nderung der Form und Lage der Schale von Leptodora hyalina Lillj. wä hrend der Entwicklung. Zoologischer Anzeiger 483:334–338, 484:341–344. Sars, G.O. 1887. On Cyclestheria hislopi (Baird), a new generic type of bivalve Phyllopoda; raised from dried Australian mud. Christiania Videnskabs-Selskabs Forhandlinger 1:1–65. Sars, G.O. 1896. Fauna Norvegiæ. Bd. I. Beskrivelse af de hidtil kjendte norske Arter af Underordnerne Phyllocarida og Phyllopoda [Descriptions of the Norwegian species at present known belonging to the suborders Phyllocarida and Phyllopoda]. Joint-Stock, Christiania, Norway. Schminke, H.K. 1981. Adaptation of Bathynellacea (Crustacea, Syncarida) to life in the interstitial (“Zoea theory”). Internationale Revue der gesamten Hydrobiologie 66:575–637. Scholtz , G. 2002. Evolution of the nauplius stage in malacostracan crustaceans. Journal of Zoological Systematics and Evolutionary Research 38:175–187. Siewing , R. 1952. Morphologische Untersuchungen an Cumaceen. Zoologische Jahrbücher. Abteilung f ü r Anatomie und Ontogenie der Tiere 72:552–559. Siewing , R. 1958. Anatomie und Histologie von Thermosbaena mirabilis. Ein beitrag zur Phylogenie der Reihe Pancarida (Thermosbaenacea). Abhandlungen der Mathematisch Naturwissenschaftlichen Klasse/Akademie der Wissenschaften und der Literatur in Mainz 7:195–270. Siewing , R. 1963. Studies in malacostracan morphology: Results and problems. Pages 85–110 in H.B. Whittington and W.D.I. Rolfe, editors. Phylogeny and evolution of Crustacea. Museum of Comparative Zoology, Cambridge, MA . Siveter, D.J., D.J. Siveter, M.D. Sutton, and D.E.G. Briggs. 2007. Brood care in a Silurian ostracod. Proceedings of the Royal Society 274:465–469. Smirnov, N.N., and A.A. Kotov. 2009. Morphological radiation with reference to the carapace valves of the Anomopoda (Crustacea: Cladocera). International Review of Hydrobiology 94:580–594. Smith, R.J., T. Kamiya, and D.J. Horne. 2006. Living males of the “ancient asexual” Darwinulidae (Ostracoda: Crustacea). Proceedings of the Royal Society of London Series B 273:1569 –1578. Stenderup, J.T., J. Olesen, and H. Glenner. 2006. Molecular phylogeny of the Branchiopoda (Crustacea)—multiple approaches suggest a “diplostracan” ancestry of the Notostraca. Molecular Phylogenetics and Evolution 41:182–194. Swanson, K.M. 1989. Manawa staceyi n.sp. (Punciidae, Ostracoda): Soft anatomy and ontogeny. Courier des Forschunginstitutes Senckenberg 113:235–249. Tattersall, W.M., and O. Tattersall. 1951. The British Mysidacea. Ray Society, London . Vannier, J., P. Boissy, and P. Racheboeuf. 1997. Locomotion in Nebalia bipes: A possible model for Palaeozoic phyllocarid crustaceans. Lethaia 30:89 –104. Walker, G. 2004 . Swimming speeds of the larval stages of the parasitic barnacle, Heterosaccus lunatus (Crustacea: Cirripedia: Rhizocephala). Journal of the Marine Biological Association of the United Kingdom 84:737–742. Walker, G., and R.J.G. Lester. 2000. The cypris larvae of the parasitic barnacle Heterosaccus lunatus (Crustacea, Cirripedia, Rhizocephala): Some laboratory observations. Journal of Experimental Marine Biology and Ecology 254:249 –257. Walker, G., A.B. Yule, and J.A. Nott. 1987. Structure and function in balanomorph larvae. Pages 307–327 in A.J. Southward, editor. Barnacle biology. Crustacean Issues, Vol. 5. Balkema, Rotterdam. Walossek , D. 1993. The Upper Cambrian Rehbachiella and the phylogeny of Branchiopoda and Crustacea. Fossils and Strata 32:1–202. Watling , L. 1999. Toward understanding the relationship of the peracaridan orders: The necessity of determining exact homologies. Pages 73–89 in F.R. Schram and J.C. von Vaupel Klein, editors. Crustaceans and the biodiversity crisis. Brill, Leiden .

The Crustacean Carapace Williamson, D.I. 1973. Amphionides reynaudii (H. Milne Edwards), representative of a proposed new order of Eucaridian Malacostraca. Crustaceana 25:35–50. Wirkner, C.S., and S. Richter. 2007. The circulatory system in Mysidacea—implications for the phylogenetic position of Lophogastrida and Mysida (Malacostraca, Crustacea). Journal of Morphology 268:311–328. Wirkner, C.S., and S. Richter. 2009. Evolutionary morphology of the circulatory system in Peracarida (Malacostraca; Crustacea). Cladistics 25:1–25. Zhang , X. -g., D.J. Siveter, D. Waloszek , and A. Maas. 2007. An epipode-bearing crown-group crustacean from the Lower Cambrian. Nature 449:595–598. Zimmer, C. 1932. Beobachtungen an lebenden Mysidaceen und Cumaceen. Sitzungsberichte der Gesellschaft Naturforschender Freunde zu Berlin 18:326 –347. Zimmer, C. 1941. Cumacea. Bronns Klassen und Ordnungen des Tierreichs, Band 5, Abteilung 1, Buch IV, Teil 5. Akademische Verlagsgesellschaft, Leipzig. Zimmer, C., and H.E. Gruner. 1956. Euphausiacea. Bronns Klassen und Ordnungen des Tierreichs, Band 5, Abteilung 1, Buch VI, Teil 3. Akademische Verlagsgesellschaft, Leipzig.

139

5 THE CRUSTACEAN INTEGUMENT: STRUCTURE AND FUNCTION

Richard M. Dillaman, Robert Roer, Thomas Shafer, and Shannon Modla

Abstract The dorsobranchial exoskeleton of decapods has served as the archetype for studies of the structure and formation of crustacean cuticle. This cuticle consists of four layers: the epi-, exo-, and endocuticles, which are mineralized with calcium carbonate, and the inner membranous layer. The inner three layers are formed from chitin-protein fibrils arranged in parallel lamellae that have a constantly changing orientation from layer to layer. This results in a plywoodlike composite, with the mineral aligning with the orientation of the fibrils. The exoskeleton is thus a composite structure with remarkable biomechanical resistance to fracture propagation. In different species and within different regions of the body of individuals, the cuticle shows many variations on this basic pattern. Both the numbers of layers and presence or degree of mineralization are highly variable. Examples of thin, pliable, uncalcified cuticles include those of the arthrodial, gill, and branchial chamber, contrasted by the heavily mineralized cuticle of the tips of the chelipeds, which may be reinforced with other minerals. The cuticle is cyclically shed and reformed to permit growth. The hypodermis first separates from the old cuticle and, in general, begins to secrete the components of the new epi- and exocuticle. When the animal emerges from the old exoskeleton, the endocuticle and membranous layer are deposited. Again, modifications of this scheme are seen in different cuticle types. The cyclical nature of cuticle formation and the temporal and spatial separation of the events of matrix deposition and calcification render the crustacean cuticle an excellent model for the study of the control of biomineralization.

INTRODUCTION The crustacean cuticle shares many structural and functional features with the cuticle of other arthropods, especially insects. One of the most obvious constraints of a rigid exoskeleton is that

140

Functional Morphology and Diversity. Edited by Les Watling and Martin Thiel. © 2013 Oxford University Press. Published 2013 by Oxford University Press.

The Crustacean Integument: Structure and Function it must be periodically shed so that the organism can increase in size. This means that cuticle deposition is not a singular event but, over a lifetime, is a succession of depositions interspersed with the shedding or loss of the old cuticle. Any review of the crustacean cuticle must recognize the outstanding work of early anatomists and physiologists who described the ultrastructure of the arthropod cuticle and who put it in both an anatomical and developmental context. Much of this early work was done on insects (Wigglesworth 1933 from Travis 1963) and served as the standard, or frame of reference, for investigations on the crustacean cuticle. For example, Neville (1975), in his volume Biology of the Arthropod Cuticle, describes a basic plan for the arthropod cuticle but one that recognizes that there is not a single type of cuticle. He therefore describes four different plans: tanned (sclerotized, cross-linked) solid cuticle, untanned solid cuticle, rubberlike cuticle, and arthrodial membrane. Revealing the plan for the crustacean cuticle is even more complex because major portions of the cuticle are calcified, thereby requiring a mechanism for deposition of the mineral as well as a means for selective removal of the mineral prior to shedding the old cuticle. The extracellular nature of the cuticle means that it accumulates over time. One consequence of this pattern is that any changes in the composition of the cuticle are a direct reflection of changes in the activity of the epithelium underlying the cuticle at a specific time. The changes in the layers can be either structural (e.g., a change in the composition or organization of the components of the cuticle) or functional (e.g., the ability or inability of the cuticle to mineralize). Furthermore, in the latter case function may not be expressed immediately upon deposition but may be expressed at a later time (e.g., the ability to calcify may be determined early in premolt but not expressed until postmolt). This type of expression pattern makes it particularly important that one be able to identify candidate proteins and glycoproteins that are associated with various aspects of cuticle deposition and also to determine the pattern of gene expression for those proteins and glycoproteins. Another factor that must be considered when characterizing cuticle formation is that different regions of the cuticle vary considerably in structure and timing of their deposition. Structural differences, of course, lead to functional differences. For example, the arthrodial membranes have a flexible, noncalcified cuticle, thereby allowing movement of the appendages, but they are deposited at the same time as the calcified dorsal carapace (Williams et al. 2003). Likewise, differences in timing have functional consequences. For example, the thin, uncalcified cuticle of the gills, which presents a minimal barrier to diffusion, is not replaced until all of the other regions of the cuticle have been synthesized (Andrews and Dillaman 1993).

STRUCTURE AND COMPOSITION OF THE CUTICLE The dorsal carapace of decapod brachyuran crustaceans (e.g., Carcinus maenas, Callinectes sapidus, Scylla serrata) has been the most widely studied and characterized region of the exoskeleton. It will thus serve as the prototype for the following description of the structure and composition of the cuticle and for comparison with other taxonomic groups and regions of the skeleton. An understanding of the function, development, and dynamics of the dorsal carapace requires knowledge of its structure when fully elaborated, at the intermolt period. While current descriptions of the crustacean exoskeleton have been refined since our previous reviews (Roer and Dillaman 1984, 1993), the basic organization remains remarkably similar to that portrayed in early studies of the insect cuticle by such icons as Locke (1959, 1960, 1961) and Neville (1975). In common, current terminology, the four layers of the crustacean cuticle are (from distal to medial) the epicuticle, exocuticle, endocuticle, and membranous layer. The organization and structure of these layers are shown in Figs. 5.1 and 5.2. The epicuticle itself is approximately 5 μm in thickness (Roer and Dillaman 1984) and comprises three layers: the outer surface coat, a cuticulin layer that has five sublayers, and a thick inner epicuticle that contains amorphous material

141

142

Functional Morphology and Diversity A

B

G

C

epicuticle exocuticle

epicuticle exocuticle

endocuticle

membranous layer hypodermis

endocuticle

F

E D

membranous layer hypodermis

Fig. 5.1. (A–F) Schematic of the molt cycle in Callinectes sapidus showing the layers present at intermolt (A; stage C4), apolysis (B; stage D0), late premolt (C; stage D3–D4), early postmolt (D; stage A 1), middle postmolt (E; stage B), and late postmolt (F; stage C3). (G) Section through an intermolt cuticle stained with acridine orange.

