Evolution of African Mammals [Reprint 2014 ed.] 9780674431263, 9780674431256

Citation preview

Evolution of African Mammals

EVOLUTION OF A F R I C A N MAMMALS

Edited by Vincent J. Maglio and H. B. S. Cooke Harvard University Press Cambridge, Massachusetts London, England 1978

Copyright © 1978 by the President and Fellows of Harvard College All rights reserved Printed in the United States of America

Library of Congress Cataloging in Publication Data Main entry under title: Evolution of African mammals. Bibliography: p. Includes index. 1. Mammals, Fossils. 2. Paleontology—Africa. 3. Mammals—Evolution. I. Maglio, Vincent J. II. Cooke, Herbert Basil Sutton. QE881.E84 569'.096 77-19318 ISBN 0-674-27075-4

To the many who have contributed to the study of paleontology and prehistory in Africa

Geological research, though it has added numerous species to existing and extinct genera, and has made the intervals between some few groups less wide than they otherwise would have been, yet has done scarcely anything in breaking down the distinction between species, by connecting them together by numerous, fine, intermediate varieties; and this not having been effected, is probably the gravest and most obvious of all the many objections which may be urged against my views . . . For my part . . . I look at the natural geological record, as a history of the world imperfectly kept, and written in a changing dialect; of this history we possess the last volume only . . . Of this Volume, only here and there a short chapter has been preserved; and of each page, only here and there a few lines. Each word of the slowly-changing language, . . . being more or less different in the interrupted succession of chapters, may represent the apparently abruptly changed forms of life, entombed in our consecutive, but widely separated formations. On this view, the difficulties . . . [arising from the poor fossil record] are greatly diminished, or even disappear. —Charles Darwin, 1859

One of the more remarkable effects of the postwar economic boom has been an explosion of intellectual investigation and a multifold increase in published data. This trend has manifested itself in nearly every area of human endeavor, including the quest for knowledge of the earth's historical past, both physical and biological. The study of man has figured prominently in this realm, along with the necessary collateral studies on the ecological setting within which our own species originated and evolved. Thus we have recently witnessed a burgeoning of investigative activities that, because of geographical history, have centered on the African continent. More than twice as many publications on African prehistory have appeared in the last two decades as in the previous two centuries. Any worker who has attempted a synthesis of these voluminous data, whether he is just beginning the study of African prehistory or has been involved for years, has undoubtedly run headlong into a wall of intellectual resistance, a wall built up of countless details of morphology, distribution, and geological setting. It has, in fact, been extremely difficult to synthesize current knowledge of any particular vertebrate group without spending much time digesting the often unpalatable morsels dispersed over an ever-widening primary literature. Thus, few workers have been able to achieve an adequate understanding of the evolutionary events that shaped Africa and the modern world, even with respect to one, or at most a few, isolated groups. For those of us who are currently pursuing such research, it has become increasingly clear that this situation cannot be allowed to persist. Limited access to scientific data perpetuates an undesirable narrowness and makes it ever more difficult to foster the infusion of ideas that can catalyze new insight and broader directions. No area of inquiry can remain closed and yet avoid stagnation. In African prehistory it has been only in the Hominidae, the one group that has received broad treatment in summary fashion, that widespread interest and diverse opinion have been brought to bear on complex evolutionary problems. Such interest is slight or nonexistent in the study of other groups, partially because of their inherently less anthropocentric value, but also because of the restricted and overtechnical presentation of available data. The description of a new species is, of course, a critical part of paleobiological studies, but once in a while we must step back to view our creations from a distance and to attempt broad correlations within the structure of related biological concepts. Such overviews are the only means of assessing the whole and of picking out our

viii

Preface

significant achievements and failures, as well as pinpointing the persistent problems. Even more important is the dissemination of knowledge. No research is significant unto itself; it is significant only to the extent that it relates to man and the world at large. With the accelerating destruction of the natural world we have a responsibility to help preserve what is left. A firm understanding of the origin and evolution of life forms and of the establishment of the ecological balance of a major continental fauna, such as that of Africa, can be significant in achieving this end. From a more practical point of view, numerous field workers are collecting fossil data in Africa, often without full knowledge of where the major problems lie, except, perhaps, for the one or two groups with which they have particular expertise. Data are lost each year simply because we do not know what kinds of evidence are most needed in each group. For example, elephant teeth are always collected, even if fragmentary, because we have learned from past studies that they hold the key to understanding this group. However, it is now abundantly clear that only cranial evidence will eventually resolve the many remaining problems in proboscidean evolution. Yet, unless exceptionally complete, such evidence is rarely gathered. Similarly with hippopotami, the mandibular symphysis has proved to be critical in the assessment of some major lineages and of evolutionary trends, but this part is not likely to be collected unless a tooth or root-bearing ramus is attached. It was for these reasons that we set out to prepare a single volume that would summarize the current state of our knowledge on the origin and evolution of the class Mammalia in Africa. We decided to restrict our efforts to that class, primarily because of the scarcity of information on the lower vertebrate assemblages of the continent. Each of the 27 systematic chapters has been written by one or several of the major authorities who have contributed to the study of the group and who, because of their intimate knowledge of most of the relevant collections, can best evaluate our current state of understanding. Each contributor was requested to speculate where possible on phyletic relationships both among African forms and between these and their Eurasiatic ancestors or descendants. For each of the 15 mammalian orders treated, this volume tries to summarize the valid taxonomic groups as now envisaged, the origin of its various subgroups, its geographic distribution, major phyletic units, and specific evolutionary trends. No attempt has been made to present each group