and fibers (~6 nm in diameter) perpendicular to the surface (Compére 1995). Intercalated within this set of fibers are others that extend down into the outer margin of the exocuticle below (Modla 2006; Fig. 5.3A). The exo- and endocuticles comprise a highly organized, organic framework of chitin-protein fibrils and, in many regions, are hardened by sclerotization and/or impregnation with mineral salts. The basis for the laminate structure of these fibrils has been a matter of study and debate for many years (Bouligand 1972). Recent reports employing more sophisticated analytical techniques have revealed a continuum of organization from the molecular to the tissue levels (Raabe et al. 2005a, 2005b, 2007, Fabritius et al. 2009). These analyses explain not only the underlying structure but also many of the composite biomechanical properties of the cuticle. At the molecular level, the chitin fibrils are composed of E -1,4-linked N-acetylglucosamine residues arranged in antiparallel chains of D-chitin. Groups of 18–25 chitin fibrils are wrapped by proteins to form nanofibrils that are 2–5 nm in diameter and approximately 300 nm in length (Raabe et al. 2005a, 2005b, 2007, Sachs et al. 2006, Romano et al. 2007, Fabritius et al. 2009). The nanofibrils, in turn, cluster to form chitin-protein fibers that are 50–250 nm in diameter. As the fibers aggregate into flat sheets parallel to the apical surfaces of the epithelial cells, they are arranged around the microvilli forming pore canals. This gives the chitin-protein sheets a fenestrated appearance when viewed in tangential planes. The chitin-protein sheets are stacked, one upon the next, with a slight rotation in the axis of the fiber orientation relative to the previous layer (Fig. 5.2A). This imparts a plywoodlike structure to the cuticle, originally described for the crustacean exoskeleton by Bouligand (1972). Each lamella in the exo- or endocuticle represents a 180° rotation in the orientation of the chitin-protein fibers within the sheets. The thickness of a lamella is greater in the endocuticle (~8 μm) than in the exocuticle (~2 μm; see Fig. 5.2), which is thought to be due to the relative angle between adjacent sheets of fibers, that is, smaller angles of rotation in the endocuticle resulting in thicker lamellae since more layers are required to effect a

The Crustacean Integument: Structure and Function

Fig. 5.2. Microstructure of lobster cuticle. (A) Schematic representation of the different hierarchical levels in the microstructure of lobster cuticle starting with the N-acetyl-glucosamine molecules (I) forming antiparallel D-chitin chains (II). Between 18 and 25 of these molecules wrapped with proteins form nanofibrils (III), which cluster to form chitin protein fibers (IV) that are arranged in horizontal planes in which the long axes of the fibers are all oriented in the same direction. The fibers are arranged around the cavities originating from the extremely well-developed pore canal system that gives the structure a honeycomb-like appearance (V). These chitin protein planes are stacked with the orientation of the fibers in superimposed layers rotating gradually around the normal axis of the cuticle, thus creating a typical twisted plywood structure (VI). (B) Scanning electron microscropic (SEM) micrograph showing a cross section through the three-layered cuticle. The different stacking density of the twisted plywood layers (tp) in the exoand endocuticle can be clearly seen. (C) SEM micrograph of obliquely fractured endocuticle displaying two superimposed twisted plywood layers (tp) and showing their typical honeycomb-like structure. The arrows indicate the pore canals. From Romano et al. (2007, fig. 1), with permission from Elsevier.

180° rotation. Presumably, the stacking angles are determined by the proteins that wrap the chitin fibrils. In fact, differences in the exo- and endocuticular proteins have been well documented (Skinner et al. 1992), as have differences in the sugar residues associated with cuticular glycoproteins (Marlowe et al. 1994; Compére et al. 2002). These differences also account, at least in part, for the differences in tanning or sclerotization that occur between these layers. The exocuticle is stabilized and hardened by quinone cross-linking effected by phenoloxidase, whereas such tanning does not occur in the endocuticle. The parallel, planar orientation of the chitin-protein lamellae and associated mineral is reflected in the predominant chitin and mineral crystallographic axes as revealed by

143

144

Functional Morphology and Diversity

Fig. 5.3. Transmission electron microscopic images of 1-hour postmolt cuticle of Callinectes sapidus fixed with 2.5% glutaraldehyde. (A) Cross section through epicuticle. Note the dense vertical fibers (dvf), epicuticular fibers (ef), epicuticular canals (ec), epicuticular roots (er), and inner (ie) and outer (oe) epicuticle. (B) Cross section through single lamella of exocuticle. Note the vertical fibers (vf) and horizontal fibers (arrowheads). (C) Tangential section through the proximal exocuticle. Note the anchoring fibers (af), horizontal fibers (hf), pore canals (pc), pore canal fibers (pcf), and pore canal sheaths (pcs). (D) Cross section through proximal exocuticle and tendinous epidermal cell. Note the insertion of the tonofibers (tf) into the cell as well as the dense structures (d) and microtubules (mt) in the cell. (E) Tangential section through the proximal exocuticle above a tendinous epidermal cell. Note the electron-dense rods (r) and tonofibers (tf). (F) Tangential section through the exocuticle. Note the region of the interprismatic septa (IPS). From Modla (2006, figs. 1a, 2d, 5e, 6d, 7d, and 9b), used with permission from the author.

synchrotron Bragg diffraction (Raabe et al. 2005a, 2007) and X-ray diffraction of cuticular samples (Raabe et al. 2006, 2007, Krywka et al. 2007, Al-Sawalmih et al. 2008). When the diffraction patterns are projected upon the {020} crystallographic plane, which runs normal to the surface in the transverse direction (90° to the long axis of the body of the lobster Homarus

The Crustacean Integument: Structure and Function

Fig. 5.4. Survey of the {020} synchrotron pole figures of the orthorhombic D-chitin taken from different parts of the cuticle. LD, longitudinal reference direction; ND, reference direction normal to the local surface; TD, transverse direction. Specimens were taken from a highly mineralized part of the cuticle on pincher claw (left cheliped), crusher claw (right cheliped), cephalothorax, and abdomen. Specimens were also taken from poorly mineralized positions on the telson or abdomen (dashed line). From Raabe et al. (2007, fig. 5), with permission from Elsevier.

americanus), pole figures (as shown in Fig. 5.4) demonstrate a strong orientation along the long axis of the body (Raabe et al. 2007). A pole figure is a crystal orientation measurement and is so named because it is often plotted in polar coordinates consisting of the tilt and rotation angles with respect to a given crystallographic orientation. These data are interpreted to reflect the orientation of the chitin-protein molecules within the lamellae that are parallel to the cuticle surface. The pole figures, however, also reveal a secondary crystallographic axis normal to the surface of the cuticle. It was concluded that this represents chitin and mineral associated with the vertical pore canals perpendicular to the lamellar elements. As discussed below, numerous vertical elements are found in the exoskeleton of Callinectes that could also contribute to the observed orientation in the normal dimension.

145

146

Functional Morphology and Diversity Investigations in our laboratory on the ultrastructure of late premolt and early postmolt blue crabs have demonstrated that numerous fibrous structures run parallel to the twisted-ribbon shaped pore canals and therefore perpendicular to the horizontal chitin protein fibers forming the Bouligand or twisted plywood layers. Modla (2006) used transmission electron microscopy after conventional glutaraldehyde fixation or uranyl acetate fixation and described the morphology and distribution of a variety of fiber types. The epicuticle consists of a thin, three- to fivelayer outer epicuticle and a thicker inner epicuticle that is composed of epicuticular roots (er, Fig. 5.3A) separated from one another by epicuticular canals (ec, Fig. 5.3A). The canals often contain dense epicuticular fibers (ef, Fig. 5.3A). The predominant fibers in the exocuticle are the chitin-protein horizontal fibers (hf, Fig. 5.3C), which run parallel to the cuticle surface and rotate in successive planes. All other fiber types are oriented perpendicular to the cuticle surface. Within the exocuticle, dense vertical fibers (dvf, Fig. 5.3A) extend from the epicuticle to distal regions of the exocuticle. Vertical fibers (vf, Fig. 5.3B) are present in the distal and medial exocuticle and are most likely contiguous with pore canal sheaths (pcs, Fig. 5.3C), which are fibers associated with pore canal membranes in the medial and proximal exocuticle. Anchoring fibers (af, Fig. 5.3C) are located in the medial and proximal exocuticle and traverse the cuticle by intersecting bundles of horizontal fibers. Pore canal fibers (pcf, Fig. 5.3C) are proximally distributed and are associated with both the outer pore canal membrane and microtubules in the hypodermis. Tonofibers (tf, Fig. 5.3D,E) and electron-dense rods (r, Fig. 5.3E) are specialized fibers occurring in regions of muscle attachment. Uranyl acetate fixation differed from conventional fixation in that it did not preserve cellular components, but it greatly increased the contrast of all fiber types. Uranyl acetate fixation also permitted the visualization of the calcification initiation sites along the epicuticle-exocuticle interface and interprismatic septa (IPS, Fig. 5.3F). The interprismatic septa delineate an array of roughly hexagonal columns in the exocuticle that have similar proportions and orientation as the lateral margins of the underlying hypodermal cells that elaborate the cuticle (Giraud-Guille 1984, Compére 1995). Fig. 5.5 summarizes the distribution of the vertical and horizontal fibers in the epi- and exocuticle. All three outer layers (epi-, exo- and endocuticles) are mineralized. The epicuticle is partially calcified, with crystals nucleated at the epi-/exocuticular margin growing up between the vertical fibers of the inner epicuticle (Hegdahl et al. 1977, Compére 1995, Dillaman et al. 2005). Electron-dense, amorphous deposits are sometimes seen to be associated with the surface layer, but it is unclear whether they are synthesized by the crustacean or are a matrix associated with microorganisms on the outer surface (Read and Williams 1991). Terminations of the pore canals are also seen on the surface of the epicuticle in some taxa (Halcrow and Bousfield 1987). Within the exo- and endocuticles, mineral is generally in the form of fused spherulites that align with the chitin-protein fibers (Roer and Dillaman 1984, Romano et al. 2007) (Fig. 5.6, inset). Originally, the form of the mineral was thought to be calcium carbonate in the form of calcite crystals. While substantial amounts of crystalline magnesian calcite have been confirmed in the carapace of lobsters and crabs (Boselmann et al. 2007), it is now clear that more than one calcium salt may be involved (depending upon the location and the species), as well as noncrystalline, amorphous forms of mineral. Pratoomchat et al. (2002) demonstrated the presence of calcium phosphate (as dicalcium phosphate dehydrate and octacalcium phosphate) as a precursor to the formation of calcium carbonate during postmolt mineralization in the carapace of the crab Scylla. Soejoko and Tjia (2003) observed that calcium phosphate minerals persist throughout the postmolt and intermolt stages in the carapace of the giant prawn Macrobrachium rosenbergii. The calcium phosphate is present both in crystalline and amorphous forms and coexists with calcium carbonate in nearly equal proportions. In the terrestrial isopods Porcellio scaber and Armadillidium vulgare, Becker et al. (2005) found the cuticle to contain crystalline

The Crustacean Integument: Structure and Function epi dvf

dvf

vf pc

pc

vf vf pcs

pcs pcs

H A

B

af

af

pcf pcf C

D

Fig. 5.5. Diagrams illustrating the various vertical fiber types and features within the cuticle as well as their distribution in tangential sections through the distal, medial, and proximal exocuticle. Each hexagonal prism represents the cuticle overlying a single hypodermal cell. (A) Diagram showing the pore canals (pc) extending from hypodermal cells (H) and dense vertical fibers (dvf) extending down from the epicuticle (epi). (B) The distribution of vertical fibers (vf) and the pore canal sheath (pcs). Note that vertical fibers are distal extensions of the more proximally located pore canal sheaths. (C) The distribution of anchoring fibers (af). (D) The distribution of pore canal fibers (pcf). Adapted from Modla (2006, fig. 27d), used with permission from the author.

magnesium calcite, amorphous calcium carbonate (ACC), and amorphous calcium phosphate. These analyses were performed by X-ray diffraction and Fourier-transform infrared spectroscopy (FTIR) of bulk specimens, so no precise distribution could be assigned to each of the minerals and their morphs. However, the application of confocal-Raman spectroscopic imaging has permitted mapping of the different minerals and their forms to the various cuticular layers. In the isopods Porcellio and Armadillidium, Hild et al. (2008) clearly demonstrated that

Fig. 5.6. Microstructure of purified chitin heat-treated at 220°C: detail images of the fibers near a pore canal. Adjacent fibers seem to be connected by small fibrillar structures (circled areas, arrows in inset). From Romano et al. (2007, fig. 7), with permission from Elsevier.

147

Functional Morphology and Diversity Epi

1

2

3

4

5

6

A

B

C

Peak area/counts a.u.

148

100 80 60 40 20 0 0

5

10

15

25 20 Distance (μm)

30

35

40

c a b 45

Fig. 5.7. Raman spectroscopic images (A–C) and line scans (D) recorded from a sagittally cleaved and microtome polished surface of the mineralized tergite cuticles (shown with the epicuticle to the left) of the isopod Armadillidium vulgare show the local distribution of the various components. Calcium carbonate (A) occurs within the whole exo- and endocuticle, whereas calcite (B) is located within the exocuticle only. Pore canals (arrowheads) appear devoid of mineral. The amount of organic material (C) increases from the distal to the proximal region of the cuticle. The membranous layer (position 6) is devoid of calcium carbonate. Horizontal dotted lines in the Raman images (A–C) indicate the position where the line scans (Epi, 1–6) were recorded for carbonate (a), calcite (b), and organic material (c) to determine material distribution. From Hild et al. (2008, fig. 5), with permission from Elsevier.

calcite was entirely restricted to the exocuticle, while the mineral in the endocuticle was uniformly ACC (Fig. 5.7). The key characteristics that the dorsal carapace should exhibit in order to protect the organism from predation are hardness and fracture resistance. The helicoidally arranged lamellae of chitin-protein fibers and associated minerals constitute a composite material with exceptional biomechanical properties in this regard. The force required to puncture the dorsal carapace of lobsters (measured with a punch test) is dependent upon the thickness of the cuticle and ranges from 6.7 kg in Homarus to 27.8 kg in the slipper lobster Scyllarides latus (Tarsitano et al. 2006). As would be expected, the cuticle is far more resistant to compressive forces than to tensile forces. In a compressive test in the normal dimension (i.e., perpendicular to the surface), the walking leg cuticle of the sheep crab Loxorhynchus grandis exhibited a stress to fracture of 101 ± 11 MPa, a value comparable to aluminum and in excess of limestone (60 MPa) (Chen et al. 2008). In contrast, the cuticle failed in a tensile test in the normal dimension at only 9.8 ± 2.6 MPa. Interestingly, the cuticle was substantially more resistant to tensile stresses in the longitudinal dimension (shear), with a stress to fracture of 31.5 ± 5.4 MPa (Chen et al. 2008).

The Crustacean Integument: Structure and Function

HYPODERMIS AND THE MOLT CYCLE The cuticle is underlain and periodically elaborated by the hypodermis in a cycle of deposition and resorption referred to as the molt cycle (Drach 1939, Travis 1955, 1957, 1965, Skinner 1962, Drach and Tchernigovtzeff 1967). These authors and others (as reviewed in Roer and Dillaman 1993) have attempted to standardize the terminology and the sequence of events of the molt cycle by focusing on dynamics of the dorsal carapace. Recent observations have highlighted, however, that this scheme may need to be modified for other tissues, as is detailed below. When the cuticle of the dorsal carapace is fully elaborated, the animal is referred to as being in intermolt (or stage C4). During this stage, the hypodermis is a simple epithelium that tends to be squamous. While growth of the internal organs of a crustacean can occur during intermolt, an increase in the external dimensions requires the shedding of the existing rigid exoskeleton. A schematic representation of the molt cycle and associated cuticular changes is presented in Fig. 5.1. The first step in this process (referred to as early premolt or stage D0) is the separation of the hypodermis from the cuticle, termed apolysis (Compére et al. 1998). The hypodermis is more active during this period and becomes cuboidal and then columnar. The “old” cuticle is partially degraded, and its components are resorbed during premolt stages D1 and D2 (Roer 1980, Compére et al. 1998). Before the cuticle can enlarge, there must be an increase in the surface area of the epithelium that secretes it and to which it is anchored by the pore canals. Stage D1 is therefore the period of increased hypodermal mitotic activity that will result in the formation of a larger new cuticle (Skinner 1965). The outer two layers of the new cuticle (the epi- and exocuticles) begin to be deposited beneath the old cuticle during late stage D1 and stage D2. These are often referred to as the preexuvial layers, and the organic components of these layers are thought to be completely formed prior to the shedding of the old cuticle (termed ecdysis). It is important to note (as discussed below) that mineralization of these layers cannot occur until the animal emerges from the old exoskeleton and is fully expanded. During this period of active cuticle synthesis, the individual columnar epithelial cells assume a packing array that results in roughly hexagonal margins (Fig. 5.8A) and leads to the formation of the interprismatic septa of the exocuticle. Furthermore, there are surface modifications of the epithelial cells (Fig. 5.8A). The most obvious are cytoplasmic extensions resembling microvilli that extend from the apical cell surface throughout the extracellular cuticular layers, referred to as pore canals. Compére et al. (1998) have extensively documented the formation, extent, and fate of the pore canals throughout the entire molt cycle. Another modification is the array of short microvilli that constitute the “plaques” (Fig. 5.8B). These were first described in insects (Locke 1961) and appear to be the sites of the initial polymerization and organization of the developing cuticular components (Compére 1995, Greenaway et al. 1995, Elliott and Dillaman 1999). The plaques may provide the template for the orientation of the chitin-protein fibers that permits them to interact with previously deposited layers. The fibers are then thought to selfassemble in such a way that their plane of orientation is offset by a fixed angle relative to the previous layer, thereby forming the Bouligand pattern of lamellae. Once the organic lamellae of the preexuvial cuticle are fully formed, the next event is the emergence of the crustacean from the old exoskeleton (exuviae). This extraordinary process of ecdysis is initiated by the uptake of water. The resultant hydrostatic pressure causes the exuviae to rupture at predetermined sites referred to as ecdysial sutures. Prior to the onset of postecdysial tanning and mineralization, the mobility of the newly molted crab is possible because of a hydrostatic support system. Taylor and coworkers (Taylor and Kier 2003, Taylor et al. 2007) have demonstrated the transition from a hydrostatic skeleton to a rigid skeleton in the very early postmolt period.

149

150

Functional Morphology and Diversity

Fig. 5.8. (A) Scanning electron micrograph of early postmolt cuticle from Callinectes sapidus. Note the fracture along the interface between the hypodermis (h) and the exocuticle (exo) revealing the roughly hexagonal margins of the individual hypodermal cells (arrows). In the inset, note the pore canals extending from the surface of the hypodermal cells (arrowheads). (B) Transmission electronic micrograph of dorsal carapace from premolt cuticle of Callinectes sapidus: high magnification of cross section through the proximal exocuticle and hypodermis. Note the assembly zone (AZ) and transition zone (TZ) above the microvilli (mv) of a tendinous epidermal cell and the electron-dense apical plaques at the microvilli (arrowheads) and abundant microtubules (mt) in the cytoplasm (from Modla 2006, fig. 8b, used with permission from the author). (C) Backscattered electron micrograph of 24-hour postmolt cuticle from Callinectes sapidus. Mineralized areas appear white. Note the calcified epicuticle (arrows) and interprismatic septa of the exocuticle (arrowheads) as well as the more fully calcified endocuticle (endo) (from Dillaman et al. 2005, fig. 4d2, used with permission).

The onset of cuticular hardening may be brought on by the release of the hormone bursicon, which has been characterized as the agent responsible for the tanning of postmolt insect cuticle. Bursicon transcripts have recently been found in the green crab Carcinus maenas (Wilcockson and Webster 2008). While the precise mechanisms by which tanning and mineralization are prevented during premolt and initiated during postmolt are not fully understood (and are addressed in more detail below), the patterns of postecdysial mineralization are apparent. Within 3 hours of ecdysis (stages A 1 and A 2) in the blue crab Callinectes sapidus, evidence of calcium carbonate precipitation can be seen at the epicuticle/exocuticle boundary (Dillaman et al. 2005). Fronts of mineralization then extend along the interprismatic septa, both distally and proximally, until they meet in the center of the exocuticle, forming calcified margins delineating the hexagonal prisms (Fig. 5.8C). This general pattern of calcification has been described in other species (Bouligand 1972, Giraud 1977, Giraud-Guille and Quintana 1982, Giraud-Guille 1984, Sakamoto et al. 2009). In addition to a defined sequence characterizing the pattern of mineralization, the form of the mineral also follows a prescribed sequence. Initially, the mineral at the epi-/exocuticular boundary is presumed to be ACC based on its

The Crustacean Integument: Structure and Function solubility. Subsequently, the ACC is transformed into crystalline calcium carbonate in the form of calcite as ACC is deposited along the interprismatic septa. This mineral also transitions to calcite, completely encasing the prisms in crystalline material. The prisms themselves then begin to calcify from the distal toward the proximal portion of the exocuticle; there is some question as to the final mineral morph in these regions. At the same time that the preexuvial layers are hardening (stages A 2, B1, and B2), the endocuticle organic matrix is being deposited and, within a short time, calcified. The mineral morph deposited in the blue crab has not been characterized, but that within the endocuticle of isopods (Hild et al. 2008) and the lobster (Al-Sawalmih et al. 2008) is in the form of ACC. Another component of the mineralized cuticle appears to be calcium phosphate, based on elemental analysis (Pratoomchat et al. 2002, Soejoko and Tjia 2003), but X-ray diffraction and FTIR spectroscopy fail to detect apatite, suggesting that this mineral is amorphous as well. The deposition and mineralization of the endocuticle continues through postmolt stages (C1–C3). The transition to intermolt is defined by the completion of the deposition of a noncalcified membranous layer.