according to a standard format. This would be not only extremely difficult but undesirable, given the nature of the various groups and the data that are currently available. Thus, for some small or poorly known groups, such as the Palaeomerycidae, Tubulidentata, and Embrithropoda, little more than an account of the fossil forms can be given. For other groups, where the data and stratigraphic distributions are more abundant, such as in the Bovidae, Proboscidea, and Suidae, a more detailed account of their origins and probable evolution is attempted. A number of orders have been subdivided into their component families for more detailed treatment. In addition to the differences in content, the treatment in each chapter differs according to the philosophy of the authors. It was felt that if a rigid organization were attempted, the volume would lose more from the imposition of the editors' biased style and from their deficiencies of personal perception than it would gain from a more predictable, textbook-like structure. Several chapters, such as those on the Hyracoidea, Deinotherioidea, and Equidae, are presented in a rather formal species-by-species treatment with orderly diagnoses, discussions, and critical revisions, whereas others, such as the Bovidae and Insectivora, are presented less formally. Still others, such as the Hippopotamidae, are discussed almost in narrative fashion. The information amassed here derives from two basic sources. Much of it stems from the published literature, where assessments must continually be made as new and more complete evidence is unearthed. But second, and more important, are the new materials and reinterpretations that are discussed here for the first time. These new data stem from the active research of the various authors and include information not otherwise available to the nonspecialist; accordingly, this should provide the reader with a more comprehensive analysis than would be possible elsewhere. Chapter 1 is a general discussion of the modern African mammalian fauna and the varied ecological environments within which it is found. The author also presents some comments on interactions between man and various species of wildlife. Chapter 2 presents a concise review of the major geological deposits of the African Cenozoic, from which the rich faunal records of the continent have been derived, and provides a framework based on available evidence for absolute and relative dating. Each succeeding chapter deals with a particular mammalian group, or in some cases a few mammalian groups; those families with more complete fossil records are treated in separate chapters.

Preface

ix

To round out the volume, the broader patterns of faunal evolution are synthesized in an attempt to cut across taxonomic boundaries and to visualize the interdependence of faunal events in the continent as a whole. Knowledge of any historical phenomenon can never be conclusive, and the degree to which interpretations command our confidence is in direct proportion to the adequacy of the documented record. This is even more true in paleobiology than in other historical subjects, because of the very incomplete nature of the fossil record, even under the best of circumstances. The present volume is not intended as a definitive treatment of the subject. Rather, it is conceived as an interim report on investigations still in progress, investigations that will not be completed by any of the present contributors and, indeed, will never be truly completed. The scope of the volume is broad and its treatment necessarily limited. But it is hoped that the book will fulfill a great need and that it may be useful to the professional who cannot be an

authority on every group, to the student who wishes to gain a wide understanding of a particular group as a basis for further research, to the evolutionary biologist and to the geologist who can deal with the larger implications of such data, and to all those interested in the processes and events that shaped the last great intact faunal province on the earth. This book represents the combined efforts of very many individual investigators and of their colleagues, far too numerous to mention here. To all these people we extend our gratitude. In particular, we wish to thank the contributors for their efforts and for their cooperation in bringing this document to fruition. A special debt is owed to Rosanne Leidy for her invaluable help in preparing the manuscripts for publication. Typing of the final draft for considerable parts of the volume was done by Ms. Jean Olsen, Ms. Evelyn Wolff, and Mrs. L. O'Hearn. Vincent J. Maglio

Η. B. S. Cooke

Contributors

Andrews, Peter Department of Palaeontology British Museum (Natural History) Barnes, Lawrence G. Natural History Museum of Los Angeles County Beden, Michel Laboratoire de Paleontologie des Vertebres et Paleontologie humaine Universite de Poitiers Bigalke, R. C. Department of Nature Conservation University of Stellenbosch Black, Craig C. Carnegie Museum of Natural History Butler, Percy M. Department of Zoology Royal Holloway College University of London Churcher, C. S. Department of Zoology University of Toronto Cooke, Η. B. S. Department of Geology Dalhousie University Coppens, Yves Musee de l'Homme Museum National' d'Histoire Naturelle Coryndon, Shirley C. (deceased) Department of Geology Bristol University Crompton, A. W. Museum of Comparative Zoology Harvard University Delson, Eric Department of Anthropology Herbert H. Lehman College City University of New York Domning, Daryl P. Department of Anatomy College of Medicine Howard University

Gentry, Alan W. Department of Palaeontology British Museum (Natural History)

Mitchell, Edward D. Arctic Biological Station Ste. Anne de Bellevue, Quebec

Hamilton, W. Roger Department of Paleontology British Museum (Natural History)

Patterson, Bryan Museum of Comparative Zoology Harvard University

Harris, John M. Department of Palaeontology Louis Leakey Memorial Institute Nairobi, Kenya

Pilbeam, David R. Department of Anthropology Yale University

Hooijer, Dirk A. Rijksmuseum van Natuurlijke Historie Leiden Howell, F. Clark Department of Anthropology University of California Berkeley Jenkins, Farish Α., Jr. Museum of Comparative Zoology Harvard University Lavocat, Rene Ecole Pratique des Hautes Etudes Institut de Montpellier

Richardson, M. L. Department of Zoology University of Toronto Savage, Robert J. G. Department of Geology Bristol University Simons, E. L. Primate Center Duke University Tanner, Lloyd G. Division of Vertebrate Paleontology State Museum University of Nebraska