VARIATIONS IN CUTICLE T YPE AND STRUCTURE Arthrodial Cuticle Much of our knowledge regarding the control of tanning and mineralization is based upon observed differences in the composition and the degree of mineralization in different cuticle types from different regions of the body. Perhaps the most striking difference is between cuticles that mineralize and those that do not. The arthrodial cuticle is found, as the name implies, in the joints of crustacean appendages and must remain flexible to allow for locomotion. The basic structure of arthrodial cuticle and the timing of its deposition are similar to those of the dorsal carapace and the mineralized cuticle with which it is contiguous. It has an epicuticle that is very similar to that of the dorsal carapace and a Bouligand pattern of chitin-protein fibers that extend from the epicuticle to the hypodermis. While the arthrodial cuticle of the blue crab has pore canals (the cuticle-encased cytoplasmic extensions of the epithelial cells; Fig. 5.9A,B), some authors have stated that in other species arthrodial cuticle lacks such structures (Raabe et al. 2007). The lamellar portion of the arthrodial cuticle appears homogeneous, but staining of the carbohydrate moieties of the cuticular glycoproteins clearly reveals that this lamellar portion can be divided into two layers (Fig. 5.9C). These two layers correspond spatially to the adjacent exo- and endocuticle of the calcified cuticle, indicating that they are deposited at the same time. Indeed, the change in thickness of the arthrodial cuticle mirrors that of the calcified cuticle in both pre- and postmolt (Williams et al. 2003). Interestingly, the boundary between the two cuticle types is not perpendicular to the cuticular surface; rather, the boundary is oblique to the surface (Fig. 5.9D). This implies that a given patch of hypodermis must begin synthesizing arthrodial cuticle and change to synthesizing calcified cuticle during the pre- and postmolt depositional periods. The advantage of this phenomenon is that the interface between the calcified and noncalcified regions has much greater surface area, reinforcing the junction between the two. Branchial Chamber Cuticle The cuticle that lines the branchial chamber of the crab is also noncalcified. In contrast to the arthrodial cuticle, this structure is only about 6.5 μm thick in the blue crab. It is lamellar

151

152

Functional Morphology and Diversity

Fig. 5.9. (A) Transmission electronic micrograph of arthrodial membrane of stage D4 Callinectes sapidus: cuticle showing the numerous lamellae and cytoplasmic extensions (arrows) roughly perpendicular to the surface of the cuticle. (B) Note the cytoplasmic extensions (arrows) extending into the newly deposited arthrodial membrane and the abundant microtubules (arrowheads) in the cytoplasm of the hypodermal cells. (C) Intermolt, stage C 4, arthrodial membrane of Callinectes sapidus stained with periodic acid-Schiff and hematoxylin. Note the boundary between the preexuvial and postexuvial cuticle (arrowhead). (D) Section of intermolt, stage C 4, cuticle of Callinectes sapidus at the boundary between the calcified cuticle and arthrodial membrane (a) and stained with hematoxylin and eosin. Note the diagonal margin for both the exocuticle (exo) and endocuticle (endo). From Williams et al. (2003, figs. 2c, 2d, 5f, and 5c), used with permission.

but possesses only approximately 30 lamellae. A feature of such thin, noncalcified cuticles in crustaceans is a modification of the timing of deposition during the molt cycle. The branchial chamber cuticle is secreted in its entirety during premolt, being completed by the late premolt stage, D3 (Elliott and Dillaman 1999). Despite the fact that this cuticle is completely elaborated at the time of the molt, it is still subject to extreme tensile forces during extraction from the old cuticle at ecdysis. To resist these tensile forces, the hypodermal cells underlying the branchial chamber cuticle, which were cuboidal to columnar during their secretory phase, become filled with microtubules along their long axis and assume the structure and function of tendon cells, anchoring the new cuticle to the connective tissue below. These microtubule-filled cells persist through ecdysis but during early postmolt revert to the intermolt/early premolt morphology. Gill Cuticle The cuticle that lines the outer surface of the gills not only protects these organs from damage but also represents a potential barrier to the free diffusion of respiratory gases and the transport of ions across the gill surfaces. Consequently, the gill cuticle is extremely thin. In

The Crustacean Integument: Structure and Function the crayfish Procambarus clarkii , the cuticle surrounding the gill filaments is between 0.3 and 2 μm thick. Like the dorsal carapace, the gill cuticle usually consists of an epi-, exo-, and endocuticle, albeit much thinner. In the crayfish, we observed differences among the cuticles of the transporting filaments and the afferent and efferent channels of the respiratory filaments (Dickson et al. 1991). The cuticle of the transporting filaments differs from the dorsal carapace model not only in lack of mineralization and dimension (~2 μm thick) but also in the relative proportions of the exo- and endocuticles. Whereas the endocuticle is the predominant layer in the carapace, the composition of the cuticle of the transporting filament is approximately two-thirds exocuticle. The cuticle of the afferent channel of the respiratory filaments is approximately the same thickness as the transport filament cuticle and is also composed of lamellate exo- and endocuticles, but the proportions are more similar to the dorsal carapace, with the endocuticle representing more than two-thirds of the thickness. The cuticle of the efferent channel of the respiratory filaments is only 0.3–0.5 μm thick. It appears to possess an outer epicuticle, but the cuticle below does not appear to be lamellate and has no obvious delineation into an exo- and endocuticle. Like the branchial chamber cuticle, the gill cuticle is fully elaborated during premolt, with no postmolt deposition (Andrews and Dillaman 1993). In fact, the formation of the respiratory cuticle is delayed until late stage D2 to early D3, and deposition in the transporting gills is delayed until early in the short, final premolt stage, D4. The net effect of these delays is that transport of ions and diffusion of respiratory gases are subjected to a double barrier (old and new cuticular layers) for only a very brief period of time. This principle may apply to the branchial chamber cuticle as well, since it has been well documented that gas exchange may occur across this surface, particularly in the case of semiterrestrial and terrestrial crabs (Greenaway and Farrelly 1984, 1990, Taylor and Greenaway 1984). Cheliped Cuticle Whereas the dorsal carapace is adapted for resistance to compression and fracture propagation to protect the organism from predation, the claws must resist wear and be hard enough to allow crushing of other mineralized structures. As mentioned in the discussion of biomechanics above, one strategy for increasing hardness is simply increasing the thickness of the cuticle (Tarsitano et al. 2006). Indeed, the cuticular thickness, even within the calcified cuticle, can vary markedly from one region to another. Additionally, the relative proportions of exo- and endocuticle is variable (Fig. 5.10). However, additional structural modifications of the cheliped cuticle provide wear resistance and impart hardness. For example, the propodus of the cheliped is some four times harder than that of the pereopods, within both the endocuticle (471 ± 50 vs. 142 ± 17 MPa) and exocuticle (947 ± 74 vs. 247 ± 19 MPa), as revealed by microindentation tests on the sheep crab Loxorhynchus (Chen et al. 2008). Local differences in hardness have been described within the claw of the lobster Homarus, the exocuticle of the lateral surface of the claw being ~325 MPa and the tooth of the claw being 590 MPa. Similar differences were noted in the stiffness (Young’s modulus) in these regions. The endocuticle of the lateral surface exhibited a value of 6.9 GPa, compared to 11 GPa for the exocuticle (roughly equivalent to bone). The endocuticle of the tooth had a stiffness value of 3.7 (comparable to polystyrene), while the tooth exocuticle was measured at 25.7 (close to that for concrete) (Chen et al. 2008). Melnick et al. (1996) compared the mechanical properties of the pigmented versus white regions of the cheliped of the stone crab Menippe mercenaria and found that the black areas had a higher density, elastic modulus, hardness, and fracture toughness compared to light areas. These differences were attributed to a decreased porosity in the dark areas and perhaps a greater degree of tanning. However, hardness may not be the desired characteristic in all chelipeds. Some crabs employ the claws for rasping and grasping, rather than crushing, and these differences in

153

154

Functional Morphology and Diversity

Fig. 5.10. Scanning electron micrographs of transversally fractured lobster cuticle from different body parts, exposing the cross sections. (A) Cuticle from the claws; the exocuticle (exo) is thin in relation to the massive endocuticle (endo). The detail image shows fibers oriented perpendicular to the fiber planes forming the twisted plywood layers in the pore canals (pc) and lining them (arrows). (B) Cuticle from the carapace; exo- and endocuticle have nearly the same thickness. In the detail image, fibers oriented in the normal direction (arrows) are also present in the pore canals (pc). (C) Cuticle from the tergites; the exocuticle is about twice as thick as the endocuticle. At a higher magnification, fibers oriented in the normal direction (arrows) are again visible in the pore canals (pc). (D) Cuticle from the uropods; the exocuticle is very thick in relation to the very thin endocuticle. Fibers oriented in the normal direction (arrows) are present in the pore canals (pc), too, but their volume fraction is smaller than in the other mineralized parts of the lobster. (E and F) Cross section of cuticle from the joint membranes (E), with detail image (F). No pore canals are present in these unmineralized parts of the lobster; fibers oriented in directions other than in plane with the cuticle surface cannot be observed. From Raabe et al. (2007, fig. 10), used with permission.

function are reflected in the structure of the cuticle. Cribb et al. (2009) studied the cheliped tips in the grapsid crab Metopograpsus frontalis and found that this region was poorly mineralized but contained high levels of halogens. The outer exocuticle was enriched with chlorine, while the inner exocuticle had elevated bromine. The inner endocuticle also contained abundant chlorine (Fig. 5.11). The cuticle tips were less hard and less stiff than the carapace but had values equivalent to those found for insect cuticle lacking metals. It was hypothesized that the high levels of

The Crustacean Integument: Structure and Function

Fig. 5.11. Scanning electron microscopic images of the cheliped tip of the crab, Metopograpsus frontalis. (A) Backscattered electron image of M. frontalis cheliped tip: darker contrast indicates lower average atomic number. Outer surface is to the left. (B–D) X-ray intensity maps for chlorine, bromine, and chlorine plus bromine, respectively, across a cheliped tip from an area similar to that shown in A. X-ray intensity maps use a thermal color scale; here, gray indicates the higher X-ray intensity for an element, white lower, and black the lowest. Scale bars, 30 μm. From Cribb et al. (2009, fig. 1f,h,i,j) with permission from Elsevier.

halogens imparted an increased hardness and low elastic modulus to the chitin-protein matrix. The cheliped tips of Metopograpsus also have an unusual structural feature. Horizontal, rodlike structures were observed in the fractured surfaces of the exocuticle (Fig. 5.12). The obvious analogy is that they resemble reinforcing rods in concrete. Ecdysial Suture Variations in cuticular structure may include local modifications of the cuticle. Such is the case with the ecdysial suture, which is the narrow region of the ventral branchial carapace in crabs that ruptures during ecdysis (Priester et al. 2005). Modifications include a significant decrease in thickness, such that the location of the suture is visible to the naked eye in the intermolt crab. Others include a decrease in mineral density and alteration in matrix composition. The decrease in mineral density in the cuticle is achieved in a number of ways. First, the exocuticle mineralizes only along the epicuticle/exocuticle boundary and the interprismatic septa. The prisms of the exocuticle never fill with mineral. Second, the wedge of endocuticle immediately beneath the less mineralized exocuticle has a demonstrably lower concentration of mineral than the adjacent endocuticle (Fig. 5.13A). Third, while it is uncertain if the mineralogy of the suture differs from the surrounding cuticle, elemental analysis conducted on the exo- and endocuticle components showed that the suture contained significantly less magnesium (0.86 ± 0.29 vs. 1.75 ± 0.24 for

155

156

Functional Morphology and Diversity

Fig. 5.12. Scanning electron micrographs of noncalcified cheliped and leg tips from Metopograpsus frontalis, in transverse section. (A) Fractured surface across a cheliped tip showing multiple layers (2–5). Outer surface is to the left. Scale bar, 20 mm. (B) Fractured surface of a cheliped tip showing rods (r) filling and protruding from holes in the layer 3 region. Scale bar, 20 mm. From Cribb et al. (2009, fig. 2a,c), with permission from Elsevier.

the exocuticle; 0.81 ± 0.30 vs. 1.31 ± 0.14 for the endocuticle). High magnesian calcite has a lower solubility, so it is possible that this difference renders the mineral of the suture more soluble and labile. Assuming that the composition and location of the mineral are affected by the associated organic matrix, one would expect to be able to detect such differences between the suture and the adjacent cuticle. In fact, lectin binding studies revealed such differences (Fig. 5.13B). The net effect of these alterations is that during premolt cuticle resorption, the suture region is preferentially removed and weakened (Fig. 5.13C–F) so that the increase in hydrostatic pressure in stages D4 and E is sufficient to allow the old carapace to break open along the line of the suture. The resorption is more intense at the posterior region of the suture line, thus allowing the carapace to open from the posterior margin, making it possible for the crab to back out of the exuviae.

TEMPORAL VARIATIONS IN CUTICULAR SYNTHESIS The timeline for deposition of some of the aforementioned types of cuticle is summarized in Fig. 5.14. Considering the wide diversity in the timing of deposition among these few cuticle types that have been examined, additional variations in depositional patterns will likely be found as other regions and types are studied. The temporal disparities in the onset and completion of cuticle synthesis among tissues within a single organism certainly pose interesting questions regarding the control of the cellular events involved in the molting process. The complexity represented by these differences speaks against a regulatory system based upon universal changes in the titers of a small number of hormones. Nowhere is this more evident than in the biphasic molting of isopod crustaceans, in which the mineral of the posterior cuticle is mobilized and stored as ACC in anterior sternal deposits, followed by the molting of the posterior cuticle, remobilization of the mineral for calcification of the posterior cuticle, and finally the molting of the anterior cuticle (Steel 1980, 1982, Ziegler et al. 2005, 2007). In the sea roach Ligia exotica, the temporal separation of posterior and anterior molts is approximately 24 hours, despite having an open circulatory system presumably carrying hormones throughout both halves of the body. In fact, injections of the molting hormone ecdysterone failed to alter the delay in molting between the two halves of the animal (Montane 1988).

The Crustacean Integument: Structure and Function

Fig. 5.13. (A) X-ray map of the calcium (Ca) distribution in an embedded and polished posterior piece of dorsal carapace of Callinectes sapidus that includes the suture. Note the region of the suture in the exocuticle (ex) as indicated by the arrowhead and in the endocuticle (en) as bounded by the arrows (from Priester et al. 2005, fig. 7b, used with permission). (B) Intermolt cuticle of Callinectes sapidus in the region of the suture stained with the fluorescein isothiocyanate–labeled lectin Lens culinaris agglutinin. Note the lack of staining in the exocuticle (ex), the heavy staining in the suture (arrowhead), and the moderate staining of the endocuticle (en) except in the region of the suture (arrow) (from Priester et al. 2005, fig. 1b). (C–F) Backscattered electron micrographs of embedded and cut samples of Callinectes sapidus cuticle in early (C) and late D2 (D–F) stages in the region containing the suture. Arrows indicate the suture region of the endocuticle (en); arrowheads, the suture region of the exocuticle (ex). Note the different rate of etching in the three regions of the cuticle: anterior (D), middle (E), and posterior (F) (from Priester et al. 2005, fig. 5). All figures used with permission.

157

Functional Morphology and Diversity

Calcified cuticle

Apolysis Intermolt

Uncalcified Cuticle

158

Ecdysis Postmolt

Premolt

Dorsal C4 Carapace

D0

D1’

D1’’

D1’’’ D2 D3 D4

A1 A2 B1

B2 C1-C3

Arthrodial C4 Membrane

D0

D1’

D1’’

D1’’’ D2 D3 D4

A1 A2 B1

B2 C1-C3

Inner Cuticle Respiratory Gill Transport Gill No deposition

Postmolt deposition

Premolt deposition

Fig. 5.14. Timeline for cuticle deposition in various tissues of Callinectes sapidus. From Williams et al. (2003, fig. 6), used with permission.

CONTROL OF MINERALIZATION As mentioned above, in order for animals to enlarge their exoskeletons and grow as they molt, they must take up water soon after ecdysis and expand the new layers of the cuticle before they begin to harden (Drach 1939, Mykles 1980). Only after this postecdysial expansion during stage A1 do the new epi- and exocuticle undergo quinone tanning (Travis 1957) and mineralization by deposition of calcium carbonate (Drach 1939, Bouligand 1970, Giraud-Guille and Quintana 1982). Thereafter, the endocuticle is synthesized and mineralizes as it is deposited (postecdysial stages A 2 through C3). Obviously, the timing of the onset of tanning and calcium carbonate deposition is of critical importance. The deposition of mineral within a biological system requires a number of criteria to be met. The prime criterion is the establishment of concentrations of the composite ions that exceed the solubility product for precipitation. This generally occurs in the context of a microenvironment with identifiable boundaries, one of which is commonly an epithelium that can transport ions into and out of this compartment. Second, there is an organic component that serves to heterogeneously nucleate the mineral by lowering the superficial energy of precipitation. The organic matrix also is responsible for determining the mineral morph and directing the second-order pattern of mineralization. Permeability Changes The microenvironment within the calcifying cuticle is bounded on the inner surface by the hypodermis, while the outer boundary is the epicuticle. Calcium is transported into this space by the hypodermis via the action of a calcium-ATPase and sodium-calcium exchanger (Roer 1980, Greenaway et al. 1995). The transport of bicarbonate is mediated by carbonic anhydrase and likely involves the activities of HCO3–-ATPase, a Cl–/HCO3– exchanger, and a Na+/ H+ exchanger (Giraud 1981, Roer and Dillaman 1993). However, during premolt, the preexuvial cuticle (including the epicuticle; Fig. 5.15A,B) is freely permeable to ions and breakdown

The Crustacean Integument: Structure and Function

Fig. 5.15. (A–C) Transmission electronic micrograph of dorsal carapace from a 1-hour postmolt crab freshly fixed by uranyl acetate: low (A) and high (C) magnification of cross sections through the epicuticle. (B) Tangential sections through the inner epicuticle and distal exocuticle. Note the dense vertical fiber (dvf), epicuticular canals (ec), epicuticular fibers (ef), epicuticular roots (er), horizontal fibers (hf), inner epicuticle (ie), and outer epicuticle (oe) (from Modla 2006, fig. 10a–c, used with permission from the author). (D) Scanning electron micrograph of the epicuticle in early postmolt cuticle in Callinectes sapidus. Note the spaces between the perpendicular fibers of the epicuticle (arrowheads) and the continuous outer layer of the epicuticle (arrows).

products of the organic matrix that pass through it during the resorption of the old exoskeleton (Roer 1980, Compére et al. 1998). Following ecdysis, mineralization cannot commence until the outer boundary of the cuticular microenvironment, the newly exposed epicuticle, becomes impermeable to calcium and bicarbonate (see Fig. 5.15C,D). This transition occurs within 15 minutes postecdysis (Williams et al. 2009). Matrix Proteins and Glycoproteins Evidence suggests that the control of calcium carbonate deposition in crustacean cuticle resides in the organic matrix of the cuticle itself, rather than in, for example, the ion-pumping activity of the underlying cellular layer (Roer and Dillaman 1984, Roer et al. 1988, Roer and Towle 2005). Using an in vitro nucleation assay, it was shown that pieces of cuticle isolated as early as 3 hours after ecdysis, stripped of all underlying cellular material, and decalcified with EDTA

159

160

Functional Morphology and Diversity can nucleate calcium carbonate crystals, whereas pieces containing the same cuticle layers removed from a preecdysial or an ecdysial animal and treated in the same way cannot nucleate calcite (Roer et al. 1988, Shafer et al. 1995). That is, during preecdysis and early postecdysis, the new cuticle is incapable of nucleating calcite. Soon after ecdysis, however, mineralization commences. Thus, in the preexuvial layers of the crustacean cuticle, we are presented with an organic matrix that at one time (preecdysis and early postecdysis) cannot mineralize but that over a period of only a few hours after ecdysis is transformed into a matrix that can and does mineralize. This synchronized temporal separation of the definitive control events potentially allows for isolation and identification of the regulatory proteins involved. Moreover, the cuticle contains regions that calcify (e.g., the dorsal carapace) and regions that neither calcify nor tan (the arthrodial cuticle). Comparisons of these two types of cuticle can provide an excellent means to elucidate those cuticular proteins important for mineralization. Comparisons of cuticular proteins across arthropod groups can also be instructive since insects do not mineralize their exoskeletons. Finally, a search for matrix proteins that stabilize ACC is also instructive. Control is necessary for the persistence of ACC in the cuticle since crystalline forms of calcium carbonate (calcite and aragonite) are thermodynamically favored and more stable. Evidence for such a matrix protein has recently been described by Shechter et al. (2008). A 65-kDa protein, GAP-65, was found in the gastrolith (the storage organ for calcium carbonate in the crayfish). This protein binds ACC and inhibits the formation of calcite. Dramatic postecdysial changes in crab cuticular glycoproteins synchronous with initial mineralization were documented at the histochemical level (Marlowe et al. 1994), by lectin blotting following SDS-PAGE (Shafer et al. 1994, 1995) and by their ability to affect calcite nucleation (Coblentz et al. 1998). Based on these data, it was hypothesized that acid-soluble proteins form nucleating sites in the cuticle and that glycoproteins active only in preecdysis and for the first few hours postecdysis shield nucleating proteins from calcium and carbonate ions or otherwise inhibit their action (Coblentz et al. 1998). A cuticular glycoprotein that disappears from lectin blots postecdysis (Shafer et al. 1995) was assumed to be the inhibitor. Its purification from relatively large amounts of ecdysial cuticle provided strong support for this hypothesis (Tweedie et al. 2004). This glycoprotein is mucinlike, being 55% carbohydrate and having both N- and O -linkages. Immunoblot analysis suggests that several changes occur in its glycosylation pattern during the first few hours after ecdysis. Immunohistochemical staining decreases in the interprismatic septa as early as 2 hours after ecdysis, coincident with the first appearance of mineralization. This temporal and fine-scale spatial correlation between the loss of a possible inhibitor and the initiation of calcium carbonate deposition, and the fact that the protein does not exist in the arthrodial membrane pre- or postecdysis, strongly suggests a role in control of nucleation. The means by which cuticular glycoproteins can be modified in situ after the molt is still a matter of active investigation. However, evidence suggests that N-acetylhexosaminidase treatment can alter cuticular glycans and change the ability of cuticle explants to calcify in a way that mimics in vivo postecdysis processes (Pierce et al. 2001). A glycosidase with this activity appears in the cuticle during the early postecdysial hours coincident with the changes in glycoprotein profiles observed in vivo (Roer et al. 2001, Roer and Towle 2004). Cuticular matrix proteins have been sequenced from several decapod crustaceans, either directly or by virtual translation of cDNA sequences. They include multiple sequences from Homarus (Kragh et al. 1997, Andersen 1998, Nousiainen et al. 1998), Cancer pagurus (Andersen 1999), Marsupenaeus japonicus (Endo et al. 2000, Watanabe et al. 2000), Callinectes (Wynn and Shafer 2005, Faircloth and Shafer 2007), and Portunus pelagicus (Kuballa et al. 2007). Functions for these proteins are at present only matters of speculation. However, a group of proteins from

The Crustacean Integument: Structure and Function the crayfish Procambarus have been produced in recombinant systems and actually evaluated for their ability to bind calcium carbonate. Calcification-associated peptides 1 and 2 (CAP-1 and CAP-2), expressed in the crayfish tail fan during postecdysis, have been shown to affect in vitro calcium carbonate formation and have been implicated as components that control the initiation of mineralization (Inoue et al. 2001, 2003, 2004). CAP-1 can alter the structure of nanocrystals deposited on chitin-coated surfaces (Sugawara et al. 2006). Additionally, Casp-2, a more soluble Procambarus cuticular protein with an unrelated structure, also regulates calcium carbonate formation when the recombinant protein is added to in vitro systems (Inoue et al. 2008).

CONCLUSION AND FUTURE DIRECTIONS While there is considerable conservation in the basic architecture of the arthropod exoskeleton across a wide array of taxa, there are also remarkable differences in cuticle characteristics. These differences manifest across the crustacean taxa and even among different species of decapods. Differences are also evident within an individual, depending upon the mechanical and physiological function that the cuticle serves in different regions of the body. Cuticles in the gills of aquatic crabs and the branchial chambers of terrestrial crabs are thin, nonmineralized, soft, and highly permeable, while that of the cheliped is heavily mineralized, hard as concrete, and completely impermeable. Moreover, cuticles in different regions of the body of an individual crustacean undergo synthesis and degradation during different times within the molt cycle, perhaps most strikingly in the biphasic molting of isopods. These variations reflect temporal delays or differential sensitivities of the underlying hypodermis to the hormonal cues that initiate these events in preparation for the molt. Clearly, the hypodermis holds the key to understanding the differences in cuticle structure, function, and formation within the molt cycle. Recent work in our laboratory and others has begun to elucidate the different expression of proteins and glycoproteins both spatially and temporally and may explain the differences in cuticular structure and thus function. This approach has perhaps been most successful in providing insight into the ability of cuticle to calcify (Inoue et al. 2001, 2003, 2004, 2008, Wynn and Shafer 2005, Sugawara et al. 2006, Faircloth and Shafer 2007). The establishment and expansion of expression libraries for crustaceans hold great promise for discovering the cuticular components that are crucial for defining the differences in cuticle morphology and function.