Madden, Cary Τ. University of Colorado Museum

Walker, Alan C. Departments of Cell Biology and Anatomy Johns Hopkins University

Maglio, Vincent J. 437 Ashwood Lane Kirkwood, Missouri

Wilkinson, Albert F. Embajada Britanica Quito, Equador

Meyer, Grant E. Raymond M. Alf Museum Webb School of California

1

Present-Day Mammals of Africa

1

R. C. Bigalke

2

Africa: The Physical Setting

17

Η. B. S. Cooke

3

Mesozoic Mammals

46

A. W. Crompton and F. A. Jenkins, Jr.

4

Insectivores and Chiroptera

56

P. M. Butler

5

Rodentia and Lagomorpha

69

R. Lavocat

6

Prosimian Primates

90

A. C. Walker

7

Cercopithecidae and Parapithecidae 100 E. L. Simons and E. Delson

8

Cenozoic Apes

120

E. L. Simons, P. Andrews, and D. R. Pilbeam

9

Ramapithecus

147

E. L. Simons and D. R. Pilbeam

10 Hominidae

154

F. C. Howell

11

Carnivora

249

R. J. G. Savage

12 Pholidota and Tubulidentata

268

B. Patterson

13 Embrithropoda

279

L. G. Tanner

14

Hyracoidea

284

G. E. Meyer

15 Deinotherioidea and Barytherioidea 315 J. M. Harris

16

Moeritherioidea

333

Y. Coppens and M. Beden

17

Proboscidea

336

Y. Coppens, V. J. Maglio, C. T. Madden, and M. Beden

18

Chalicotheriidae

368

25

P. Μ. Butler

C. S. Churcher

19 Rhinocerotidae

371

26

Equidae

27

379

Anthracotheriidae

423

28

Suidae and Tayassuidae

435

29

Η. B. S. Cooke and A. F. Wilkinson

23

Hippopotamidae

30

483

Cervidae and Palaeomerycidae W. R. Hamilton

540

Sirenia

573

Cetacea

582

L. G. Barnes and E. D. Mitchell

Patterns of Faunal Evolution V. J. Maglio

S. C. Coryndon

24

Bovidae

D. P. Domning

C. C. Black

22

536

A. W. Gentry

C. S. Churcher and M. L. Richardson

21

Tragulidae and Camelidae A. W. Gentry

D. A. Hooijer

20

Giraffidae 509

496

Index

621

603

1

1 Present-Day Mammals of Africa R. C. Bigalke

Faunistics It has become almost a commonplace to write of Africa that it has a remarkably rich and diverse mammalian fauna. As it is mainly a tropical region, and tropical faunas in general are larger and more diverse than those of temperate zones, this is to be expected. Tropical Asia and Neotropica also have many mammals. There are approximately 740 species and 52 families in Africa (Bigalke 1972). For Neotropica, although it is somewhat smaller (7 million sq mi as against 11.5 million sq mi), Hershkovitz (1972) gives the surprisingly high figure of 810 species in 50 families. The considerably smaller Oriental Region has approximately 470 species and 40 families (Darlington 1957). By way of a general explanation, Darlington's observation (1957) that tropical faunal diversity results from the presence of both widely distributed and localized groups may be noted. Temperate zones, on the other hand, are populated mainly by widely distributed groups, and localized taxa are unimportant. The question of most significance is to what extent Africa differs from other tropical areas. There are basic and well-known differences in faunal composition between Neotropica on the one hand and the Ethiopian and Oriental regions—the Old World tropics—on the other. These may be simply illustrated by table 1.1. The proportions of bats, "gnawers," and other land mammals in Africa and the Oriental Region are similar, although the last group has rather more species in Africa. Bats make up much the same fraction of all three faunas. But in Neotropica almost half of the species are "gnawers" and other land mammals are poorly represented. The close faunal relationship between Africa and the Oriental region is shown by the fact that 10 families occur only in these two zones. They are lorisids, cercopithecid monkeys, apes, pangolins, bamboo rats, Old World porcupines, viverrids, elephants, rhinoceroses, and chevrotians (porcupines and viverrids each have one species in southern Europe). In a few cases genera and even species are shared. Some of the families do, however, have fossil representatives in European deposits, indicating a wider original distribution. For all that Africa is, in a sense, a typical tropical region and has strong links with tropical Asia, the mammal fauna is distinctive. Its variety is due to the presence of a great mixture of groups of varying Zoogeographie affinities. In addition to the 10 families shared with the Oriental Region, 17 are shared in varying degrees with Eurasia, and in a few cases with other regions. Eight families—shrews, vesper-

Bigalke

Table 1.1

Present-Day Mammals of Africa

Proportions of land mammals in major Zoogeographie regions.

Species of land mammals (except gnawers and bats) Species of gnawers (Lagomorphs and rodents) Species of bats

Africa (after Bigalke 1972)

Oriental region, excluding Philippines (after Darlington 1957)

Neotropica (after Hershkovitz 1972)

330 (44%)

180 (38%)

205 (25%)

237 (32%) 174 (24%)

135 (29%) 154 (33%)

380 (47%) 222 (28%)

tilionid bats, lagomorphs, squirrels, cricetids, canids, mustelids, and cats—are widespread or worldwide. Distinction is enhanced by 15 families (almost a quarter of the total number) and two subfamilies that are endemic or virtually so. Fossil remains of five of these families, Thryonomyidae, Orycteropodidae, Procaviidae, Hippopotamidae, and Giraffidae, are, however, known from Eurasia, but thryonomyids and procaviids are believed to be of African origin (Cooke 1972). The orders, families, and approximate numbers of species are listed in table 1.2. Figures in this table refer to the continent of Africa without the extreme northwest corner, north of the Atlas Mountains. Groups with the greatest number of genera and species are insectivores (Lipotyphla and Menotyphla), bats, primates, rodents, carnivores, and artiodactyls. There are several notable features about the fauna (see also Keast 1972). The many unusual endemic insectivores include aquatic otter shrews, subterranean golden "moles," and the elephant shrews, many of them kangaroolike in form. Catarrhine monkeys are the dominant primates. The Cercopithecidae have no fewer than 47 species, most of them arboreal but some terrestrial. The endemic galagos are of particular interest, some of them only as big as mice. Chimpanzees and gorillas are distinctive great apes. Dominant rodents are Muridae, with about 79 species, and Cricetidae, with about 109 species. While by no means as important as in the Neotropical Region, the rodent group is extremely diverse in form. Ellerman (1940-41) considered that Africa must be considered the present headquarters of the order so far as variation in character goes. Anomaluridae ("scaly tails" or "flying squirrels") have converged with the "flying" sciurids (such as Pteromys) of the Northern Hemisphere. Pedetes is an archaic, monotypic, kangaroolike endemic. CanQ rats (Thryonomyidae) resemble large South American caviomorphs. Bathyergids are numerous and successful subterranean vegetarians, whereas gundis (Ctenodactylidae) are morphologi-

cally and ecologically remarkably similar to hyraxes. Among conventional canids and felids, Lycaon, occupying the wolf niche, and the long-legged, coursing cheetah stand out. Proteles, a small, insectivorous hyenalike animal, is endemic. The robust hyenas are survivors of a group of hunter-scavengers that was once much richer in species. Mustelids are poorly represented, but the Viverridae, with 37 species, have speciated actively and fill most of the small carnivore and omnivore niches. The extremely specialized aardvark Orycteropus and the successful, terrestrial and arboreal, herbivorous dassies or hyraxes are endemics of great interest. Finally, the spectrum of almost a hundred large ungulates, including elephants, rhinoceroses, giraffes, hippopotamuses, pigs, and more species of bovids (78) than are found anywhere else, give the African fauna much of its unique pre-Pleistocene character. As the land of big game par excellence, more than any other continent it has retained something of the richness and strangeness of the mammal fauna that dominated the world before the Pleistocene extinctions took their toll. The complex geological and geomorphological changes, climatic fluctuations, evolutionary developments, and migrations of animal groups, of which the recent mammal fauna is the end product, are the subjects of succeeding chapters. Faunal evolution will not be dealt with here. But the role of topography, climate, and vegetation in preserving ancient, relict forms while also providing isolated habitats in which speciation was stimulated must be appreciated. It is therefore necessary to review the general ecological characteristics of the continent.