ACKNOWLEDGMENTS We acknowledge the invaluable assistance of Mr. D. Mark Gay with the preparation of the figures. We thank the Drs. Raabe, Romano, Cribb, and Hild for supplying high-resolution images of their figures and their publishers for permission to use them.

REFERENCES Al-Sawalmih, A., C. Li, S. Siegel, H. Fabritius, S. Yi, D. Raabe, P. Fratzl, and O. Paris. 2008. Microtexture and chitin/calcite orientation relationship in the mineralized exoskeleton of the American lobster. Advanced Functional Materials 18:3307–3314. Andersen, S.O. 1998. Characterization of proteins from arthrodial membranes of the lobster Homarus americanus. Comparative Biochemistry and Physiology Part A 121:375–383. Andersen, S.O. 1999. Exoskeletal proteins from the crab, Cancer pagurus. Comparative Biochemistry and Physiology Part A 123:203–211.

161

162

Functional Morphology and Diversity Andrews, S.C., and R. M. Dillaman . 1993. Ultrastructure of the gill epithelia in the crayfish Procambarus clarkii at different stages of the molt cycle. Journal of Crustacean Biology 13:77–86. Becker, A., A. Ziegler, and M. Epple. 2005 . The mineral phase in the cuticles of two species of Crustacea consists of magnesium calcite, amorphous calcium carbonate, and amorphous calcium phosphate. Dalton Transactions 2005:1814–1820. Boselmann, F., P. Romano, H. Fabritius, D. Raabe, and M. Epple. 2007. The composition of the exoskeleton of two Crustacea: The American lobster Homarus americanus and the edible crab Cancer pagurus. Thermochimica Acta 463:65–68. Bouligand, Y. 1970. Aspects ultrastructuraux de la calcification chez les crabes. Septieme Congres Internationale de Microscopie Electronique (Grenoble) 3:105–106. Bouligand, Y. 1972. Twisted fibrous arrangements in biological materials and cholesteric mesophases. Tissue and Cell 4:189 –190, 192–217. Chen, P.-Y., A.Y.-M. Lin, J. McKittrick , and M.A. Meyers. 2008. Structure and mechanical properties of crab exoskeletons. Acta Biomaterialia 4:587–596. Coblentz , F.E., T.H. Shafer, and R.D. Roer. 1998. Cuticular proteins from the blue crab alter in vitro calcium carbonate mineralization. Comparative Biochemistry and Physiology Part B 121:349 –360. Compére, P. 1995 . Fine structure and morphogenesis of the sclerite epicuticle in the Atlantic shore crab Carcinus maenas. Tissue and Cell 27:525–538. Compére, P., M.F. Jaspar-Versali, and G. Goffinet. 2002. Glycoproteins from the cuticle of the Atlantic shore crab Carcinus maenas: I. Electrophoresis and Western-blot analysis by use of lectins. Biological Bulletin 202:61–73. Compére, P., A. Thorez , and G. Goffinet. 1998. Fine structural survey of old cuticle degradation during pre-ecdysis in two European Atlantic crabs. Tissue and Cell 30:41–56. Cribb, B.W., A. Rathmell, R. Charters, R. Rasch, H. Huang , and I.R. Tibbetts. 2009. Structure, composition and properties of naturally occurring non-calcified crustacean cuticle. Arthropod Structure and Development 38:173–178. Dickson, J.S., R.M. Dillaman, R.D. Roer, and D.B. Roye. 1991. Distribution and characterization of ion transporting and respiratory filaments in the gills of Procambarus clarkii. Biological Bulletin 180:154–166. Dillaman, R., S. Hequembourg , and M. Gay. 2005 . Early pattern of calcification in the dorsal carapace of the blue crab, Callinectes sapidus. Journal of Morphology 263:356 –374. Drach, P. 1939. Mue et cycle d’intermue chez les crustacés décapodes. Annales d’Institut Oceanographie, Paris 19:103–391. Drach, P., and C. Tchernigovtzeff. 1967. Sur la méthode de détermination des stades d’intermue et son application générale aux crustaces. Vie et Milieu 18:595–610. Elliott, E.A., and R.M. Dillaman. 1999. Formation of the inner branchiostegal cuticle of the blue crab, Callinectes sapidus. Journal of Morphology 242:47–56. Endo, H., P. Persson , and T. Watanabe. 2000. Molecular cloning of the crustacean DD4 cDNA encoding a Ca 2+-binding protein. Biochemical and Biophysical Research Communications 276:286 –291. Fabritius, H.-O., C. Sachs, P.R. Triguero, and D. Raabe. 2009. Influence of structural principles on the mechanics of a biological fiber-based composite material with hierarchical organization: The exoskeleton of the lobster Homarus americanus. Advanced Materials 21:391–400. Faircloth, L.M., and T.H. Shafer. 2007. Differential expression of eight transcripts and their roles in the cuticle of the blue crab, Callinectes sapidus. Comparative Biochemistry and Physiology Part B 146:370 –383. Giraud, M.M. 1977. Histochemistry of 1st stages of mineralization of Carcinus maenas cuticle—attempt of localization of carbonic anhydrase. Comptes Rendus Hebdomadaires des Seances de l’Academie des Sciences Serie D 284:1541–1544. Giraud, M.M. 1981. Carbonic anhydrase activity in the integument of the crab Carcinus maenas during the intermolt cycle. Comparative Biochemistry and Physiology Part A 69:381–387. Giraud-Guille, M.M. 1984 . Calcification initiation sites in the crab cuticle: The interprismatic septa. An ultrastructural cytochemical study. Cell and Tissue Research 236:413–420.

The Crustacean Integument: Structure and Function Giraud-Guille, M.M., and C. Quintana. 1982. Secondary ion microanalysis of the crab calcified cuticle: Distribution of mineral elements and interactions with the cholesteric organic matrix. Biologie Cellulaire 44:57–67. Greenaway, P., R.M. Dillaman, and R.D. Roer. 1995 . Quercitin-dependent ATPase activity in the hypodermal tissue of Callinectes sapidus during the moult cycle. Comparative and Biochemical Physiology Part A 111:303–312. Greenaway, P., and C.A. Farrelly. 1984 . The venous system of the terrestrial crab Ocypode cordimanus (Desmarest 1825) with particular reference to the vasculature of the lungs. Journal of Morphology 181:133–142. Greenaway, P., and C. Farrelly. 1990. Vasculature of the gas-exchange organs in air-breathing brachyurans. Physiological Zoology 63:117–139. Halcrow, K., and E.L. Bousfield. 1987. Scanning electron microscopy of surface microstructures of some gammaridean amphipod crustaceans. Journal of Crustacean Biology 7:274–287. Hegdahl, T., F. Gustavsen, and J. Silness. 1977. The structure and mineralization of the carapace of the crab (Cancer pagurus L.): The exocuticle. Zoologica Scripta 6:101–105. Hild, S., O. Marti, and A. Ziegler. 2008. Spatial distribution of calcite and amorphous calcium carbonate in the cuticle of the terrestrial crustaceans Porcellio scaber and Armadillidium vulgare. Journal of Structural Biology 163:100 –108. Inoue, H., T. Ohira, N. Ozaki, and H. Nagasawa. 2003. Cloning and expression of a cDNA encoding a matrix peptide associated with calcification in the exoskeleton of the crayfish. Comparative Biochemistry and Physiology Part B 136:755–765. Inoue, H., T. Ohira, N. Ozaki, and H. Nagasawa. 2004 . A novel calcium-binding peptide from the cuticle of the crayfish, Procambarus clarkii. Biochemistry and Biophysics Research Communications 318:649 –654. Inoue, H., N. Ozaki, and H. Nagasawa. 2001. Purification and structural determination of a phosphorylated peptide with anti-calcification and chitin-binding activities in the exoskeleton of the crayfish, Procambarus clarkii. Bioscience Biotechnology and Biochemistry 65:1840–1848. Inoue, H., N. Yuasa-Hashimoto, M. Suzuki, and H. Nagasawa. 2008. Structural determination and functional analysis of a soluble protein associated with calcification of the exoskeleton of the crayfish, Procambarus clarkii. Bioscience Biotechnology and Biochemistry 72:2697–2707. Kragh, M., L. Mølbak , and S.O. Andersen. 1997. Cuticular proteins from the lobster, Homarus americanus. Comparative Biochemistry and Physiology Part B 118:147–154. Krywka, C., C. Sternemann, M. Paulus, N. Javid, R. Winter, A. Al-Sawalmih, S. Yi, D. Raabe, and M. Tolan. 2007. The small-angle and wide-angle X-ray scattering set-up at beamline BL9 of DELTA. Journal of Synchrotron Radiation 14:244–251. Kuballa, A.V., D.J. Merritt, and A. Elizur. 2007. Gene expression profiling of cuticular proteins across the moult cycle of the crab Portunus pelagicus. BMC Biology 5:45 . Locke, M. 1959. The cuticular pattern in an insect, Rhodnius prolixus Stal. Journal of Experimental Biology 36:459 –476. Locke, M. 1960. The cuticular pattern in an insect—the intersegmental membranes. Journal of Experimental Biology 37:398 –407. Locke, M. 1961. Pore canals and related structures in insect cuticle. Journal of Biophysical and Biochemical Cytology 10:589 –618. Marlowe, R.L., R.M. Dillaman, and R.D. Roer. 1994 . Lectin binding by the crustacean cuticle: The cuticle of Callinectes sapidus throughout the molt cycle, and the intermolt cuticle of Procambarus clarkii and Ocypode quadrata. Journal of Crustacean Biology 14:231–246. Melnick , C.A., Z. Chen, and J.J. Mecholsky. 1996. Hardness and toughness of exoskeleton material in the stone crab, Menippe mercenaria. Journal of Materials Research 11:2903–2907. Modla, S. 2006. The effect of fixation on the morphology of the late premolt and early postmolt cuticle of the blue crab, Callinectes sapidus . MS thesis, University of North Carolina, Wilmington . Montane, M.M. 1988. A study of the biphasic molt mechanism of the supralitoral isopod, Ligia exotica . MS thesis, University of North Carolina, Wilmington . Mykles, D.L. 1980. The mechanism of fluid absorption at ecdysis in the American lobster, Homarus americanus. Journal of Experimental Biology 84:89 –101.

163

164

Functional Morphology and Diversity Neville, A.C. 1975 . Biology of the arthropod cuticle. Springer, New York. Nousiainen, M., K. Rafn, L. Skou, P. Roepstorff, and S.O. Andersen. 1998. Characterization of exoskeletal proteins from the American lobster, Homarus americanus. Comparative Biochemistry and Physiology Part B 119:189 –199. Pierce, D.C., K.D. Butler, and R.D. Roer. 2001. Effects of exogenous N-acetylhexosaminidase on the structure and mineralization of the post-ecdysial exoskeleton of the blue crab, Callinectes sapidus. Comparative Biochemistry and Physiology Part B 128:691–700. Pratoomchat, B., P. Sawangwong , R. Guedes, M. de Lurdes Reis, and J. Machado. 2002. Cuticle ultrastructure changes in the crab Scylla serrata over the molt cycle. Journal of Experimental Zoology 293:414–426. Priester, C., R.M. Dillaman, and D.M. Gay. 2005 . Ultrastructure, histochemistry, and mineralization patterns in the ecdysial suture of the blue crab, Callinectes sapidus. Microscopy and Microanalysis 11:479 –499. Raabe, D., A. Al-Sawalmih, S.-B. Yi, and H. Fabritius. 2007. Preferred crystallographic texture of а-chitin as a microscopic and macroscopic design principle of the exoskeleton of the lobster Homarus americanus. Acta Biomaterialia 3:882–895. Raabe, D., P. Romano, C. Sachs, A. Al-Sawalmih, H.-G. Brokmeier, S.-B. Yi, G. Servos, and H.G. Hartwig. 2005a . Discovery of a honeycomb structure in the twisted plywood patterns of fibrous biological nanocomposite tissue. Journal of Crystal Growth 283:1–7. Raabe, D., P. Romano, C. Sachs, H. Fabritius, A. Al-Sawalmih, S.-B. Yi, G. Servos, and H.G. Hartwig. 2006. Microstructure and crystallographic texture of the chitin-protein network in the biological composite material of the exoskeleton of the lobster Homarus americanus. Materials Science and Engineering A 421:143–153. Raabe, D., C. Sachs, and P. Romano. 2005b. The crustacean exoskeleton as an example of a structurally and mechanically graded biological nanocomposite material. Acta Materialia 53:4281–4292. Read, A.T., and D.D. Williams. 1991. The distribution, external morphology and presumptive function of the surface microstructures of Gammarus pseudolimnaeus (Crustacea: Amphipoda) with emphasis on the calceolus. Canadian Journal of Zoology 69:853–865. Roer, R.D. 1980. Mechanisms of resorption and deposition of calcium in the carapace of the crab, Carcinus maenas. Journal of Experimental Biology 88:205–218. Roer, R.D., S.K. Burgess , C.G. Miller, and M.B. Dail. 1988. Control of calcium carbonate nucleation in pre- and postecdysial crab cuticle. Pages 21–24 in C.S. Sikes and A.P. Wheeler, editors. Chemical aspects of regulation of mineralization. University of South Alabama Publication Service, Mobile. Roer, R.D., and R.M. Dillaman. 1984 . The structure and calcification of the crustacean cuticle. American Zoologist 24:893–909. Roer, R.D., and R.M. Dillaman. 1993. Molt-related change in integumental structure and function. Pages 1–37 in M. Horst and J.A. Freeman, editors. The crustacean integument—morphology and biochemistry. CRC Press, Boca Raton, FL. Roer, R.D., K.E. Halbrook , and T.H. Shafer. 2001. Glycosidase activity in the post-ecdysial cuticle of the blue crab, Callinectes sapidus. Comparative Biochemistry and Physiology Part B 128:683–690. Roer, R.D., and D.W. Towle. 2004 . Partial nucleotide sequence of a putative cuticular hexosaminidase from the blue crab, Callinectes sapidus. Bulletin of the Mount Desert Island Biological Laboratory 43:40–42. Roer, R.D., and D.W. Towle, 2005 . Partial nucleotide sequence and expression of plasma membrane Ca-ATPase in the hypodermis of the blue crab, Callinectes sapidus. Bulletin of the Mount Desert Island Biological Laboratory 44:40 –42. Romano, P., H. Fabritius, and D. Raabe. 2007. The exoskeleton of the lobster Homarus americanus as an example of a smart anisotropic biological material. Acta Biomaterialia 3:301–309. Sachs, C., H. Fabritius, and D. Raabe. 2006. Hardness and elastic properties of dehydrated cuticle from the lobster Homarus americanus obtained by nanoindentation. Journal of Materials Research 21:1987–1995. Sakamoto, K., W. Honto, M. Iguchi, N. Ogawa, K. Ura, and Y. Takagi. 2009. Post-molt processes of cuticle formation and calcification in the Japanese mitten crab Eriocheir japonicus. Fisheries Science 75:91–98.

The Crustacean Integument: Structure and Function Shafer, T.H., R.D. Roer, C. Midgette-Luther, and T.A. Brookins. 1995 . Postecdysial cuticle alteration in the blue crab, Callinectes sapidus: Synchronous changes in glycoproteins and mineral nucleation. Journal of Experimental Zoology 271:171–182. Shafer, T.H., R.D. Roer, C.G. Miller, and R.M. Dillaman. 1994 . Postecdysial changes in the protein and glycoprotein composition of the cuticle of the blue crab Callinectes sapidus. Journal of Crustacean Biology 14:210 –219. Shechter, A., L. Glazer, S. Cheled, E. Mor, S. Weil, A. Berman, S. Bentov, E.D. Aflalo, I. Khalaila, and A. Sagi. 2008. A gastrolith protein serving a dual role in the formation of an amorphous mineral containing extracellular matrix. Proceedings of the National Academy of Science USA 105:7129 –7134. Skinner, D.M. 1962. The structure and metabolism of a crustacean integumentary tissue during a molt cycle. Biological Bulletin 123:635–647. Skinner, D.M. 1965 . Amino acid incorporation into protein during the molt cycle of the land crab, Gecarcinus lateralis. Journal of Experimental Zoology 160:225–233. Skinner, D.M., S.S. Kumari, and J.J. O’Brien . 1992. Proteins of the crustacean exoskeleton. American Zoologist 32:470–484. Soejoko, D.S., and M.O. Tjia. 2003. Infrared spectroscopy and X ray diffraction study on the morphological variations of carbonate and phosphate compounds in giant prawn (Macrobrachium rosenbergii) skeletons during its moulting period. Journal of Materials Science 38:2087–2093. Steel, C.G.H. 1980. Mechanisms of coordination between moulting and reproduction in terrestrial isopod Crustacea. Biological Bulletin 159:206 –218. Steel, C.G.H. 1982. Stages of the intermoult cycle in the terrestrial isopod Oniscus asellus and their relation to biphasic cuticle secretion. Canadian Journal of Zoology 60:429 –437. Sugawara, A., T. Nishimura, Y. Yamamoto, H. Inoue, H. Nagasawa, and T. Kato. 2006. Self-organization of oriented calcium carbonate/polymer composites: Effects of a matrix peptide isolated from the exoskeleton of a crayfish. Angewandte Chemie International Edition 45:2876 –2879. Tarsitano, S.F., K.L. Lavalli, F. Horne, and E. Spanier. 2006. The constructional properties of the exoskeleton of homarid, palinurid, and scyllarid lobsters. Hydrobiologia 557:9 –20. Taylor, H.H., and P. Greenaway. 1984 . The role of the gills and branchiostegites in gas exchange in a bimodally breathing crab, Hothuisana transversa: Evidence for a facultative change in the distribution of the respiratory circulation. Journal of Experimental Biology 111:103–121. Taylor, J.R.A., J. Hebrank , and W.M. Kier. 2007. Mechanical properties of the rigid and hydrostatic skeletons of molting blue crabs, Callinectes sapidus Rathbun. Journal of Experimental Biology 210:4272–4278. Taylor, J.R.A., and W.M. Kier. 2003. Switching skeletons: Hydrostatic support in molting crabs. Science 301:209 –210. Travis, D.F. 1955 . The molting cycle of the spiny lobster, Panulirus argus Latreille. II. Pre-ecdysial histological and histochemical changes in the hepatopancreas and integumental tissues. Biological Bulletin 108:88 –112. Travis, D.F. 1957. The molting cycle of the spiny lobster, Panulirus argus Latreille. IV. Post-ecdysial histological and histochemical changes in the hepatopancreas and integumental tissues. Biological Bulletin 113:451–479. Travis, D.F. 1963. Structural features of mineralization from tissue to macromolecular levels of organization in the decapod Crustacea. Annals of the New York Academy of Science 109:177–245. Travis, D.F. 1965 . The deposition of skeletal structures in the Crustacea. 5. The histomorphological and histochemical changes associated with the development and calcification of the branchial exoskeleton in the crayfish, Orconectes virilis Hagen. Acta Histochemica 20:193–222. Tweedie, E.P., F.E. Coblentz , and T.H. Shafer. 2004 . Purification of a soluble glycoprotein from the uncalcified ecdysial cuticle of the blue crab Callinectes sapidus and its possible role in initial mineralization. Journal of Experimental Biology 207:2589 –2598. Watanabe, T., P. Persson, H. Endo, and M. Kono. 2000. Molecular analysis of two genes, DD9A and B, which are expressed during the post molt stage in the decapod crustacean Penaeus japonicus. Comparative Biochemistry and Physiology Part B 125:127–136.

165

166

Functional Morphology and Diversity Wigglesworth, V.B. 1933. The physiology of the cuticle and ecdysis in Rhodnius prolixus (Triatomidae, Hemiptara); with special reference to the function of the oenocytes and of the dermal glands. Quarterly Journal of Microscopical Science 76:269 –318. Wilcockson, D.C., and S.G. Webster. 2008. Identification and developmental expression of mRNAs encoding putative insect cuticle hardening hormone, bursicon in the green shore crab Carcinus maenas. General and Comparative Endocrinology 156:113–125. Williams, C.L., R.M. Dillaman, E.A. Elliott, and D.M. Gay. 2003. Formation of the arthrodial membrane in the blue crab, Callinectes sapidus. Journal of Morphology 256:260–269. Williams, D.L., S. Modla , R.D. Roer, and R.M. Dillaman. 2009. Post-ecdysial change in the permeability of the exoskeleton of the blue crab, Callinectes sapidus. Journal of Crustacean Biology 29:550–555. Wynn, A., and T.H. Shafer. 2005 . Four differentially expressed cDNAs in Callinectes sapidus containing the Rebers-Riddiford consensus sequence. Comparative Biochemistry and Physiology Part B 141:294–306. Ziegler, A., H. Fabritius, and M. Hagedorn. 2005 . Microscopical and functional aspects of calcium-transport and deposition in terrestrial isopods. Micron 36:137–153. Ziegler, A., M. Hagedorn, G.A. Ahearn, and T.H. Carefoot. 2007. Calcium translocations during the moulting cycle of the semiterrestrial isopod Ligia hawaiiensis (Oniscidea, Crustacea). Journal of Comparative Physiology B 177:99 –108.

6 THE CRUSTACEAN INTEGUMENT: SETAE, SETULES, AND OTHER ORNAMENTATION

Anders Garm and Les Watling

Abstract The cuticle plays an important role in many aspects of crustacean biology, since it is the interface to the surrounding world. Thus, the cuticle displays many structural specializations all over the body. The structures considered here are setae, setules, denticles, and spines. We provide definitions for them and discuss their functional morphology and development, with the main focus on setae. We recognize seven types of setae based on their detailed external morphology: plumose, pappose, composite, serrate, papposerrate, simple, and cuspidate. In support of the categorization of these setae, each seems to correlate with a specific functional outcome such as feeding, grooming, and locomotion. Setae are also important sensory organs, and in crustaceans they are normally bimodal chemo- and mechanoreceptors, but there are also indications of thermo-, osmo-, and hygrosensitivity. Little can be learned about the sensory functions from the external morphology of setae, but their ultrastructure seems to provide better cues. In particular, mechanoreceptors display structures related to transduction mechanisms, with the scolopale as a good example. Still, too few data are available outside malacostracans to draw general conclusions for all crustaceans, underlining the need for multidisciplinary and broad intertaxon studies. Less is known about the functional morphology and development of setules and denticles in the general cuticle, but they seem to be homologous with similar structures on the setae. Arthropods outside Crustacea also have setae in their cuticle, and many shared features can be found. They are especially well studied in insects, where many correlations between structure and function have been shown.

INTRODUCTION TO THE STRUCTURES OF THE CRUSTACEAN CUTICLE: DEFINITIONS/CLASSIFICATION One of the defining characters of crustaceans as well as other arthropods is their external skeleton, the cuticle (see also chapter 5). The cuticle plays a major role in most aspects of crustacean Functional Morphology and Diversity. Edited by Les Watling and Martin Thiel. © 2013 Oxford University Press. Published 2013 by Oxford University Press.