General Ecological Features Topographically, Africa can be simply described as an eroded peneplain of largely uplifted, flat land (Keast 1972). Perhaps the most important physical features that affect, or have affected, mammal dis-

Bigalke

Table 1.2 Families and approximate numbers of species and superspecies of contemporary mammals of continental Ethiopian Africa. (After Bigalke 1972.)

Order and family INSECTIVORA Potamogalidae (Otter Shrews) Chrysochloridae (Golden Moles) Erinaceidae (Hedgehogs) Soricidae (Shrews) Macroscelididae (Elephant Shrews) CHIROPTERA Pteropodidae (Fruit Bats) Rhinopomatidae (Mouse-tailed Bats) Emballonuridae (Sheath-tailed Bats) Nycteridae (Hollow-faced Bats) Megadermatidae (Big-eared Bats) Rhinolophidae (Horseshoe Bats) Hipposideridae (Leaf-nosed Bats) Vespertilionidae (Simple-nosed Bats) Molossidae (Mastiff Bats) PRIMATES Lorisidae (Pottos) Galagidae (Galagos) Cercopithecidae (Monkeys) Pongidae (Apes)

Species

Superspecies

3 16 6 56

3 7 3 41

13

7

26

22

2

2

7 11 2 17 14

7 7 2 16 14

64 31

?56 26

2 6 47 3

2 4 20 2

Present-Day

Table 1.2

Mammals

(continued)

Species

Rhizomyidae (Bamboo rats) Muscardinidae (Dormice) Dipodidae (Jerboas) Hystricidae (Porcupines) Thryonomyidae (Cane Rats) Petromyidae (Dassie Rats) Bathyergidae (Mole Rats) Ctenodactylidae (Gundis) Spalacidae (Blind Mole Rats)

2 7 3 5 2 1 13 5 1

2 5 3 3 2 1 8 5 1

CARNIVORA Canidae (Jackals, etc.) Mustelidae (Weasels, etc.) Viverridae (Genets, etc.) Hyaenidae (Hyenas) Protelidae (Aardwolf) Felidae (Cats)

11 7 37 3 1 10

10 6 32 2 1 10

TUBULIDENTATA Orycteropodidae (Aardvark)

1

1

PROBOSCIDEA Elephantidae (Elephants)

1

1

11

4

1 1

1 1

5 2

4 2

3 2 1 2 78

3 2 1 2 67

HYRACOIDEA Procaviidae (Dassies)

3

LAGOMORPHA

PERISSODACTYLA

Leporidae (Hares)

10

8

Equidae (Zebras) Rhinocerotidae (Rhinos)

31 7 1

17 7 1

79

54

14 12 1 33 1 5 3 1

11 11 1 26 1 3 3 1

Manidae (Scaly Anteaters)

RODENTIA Sciuridae (Squirrels) Anomaluridae (Scaly Tails) Pedetidae (Springhare) Muridae Murinae Cricetidae Dendromurinae Otomyinae Cricetinae Gerbillinae Lophiomyinae Cricetomyinae Petromyscinae Microtinae

4

Superspecies

Order and family

SIRENIA Trichechidae (Manatees) Dugongidae (Dugongs)

PHOLIDOTA

of Africa

ARTIODACTYLA Suidae (Pigs) Hippopotamidae (Hippos) Tragulidae (Chevrotain) Giraffidae (Giraffe and Okapi) Bovidae (Antelopes)

Note: Families in italics are endemic. The order Insectivora is retained here although Lipotyphla and Menotyphla (for Macroscelididae) are used in the Smithsonian identification manual, The Mammals of Africa. Similarly, the Potamogalidae are retained as a family in this text although Corbet ( 1 9 7 1 ) places them in the Tenrecidae in the same manual.

4

Bigalke

tribution and speciation are the high, isolated mountains, most of them of volcanic origin, which were formed in East Africa and Ethiopia during the Tertiary, and the great valleys that developed as a result of rift faulting in East Africa, probably mainly toward the end of the Pliocene (Cooke 1972). Being situated across the equator, Africa has a wide range of climates, with almost 80% of its surface area in the tropics and extending far enough beyond them to include temperate zones at both northern and southern ends. It is, however, predominantly a dry continent. Keast (1972) states that only 28% of the surface area has an annual rainfall in excess of 40 in, while 31% receives less than 5 in of rain per year. The corresponding figures for South America, for example, are 68% and 2%. Accordingly the vegetation of much of Africa falls within the broad categories of woodland, savanna, grassland, and low shrub steppe. Rough calculations by Keast (1972) indicate that tropical rain forest covers only 9% of the continent. Figures for the other major vegetation types are woodland and open forest, 31%; savanna, grassland, and steppe, 19%; and desertic areas, 30%. Oversimplifying somewhat and using a wide definition for the hard-worked term "savanna," it can be said that savanna covers about one-half of Africa. Most of the dominant groups of African mammals are savanna forms and more species occupy this biome than any other.

Biotic Zones and Mammal Communities Although savanna areas are important by virtue of their size, forests are ancient environments of great significance, with a fauna of their own. The interspersion of forest, savanna, and arid zones, together with other more minor zones, is responsible for much of the faunal diversity. The principal biotic zones, delimited by Davis (1962) from the distribution of vegetation types and faunal regions, are shown in figure 1.1.