167

168

Functional Morphology and Diversity biology, and this has led to a vast number of structural and functional specializations. Many of these specializations lie within the detailed surface structures, and they are the topic of this chapter. First, we provide an overview of the diversity of these structures and their functions and use this to suggest a classification system. The main part of the review then focuses in detail on the major group of cuticular specializations, the setae, since they are by far the most studied and have the greatest functional diversity and importance. We end by comparing with data from other arthropod groups and listing suggestions for where future research in this field is most needed and will be most fruitful. When observing crustaceans with the naked eye, many of the cuticular specializations are visible in a large number of species (Fig. 6.1A). Some body parts and especially the appendages appear furry (Fig. 6.1B), and the hairlike structures found in these areas are outgrowths of the general cuticle, normally with a distinct articulation at the base, making them flexible (Figs. 6.1D, 6.2). There is a general consensus that these structures are homologous within Crustacea and are also probably homologous with similar structures in other arthropods. Many terms have been used for these structures, such as setae, sensilla, bristles, or even “hairs.” For crustaceans, the most often used term is setae, and it will therefore be used here. Even though setae are in general considered homologous, it is difficult to decide which cuticular projections to include in this term. A number of authors have addressed this problem and provided definitions of what they considered setae. Thomas (1970) was one of the first to do so, and he proposed that all elongate outgrowths with distal pores were setae. His work was based on light microscopy, and electron microscopy work has since shown his definition to be far too narrow. Fish (1972) considered elongate outgrowths filled with “cytoplasm” as setae, but this very broad definition will include many other structures, such as spines (see below), and exclude setae with no cells in the lumen. Some authors have used the size of the cuticular structures as a basis for classification. This has led to such terms as microsetae (Jacques 1989) and microtrichs (Cuadras 1982, Steele and Steele 1997, 1999), but we do not approve of this approach. If a structure complies with our given definition (see below), we will consider it a seta no matter the size, and we see no reason to believe that they cannot be small. In fact, we believe that in small crustaceans, such as nauplius larvae, there has been strong selection pressure for miniaturizing the setae. An evolutionary perspective was taken by Watling (1989), who stressed the need for a definition based on homologies. He suggested that the articulation with the general cuticle is such a homology and used this structure to define setae from other cuticular outgrowths. This definition has been widely accepted as it seems to hold true for the vast majority of setae, and the “stem seta” probably also had such an articulation. When considering the diversity of present-day crustaceans, though, Watling’s definition runs into some problems, which were first addressed in an earlier review (Garm 2004b). Some of the articulated outgrowths have an external and internal morphology so similar to long setules found on some of the setae that there are no structural arguments to consider them as being different. They are commonly found on the mouthparts of decapods and peracarids (Fig. 6.1C), and we suggest that they should be included in the term setules (see below). The other problem concerns a loss of the articulation between the setae and the general cuticle. This has probably happened a number of times in several crustacean lineages to encompass mechanical functions requiring a very sturdy seta (Garm and Høeg 2001, Garm 2004a). Clear examples of such loss are seen for the spinelike projection found on the basis of maxilla 1 of the squat lobster Munida sarsi (Fig. 6.2B). These unarticulated projections are innervated, have a continuous lumen, and have a cellular arrangement very similar to other setae. Further, structures undoubtedly homologous with the spinelike projections (they are situated in the same place and arranged in the same two parallel rows in other decapod species) are typical setae with clear articulation (Garm 2004b). The same situation is seen for unarticulated spinelike

Setae, Setules, and Other Ornamentation

A

B

C

E

D

F

G

Fig. 6.1. Structures in the crustacean cuticle. (A) At the macroscopic level, many crustaceans, such as the hermit crab Parapagurus sulcata, appear furry because of very heavy setation (picture courtesy of Dr. Jens T. Høeg). (B) Maxilliped 1 of the hermit crab Pagurus bernhardus displaying heavy setation, especially on the medial edges of the coxa and basis. Several types of setae are present. (C) Setules from paragnath of P. bernhardus are clearly articulated with the general cuticle (inset). (D) Between the setae (S) on the mouthpart of Panulirus argus, the cuticle is filled with teethlike structures (denticles). (E) Ultrastructure of setules from the paragnath of Penaeus monodon shows that they are made entirely of cuticle and lack a lumen and innervation. (F) Ultrastructure of setae show a round, hollow base filled with sheath cells (ShC). Cu, setal cuticle. (G) Close-up of the central part of the setal lumen showing that the semicircular sheath cells (ShC) encircle the outer dendritic segments (ODS) of a number of sensory cells.

169

170

Functional Morphology and Diversity

Fig. 6.2. Details of the external morphology of setae. (A) Spinelike setae from maxilla 1 of Munida sarsi, with no apparent articulation at the base (arrows). (B) Plumose setae displaying a supracuticular articulation (arrows) with the general cuticle, making them very flexible. (C) Most setae have an infracuticular articulation (arrows) with the general cuticle, reducing their flexibility. (D) Setules from a composite seta showing an articulation (arrows) with the setal shaft. (E) Some setae display unarticulated teethlike structures arranged parallel to the setal shaft. These denticles (De) are found in two rows on the distal half of the setae often together with small setules (S). (F) On some setae, there is a graduated transformation between setules (S) and denticles (De). (G) Many setae display a terminal pore (TP) at the tip, often associated with small scalelike setules. (H) The tip of a seta used for grooming the gills. Such setae often have a specialized tip. (I) A newly molted composite seta displaying a very distinct annulus (ringed) as a by-product from the invagination during development.

Setae, Setules, and Other Ornamentation setae on the dactylus of maxilliped 3 of the shrimp Palaemon adspersus. The definition we will follow here was put forward by Garm (2004b, 1): “A seta is an elongate projection with a more or less circular base and a continuous lumen. The lumen has a semicircular arrangement of sheath cells basally.” The available data on the ultrastructure of setae provide good support for the internal characteristics—the continuous lumen and the semicircular sheath cells (Fig. 6.1F,G), also called enveloping cells (Alexander et al. 1980, Hallberg et al. 1992, Crouau 1997, Paffenhöfer and Loyd 2000). That the sheath cells seem to be a unifying character for setae indicates that they play important functional roles. They are involved in setal development, and this complex process could possibly provide a more detailed definition. The continuous lumen is also functionally significant since it is closely connected with the sensory properties of setae. Both of these topics are discussed in detail in later sections. It is often problematic to use internal characters because categorization is normally based on light or scanning electron microscopy. The round shape of the basal part of a seta is therefore an important character, and it seems to be very consistent for setae found on all body parts of many groups of crustaceans (see Garm 2004b for review). Still, using this character alone will not suffice since it will not separate unarticulated setae and spines dealt with below. Besides setae, there are other surface structures of the cuticle that we will briefly consider. One group has already been mentioned—the setules. As said above, this is a term widely used for certain outgrowths on setae, but we believe them to be a general feature of the cuticle. They are elongate structures, 10–150 μm long, often inserted into the cuticle in a socket, making them flexible (Fig. 6.1C). They are flattened in cross section and made entirely of cuticle, so they are never innervated and do not contain semicircular sheath cells basally (Fig. 6.1E). Most often they have a serrated edge distally. Such setules are commonly found in the general cuticle throughout the Crustacea, especially on the mouth apparatus and in the foregut, but they have typically been referred to as setae (Halcrow and Bousfield 1987, Holmquist 1989, Martin 1989, Olesen 2001). Another expression often used when describing the cuticle of crustacean is denticles. Like setules, denticles are commonly found on setae but also in the general cuticle. This again stresses that some of the structures generally considered special features of setae are in fact general cuticular characteristics. Denticles are relatively small structures (normally < 30 μm long) and, as the name implies, more or less tooth shaped (Fig. 6.1D). They are unarticulated, made entirely of cuticle, and never innervated. There is some evidence that they are in fact evolutionarily related to setules (Garm 2004b). A common cuticular outgrowth is the spine. This term should be used with care since it can be very hard to tell a true spine from an unarticulated seta. We consider a spine to be an unarticulated extension of the general cuticle. It is hollow, and the lumen is lined with normal epithelial cells; no innervation is present unless the spine carries setae (see, e.g., Martin and Cash-Clark 1995, their fig. 19A,C). If ultrastructural data are not available, then comparison with closely related species should be used to verify that they do not have setae in the same position. While the other types of structures are probably homologies, we find it very likely that spines have arisen several times and represent convergent evolution. Scales, like spines, are cuticular outgrowths, but they are generally wider than long and are not usually hollow (Klepal and Kastner 1980). Most often, scales follow one side of the outline of the polygons often visible on crustacean cuticles when observed with the scanning electron microscope. While it was long known that the crustacean cuticle was often sculptured, the exact details could not be seen until the invention of the scanning electron microscope. Some of the main features are summarized by Meyer-Rochow (1980), Holdich (1984), and Halcrow

171

172

Functional Morphology and Diversity and Bousfield (1987). A terminology of surface sculpturing was proposed by Harris (1979) for insects, but it seems equally applicable to crustaceans. The basic unit of sculpturing seems to be a more or less well-defined polygon, which Hinton (1970) and Duncan (1985) assert represents the surface mani festation of underlying epidermal cells (see also chapter 5). Scales, microspines, micropores, and a large variety of other structures can be found within and along the boundaries of polygons (e.g., Klepal and Kastner 1980, Halcrow and Bousfield 1987). In many other crustaceans and insects, however, the polygon is obliterated by cuticular secretions that form more elaborate sculpture (e.g., Hinton 1970, MeyerRochow 1980).

THE EXTERNAL STRUCTURE OF SETAE As discussed in the preceding section, setae constitute the largest and most diverse group of structures, which is also why providing a general definition is not a trivial task. This diversity is seen between species, but sometimes a single species carries close to the full diversity of seta types. Most of the setae are found on the appendages, and especially the mouthparts are heavily ornamented with setae, and a single segment (= article) of, for example, a maxilliped can display quite a number of setal types (Fig. 6.1B). In the following we will try to deconstruct this diversity and pinpoint some of the important structures that cause the diversity. Structures that share some kind of similarities can be a product of the evolutionary history of setae and thereby be considered homologies, and/or they can be products of shared functionality. As we discuss further below, most of the similarities of setae stem from shared functions, and this will be used to suggest a classification system. First, it is important to recognize that all setae can be seen as having a more or less elongate and round (at least at the base) central part, the shaft, which may or may not have specializations, including different types of outgrowths. The length of the shaft varies from just a few micrometers to several millimeters in large decapods. They are found on all body parts, including internally in the foregut (Altner et al. 1986, Johnston 1999), and serve many different functions. This diversity of function has undoubtedly added to the wide range of external morphology. Some setae are long and slender with no apparent specializations along the setal shaft (Fig. 6.2F,G), whereas others have many types of outgrowths, resembling feathers or pine trees (Fig. 6.2A,B). Still others are thorn shaped or bent and appear as hooks. Despite the diversity, several substructures can be recognized in many setae and can be used to group the setae into different types. If a classification includes too many details, there is a high risk that the designated setal types will be highly specialized and appear only in a very limited number of taxa. Here, we try to avoid this problem and consider only overall structural similarities found in most major crustacean taxa, since this will have the broadest application and interest. One of the prominent substructures concerns how the setae attach to the general cuticle. Three types of attachments are seen: (1) an articulation in the form of a socket, which is drawn into the general cuticle and gives the seta an infracuticular articulation (Fig. 6.3C)—this is by far the most common type of attachment; (2) the socket is extended from the general cuticle, giving the seta a supracuticular articulation (Fig. 6.3B)—this gives the seta great flexibility and is often seen in setae experiencing large drag forces; (3) no articulation is seen, and the general cuticle has a direct transition into the cuticle of the seta (Fig. 6.3A)—as mentioned earlier, the articulation is probably reduced to obtain sturdiness. Another feature concerning the sturdiness is the length:width (L:W) ratio of the shaft, where the width is measured at the base of the seta. The vast majority of setae are slim, with a L:W ratio of >15 (Fig. 6.2), but some setae are more stout and robust, with a L:W ratio 15 – Infra – + – >15 Ter/– Infra – +/– + >15 Ter/ Sub/– Infra + +/– + >15 Ter/– Infra – – – >15 Ter/– Infra/absent – +/– –