Forest The forested regions may conveniently be divided into the extensive, more or less continuous blocks of tropical lowland forest and the small relic forest patches scattered over mountains and coastal lowlands as far as the southern extremity of the continent. Lowland forest has long been considered the most distinct African biome. Its mammal fauna includes many specialized, ancient forms and must have evolved over a long period in a stable environment. Among the insectivores, the endemic otter shrews—

Present-Day Mammals of Africa

Figure 1.1 The main biotic zones of Africa south of the Sahara. Montane and coastal forests are in black. (After Davis 1962.)

traditionally considered a family, Potamogalidae, but classified as a subfamily of the Tenrecidae by Corbet (1971)—are an old forest group. They are highly adapted aquatic creatures that live in streams and feed on invertebrates and vertebrates. Soricids are well represented, partly because they are adapted to moist environments with dense vegetation. Their predominance in the tropics may also be due in part to their relatively recent immigration from Europe in the late Oligocene. Radiation further southward may have been restricted by competition from well-established endemic golden moles, Chrysochloridae, or, more likely, from elephant shrews, Macroscelididae (see Bigalke, forthcoming). The former is essentially a southern family not well represented in forest. The Macroscelididae are also mainly southern African but there are several forest forms such as the giant Rhynchocyon and a race of Petrodromus tetradactylus (Kingdon 1971). As in other tropical regions, Megachiroptera are prominent members of the African forest fauna. As Kingdon (1971) points out, fruit bats are best represented in forests since they depend on a year-round supply of fruit from trees and shrubs. Some use is also made of nectar, pollen, and flowers and Megaloglossus is a specialized nectar feeder. Myonycteris, Megaloglossus, Hypsignathus, and Epomops are typ-

Bigalke

ical lowland forest genera. There are also a number of forest-dwelling Microchiroptera and Kingdon (1971) has drawn attention to the surprising correspondence between areas of distribution of these insectivorous bats—and of Megachiroptera—and those of other mammals, in spite of their mobility. One finds lowland forest species, relic forms confined to mountain refuges, and so on. Fleming (1975) considers insectivorous bats to be trophically much more important in the tropics than frugivorous and nectar-feeding species. However the vast numbers of colonial fruit bats, such as Eidolon helvum, which roost in some areas, must be of considerable local significance for seed dispersal and perhaps pollination. The Lowland Forest Zone is the center of primate diversity. There are two families of prosimians. Pottos (Lorisidae) are specialized, somewhat slothlike, arboreal omnivores that are confined to this zone, although not only occupying high forest. The second family, Galagidae, is endemic to Africa. Four of the six species of bush babies are forest forms. In contrast to pottos they are agile and quick. Some, such as Galagoides demidovii, are little bigger than mice and their presence may have limited rodent radiation. The Cercopithecidae is the largest of the two families of higher primates in Africa and most species are confined to the forest biome. Colobus monkeys of the subfamily Colobinae are sluggish arboreal leaf eaters. Most Cercopithecinae are small to medium-sized, wholly or partly arboreal monkeys feeding on fruit or leaves. The genus Papio consists of large forms, baboons found both in forest and savanna, and the mandrill and drill, forest apes of the subgenus Mandrillus. Booth (1956) was perhaps the first to draw attention to the nature of niche differences in sympatric forest monkeys. Many of them use distinctly different levels in the canopy although there are also examples of considerable niche overlap. The great apes, Pongidae, like the lorisids and cercopithecids, are restricted to Africa and Asia and are one of the groups that provide evidence for past contact between forests of these regions. This is unlikely to have been later than the Oligocene or early Miocene (Misonne 1963). Since that time African and Oriental forms have diverged greatly. Gorilla gorilla is a herbivore strictly confined to forest while the more omnivorous chimpanzee Pan troglodytes is also found outside the Lowland Forest Zone in woodland and savanna. The pygmy form P. paniscus is a localized forest species. The Pholidota are another of the old, specialized groups shared with the Oriental Region. While there are two partly arboreal forest pangolins, the

Present-Day

Mammals

of Africa

5

order is not wholly confined to this biome. A large form occupies both forest and savanna and the remaining species inhabits savanna. All are ant and termite feeders. The forest biome supports a distinctive rodent fauna, the most singular members of which are the Anomaluridae, scaly tails or "flying squirrels." This old, endemic family of arboreal, squirrellike gliding creatures is essentially confined to high forest. The animals feed on leaves, fruit, and in some cases perhaps insects. Sciuridae are particularly well represented by forms ranging from the mouselike pygmy squirrel Myosciurus pumilio to giant species of Protoxerus, 60 cm long and weighing a kilogram. Other forest genera of note are Heliosciurus, Epixerus, and Funisciurus. Dormice (Muscardinidae) of the genus Graphiurus are present and the tree porcupine Atherurus is an interesting, highly adapted forestdwelling member of the Hystricidae. Murid rodents are, however, poorly represented and there are no really specialized forest species. This is probably because the group only entered Africa in the late Pliocene when many rodent niches were filled by small anomalurids, squirrels, galagos, and cricetids so that there was little stimulus to radiation (Misonne 1963; Kingdon 1971).1 Forest-inhabiting genera include Hybomys, Oenomys, Stochomys, Lophuromys, and Praomys morio. Cricetomys is the most important representative of the Cricetidae but is not confined to forests. Carnivora of the forests are limited in variety and there are few confined to this biome. The only Mustelids are the widespread ratel, Mellivora, and otters of the genera Lutra and Aonyx, the latter being represented in the Lowland Forest Zone by a distinct species, A. congica. Members of the Viverridae are the commonest carnivores. Nandinia binotata, the palm civet, is an ancient, mainly frugivorous forest species, the only member of the subfamily Nandiniinae. The Viverrinae includes several purely forest forms: Poiana (linsang), Osbornictis (aquatic civet), the giant genet Genetta victoriae, and other species of this genus. Mongooses of the genera Bdeogale, Crossarchus, and Herpestes are also confined to this biome. Felis aurata is the only true forest cat, but both the leopard Panthera pardus and F. serval are found in forested habitats. One genus of the Hyracoidea, Dendrohyrax, is an arboreal forest form that also lives among rocks above the tree line on high mountains such as the Ruwenzori (Dorst and Dandelot 1970). The elephant 1 According to Lavocat (this volume) the murids are known from the middle Miocene. —Ed.

6

Bigalke

Loxodonta africana inhabits forests as well as most other habitats and a distinct small forest form, cyclotis, is recognized. Artiodactyls are represented by a fair range of species, but they are generally not plentiful in forested environments because of the limited food supply for ground dwellers. The giant forest hog Hylochoerus is essentially a forest species and the widespread bushpig Potamochoerus also occurs. The chevrotian Hyemoschus is an ancient specialized forest denizen of great interest, the only representative of the Tragulidae. Choeropsis, the pygmy hippopotamus, and Okapia johnstoni are two well-known forest specialists with restricted ranges. The duikers, Cephalophinae, are the dominant bovids. They are clearly of ancient lineage and have speciated to a remarkable extent. About 13 forest species are recognized, ranging from dwarf forms, e.g., Cephalophus monticola, to C. sylvicultor, which weighs about 75 kg. The bovid tribe Neotragini also contains tiny forest species, namely the royal and pygmy antelopes of the genus Neotragus. The bongo Boocercus is a true forest antelope and a small forest form of buffalo, Syncerus, also occurs. Species of wider distribution that tolerate forest conditions include bushbuck, Tragelaphus scriptus, and sitatunga, T. spekei. Terrestrial herbivores are largely dependent on fruits and leaves as grasses are very scarce on the poorly lit forest floor. Since relatively little of the net production is available to forest herbivores, mammal biomass is characteristically low. For example, Collins (1959, quoted by Bourliere 1965) found the biomass of three ungulate species and seven species of primates in a forest in Ghana to be only 72.2 kg per sq km. For the ungulates alone, the figure was only 5 kg per sq km. Holloway (1962, quoted by Bourliere 1965) gives a figure of approximately 420 kg per sq km for a montane forest in Kenya. The uniformity of the forest environment is such that the existence of well-marked regional differences in the fauna is rather surprising. Three centers of endemism are distinguished (Booth 1954; Misonne 1963). Liberia, or the Guinea forest block, is separated from the others by the Dahomey Gap. It has the most distinctive fauna. The Gabon (or Gabon-Cameroon) and Upper Congo (Ituri-Maniema forests) regions have sufficient faunal differences to suggest that they were separated by a nonforested corridor during dry periods in the late Pliocene or early Quaternary. Speciation in the three forest refuges during periods of isolation appears to have been a major factor in the enrichment of the African forest fauna. Relic forests on the central and eastern African

Present-Day

Mammals

of Africa

mountain massifs, in Zambia and Angola, as well as in coast lowlands and along escarpments down the east coast to South Africa, support mammal faunas that are, in general, impoverished versions of those of the lowland forest blocks. There are fewest species in forests most distant from the equator. This simplified statement requires some qualification. Kingdon (1971) shows that relic forests in northern East Africa have a homogeneous fauna very similar to that of the central African forests. Southern and coastal lowland forest relics on the other hand tend to retain older forms and are sufficiently different faunistically to suggest that they have been isolated for a much longer time. The presence of some endemic species on the high mountains of central and eastern Africa also indicates a considerable period of isolation. The faunal similarities between these widely separate mountains—Mount Cameroon, Ruwenzori, Elgon, Kenya, Kilimanjaro, and the Aberdares—show that they were once in contact through the intervening lowlands and the fauna is not a true montane one. In summary, the forest biome is inhabited by ancient mammal groups that have survived in the refuge provided by this stable and favorable environment, and by more recent taxa. Together they make up a typical tropical fauna, rich in species but for the most part poor in numbers of individuals. Bats and arboreal animals—lorises, galagos, cercopithecid monkeys, anomalures, squirrels and other rodents, tree pangolins, tree hyrax, and viverrids— occupy the canopy and utilize its resources. Members of some groups that are mainly ground dwellers such as felids, small rodents, insectivores, and great apes also use the trees to some extent. Food supplies on the ground are limited and there are but small populations of terrestrial herbivores. These are specialized browsers and fruit eaters, solitary or with a simple social organization associated with the difficulties of communication in a closed environment. Most forest mammals are closely tied to this biome but the relic forests carry a surprisingly large array of these forms well beyond the tropics and deep into the Savanna Zone.

Savanna The very distinct difference between African forest and savanna biomes was recognized by Sclater (1896) when he named the former the West African subregion and the latter the East and South African subregion. Within the extensive savanna biome there is a great diversity of vegetation types. From open temperate grasslands such as the South African high veld there are transitions through many

Bigalke

plant associations of grass and woody shrubs and trees to dense woodlands. The boundaries between the savanna and arid zones are not clearcut and differences tend to be mainly a matter of degree of aridity. The vegetation of great areas of the arid zones is simply dry savanna. The high constant humidity and even temperatures of the forest are in strong contrast to large daily and seasonal fluctuations of temperature and humidity in savanna. Droughts are common, the availability of water may play a critical role, and winds have a significance entirely lacking in forest. Structurally the environment is simpler than in forest and the fauna is dominated by mobile terrestrial animals. Grass is a major source of food for herbivores and seasonal changes in its availability and nutritional value affect them materially, in some cases inducing nomadic or migratory movement. Subterranean forms are common but arboreal and volant groups are unimportant. The openness of the vegetation facilitates visual communication and many mammals occur in groups with a complex social organization. Coursing predators, and not only those that stalk their prey, are found. Approximately 40% of African mammal species occur in the savanna biome. They belong to a great variety of families, many of the largest of which are essentially savanna groups. This is true of chrysochlorids, which are most prominent in the south, their probable evolutionary center. They exploit invertebrates, mainly those living underground. Macroscelidids also have a preponderance of species living in the Savanna Zone. They shelter in dense vegetation and among rocks and feed on invertebrates and, in some cases, fruit. There are few African species of hedgehogs and most of them occupy the northern savanna and arid zones, as may be expected of an immigrant group from Europe. Shrews are not primarily a savanna family, but there are many species and individuals, especially in regions receiving more than 600 mm of rain a year (Meester 1962). Most savanna bats are microchiropterans but a few fruit bats, for example species of Epomophorus, have a wide distribution in this biome. They tend to move a great deal in order to obtain the fruit on which they depend. There are few savanna primates but all tend to be numerous and successful. They include two species of Galago, baboons (Papio), the vervet monkey Cercopithecus aethiops, and the patas monkey C. (.Erythrocebus) patas, all species that forage to a greater or lesser extent on the ground. Lagomorpha fill a small herbivore niche successfully in both savanna and arid zones.

Present-Day Mammals of Africa

7

The fossorial Bathyergidae are found almost entirely in the savanna and arid zones, where they exploit subterranean plant storage organs most successfully. No other rodent families are so exclusively associated with savanna, save perhaps the Pedetidae. The spring hare Pedetes, however, is also widespread in the Southwest Arid Zone. It is a burrowing, nocturnal grazing herbivore. Ground squirrels such as Xerus and partly arboreal genera like Paraxerus represent the Sciuridae. Many murids and most of the cricetids occupy the savanna biomes and form the backbone of the small omnivore and herbivore stratum in food chains. There are a great many burrowing species. Arboreal rodents include the tree rat Thallomys, Praomys, and Thamnomys. The Dendromurinae are a specialized subfamily of the Cricetidae, with semiprehensile tails and modified feet; they feed on grass seeds and insects and suspend their nests on plants. Rodents living on the ground surface may inhabit dense grass, e.g., Otomyinae, bush, e.g., Aethomys, or rocky situations, e.g., Acomys. The savanna biome is the environment in which' by far the most carnivores, perissodactyls, and artiodactyls occur. Small carnivores include most of the Viverridae, of which mongooses are especially typical savanna dwellers, preying on small vertebrates and invertebrates. There are also the few African mustelids—the skunklike Ictonyx, Poecilogale (weasels), and the rather wolverinelike ratel Mellivora. Felidae, from the small black-footed cat Felis nigripes to lion, leopard, and cheetah; small canids of the genera Vulpes (foxes), Canis (jackals), and the wolflike hunting dog Lycaon pictus; the termitophagous specialist Proteles (aardwolf); and three species of hyenas are all mainly inhabitants of savanna biomes. They hunt and scavenge in varying proportions, each species tending to take prey of a certain size range but accepting smaller or larger species in case of need and probably often competing with one another for food (Ewer 1973). Foxes and jackals may also take fruit when it is available. Both the large grazing rhinoceros, Ceratotherium, and the smaller, mainly browsing Diceros, as well as the zebras, are savanna forms; in some cases they extend into the arid zone as well. Pigs are represented by the warthog, Phacochoerus, and the bushpig, the latter already mentioned as an inhabitant of forests. The giraffe utilizes both savanna and arid biomes and is most successful in the high browsing niche where the elephant is the only possible minor competitor. The elephant is or was widespread. Most species of Hyracoidea inhabit rocky habitats in this biome.

8

Bigalke

The large herbivore fauna of the savanna is dominated by Bovidae, of which approximately 80% of the species occur in this biome alone or both there and in arid zones. There is only one savanna duiker, Sylvicapra, but many small neotragine antelope occupy specialized microhabitats, for example rocks in the case of the klipspringer Oreotragus and thickets in the case of dik-diks (Madoqua spp.) and the suni Nesotragus. Among the Antilopini, gazelles and the springbok {Antidorcas) select open habitats and tend to be nomadic or migratory; the gerenuk Litocranius lives in bush. The impala (Aepyceros, Aepycerotinae) selects open savanna (Hirst 1975). Buffalo, Syncerus (Bovinae, tribe Bovini), are not habitat specialists but usually inhabit fairly well wooded country near water. Tragelaphine antelopes, which include bushbuck, kudu, nyala (Tragelaphus), and eland (Taurotragus), are in general characteristic of more or less closed savanna environments while reedbucks, Waterhuck, and related forms (Reduncini) are associated with grassy habitats near water. Roan and sable antelopes (Hippotragus, Hippotragini) are found in various savanna vegetation types. The tribe Alcelaphini, consisting of blesbok, bontebok, and tsessebe (Damaliscus), the hartebeests, Alcelaphus, and the wildebeests, Connochaetes, are mainly medium-sized antelope of mesic grassland and savanna or woodland habitats. Many species tend to aggregate into large, irregularly nomadic or migratory groups and are thereby able to seek the best feeding grounds in areas of fluctuating environmental conditions (Estes 1974). Within the savanna some common mammals are widely distributed, but there are sufficient faunal differences to warrant the separation of Northern and Southern Savanna zones. They meet at the "Sclater line" just north of the equator (Davis 1962). Expansion of forests during wet periods is likely to have separated northern from southern savannas and provided opportunities for faunal differentiation. The northern savanna fauna is poorer than that of the south. Kingdon (1971) compares some mammals endemic to the two zones, and occupying broadly comparable habitats, in East Africa. He lists 10 northern as against 23 southern forms, also pointing out that most of the northern species are isolates of widely distributed genera and have related species in the south. These differences, Kingdon suggests, may be due to the relatively simple, homogeneous nature of the northern savanna vegetation belts, which are not divided into the complex mosaic pattern found in southern and eastern Africa. They may thus have provided fewer opportunities for spe-

Present-Day

Mammals

of Africa

ciation and may have been unattractive to potential colonizing species from the south.

Ecological Separation in Savanna Communities Two aspects of the ecology of savanna communities deserve particular mention: the manner in which so many different large herbivore species are sustained, and the high biomasses supported. Keast (1972) has summarized the mechanisms that permit the continent to support so many herbivore species. In brief, and with some modifications to Keast's list, they are the following: 1. Habitat selection. Many species have welldeveloped preferences for particular vegetation formations. Lamprey (1963) has shown these quantitatively for 14 ungulate species in Acacia savanna in Tanganyika (Tanzania). Recently, Ferrar and Walker (1974) and Hirst (1975), using sophisticated mathematical techniques of analysis, have confirmed that most of the ungulates studied are more or less closely associated with certain vegetation types. They are therefore spatially separated from one another. The preferences of some species overlap, however, and fire and rainfall may modify intrinsic preferences. The proportion of total plant biomass contributed by woody plants, especially shrubs, was the most important single site characteristic responsible for habitat separation between the ungulates that Ferrar and Walker studied. The role played by physiological characteristics of animals in influencing the choice of habitat is beginning to be appreciated. The water demands of the Waterhuck, Kobus ellypsiprymnus, for example, are such that it cannot live far from sources of free water (Taylor, Spinage, and Lyman 1969). Oryx, O. gazella, and dorcas gazelle, Gazella dorcas, which inhabit extremely arid areas, have very low water turnovers (Macfarlane and Howard 1972). Morphological (and prbbably also physiological) adaptations of lechwes, Kobus leche and K. megaceros, and the even more specialized sitatunga, Tragelaphus spekei, with elongated hooves, restrict them to floodplains and swamps. The klipspringer, Oreotragus oreotragus, is a specialized rock jumper. Specializations also result in different microhabitats within a major vegetation formation being inhabited by different species. For example, dik-diks CMadoqua spp.) are confined to low, dense thickets in savannas. 2. Feeding at different levels. Giraffe, Giraffa camelopardalis, feed up to 18 ft above the ground, gerenuk, Litocranius walleri, at 4 to 8 ft, and kudu, Tragelaphus strepsiceros, at 3 to 6 ft.

Bigalke

3. Food selection. In place of the traditional categories of browsers and grazers, Hofmann and Stewart (1972) have classified 31 species of East African ruminants as bulk and roughage eaters (i.e., grazers—roughage, fresh grass, and dry region grazers); selectors of juicy, concentrated herbage (tree and shrub eaters, fruit and dicotyledon selectors), and intermediate feeders (some preferring grasses, others dicotyledons). Gwynne and Bell (1968), Field (1968), and Sinclair and Gwynne (1972) have demonstrated that grazing species may not only select different grass species but also varying proportions of plant parts, e.g., wildebeest take a high proportion of leaf, buffalo and topi less, and zebra select mainly stem material. It is generally agreed, however, that there is often a considerable overlap in the diets of animals occupying the same area, and that in periods of food shortage direct competition is probably common. 4. Seasonal patterns of environmental utilization. Vesey-Fitzgerald (1960, 1965) has shown that the floodplains surrounding Lake Rukwa are used by 18 species of herbivorous ungulates—eight of them common—on a seasonally fluctuating basis. Different species use different communities at different times. The heaviest—elephant, hippopotamus, and buffalo—reduce the tall grassland of the seasonal swamps to a short-grass pasture that becomes available to the smaller antelope during the dry season. In Serengeti National Park, an entirely different community, Gwynne and Bell (1968) describe another grazing succession. Zebra and buffalo enter long grass areas first during the dry season, use the stemmy material at the top of the herb layer and so prepare the way for topi and wildebeest, and finally for Thomson's gazelle; these follow successively and feed on the intermediate and lowest levels respectively, taking different components. 5. Duplication of faunas in equivalent vegetation formations. The ecologically similar but widely separated arid areas in southwestern, northeastern, and northern Africa have, to some extent, distinct mammal faunas. These species contribute to the large numbers inhabiting the continent.

Herbivore Biomasses Records of high ungulate biomasses in savanna habitats have attracted much comment. Some early published figures were based on rather scanty data (see Talbot et al. 1965) but recent studies have a firmer foundation. Foster and Coe (1968) believe that the carrying capacity of the Nairobi Park is about 6,300 kg per sq km and state that this standing crop biomass is of the order one would expect in

Present-Day

Mammals

of Africa

9

the acacia savanna habitat of East Africa. Watson, Graham, and Parker (1969) give somewhat lower figures of 4,027 kg per sq km and 3,907 kg per sq km respectively for the Serengeti Park ecosystem and the Loliondo Controlled Area in Tanzania. These data are for mixed populations of small- and medium-sized ungulates: zebra, giraffe, and various bovids such as wildebeest, hartebeest, eland, gazelle, and impala. Field and Laws (1970) report a much higher figure from bushed grassland in the Queen Elizabeth National Park, Uganda. The mean year-round standing crop biomass over a four-year period was 29,490 kg per sq km. Of this 65% was contributed by hippo, 19% by buffalo, and 12% by elephant. A count by Watson and Turner (1965) in the Lake Manyara Park, Tanzania, where elephant and buffalo predominated, also revealed a high biomass of 21,870 kg per sq km. The presence of very heavy species a p parently increases the total biomass considerably. Comparison with domestic animal populations is instructive. Foster and Coe (1968) quote various authors. On tribal grazing land in East African savanna, domestic stock totaled 1,960 to 2,800 kg per sq km, and on managed European ranches, 3,728 to 5,600 kg per sq km. The average of all virgin ranges carrying domestic animals in the western United States is given as 3,448 kg per sq km while Petrides and Swank (1965) write that the best ranges there have a capacity of 4,300 kg per sq km. In the Nairobi Park major predators have a biomass of only 1.4% of the total ungulate biomass. They are estimated to remove 15.5% of the total biomass, about 700 kg per sq km, per annum and this probably represents what man could take in the absence of predators (Foster and Coe 1968). At Loliondo 10% of the known populations could probably be removed without exceeding the sustained yield (Watson, Graham, and Parker 1969).

Arid Biomes In addition to the Sahara Desert there are three distinct dry regions distinguished as the Southwest, Somali, and Sudanese Arid Zones. They are occupied by mammals adapted to heat and drought, some shared but many endemic to each region. This suggests a long history of aridity and periods of isolation during which local speciation led to the development of separate faunas. Conditions are more stringent than in the savanna zones. Rainfall is low and unpredictable and as a result primary productivity is limited. Mammals are faced with serious problems of thermoregulation because of high ambient temperatures and a

10

Bigalke

scarcity of water for evaporative cooling. Many small mammals escape by burrowing and emerge to feed during the cool night hours. There are only a few large species with the physiological adaptations necessary for survival and they must often move over great distances to find food. As in other very dry areas, many mammals are pale in color and reflect radiant energy efficiently. The Southwest Arid Zone is the most distinct and endemic mammals are prominent in the fauna. They include five insectivores, two elephant shrews, Elephantulus vandami and Macroscelides proboscideus, and three golden moles, Eremitalpa granti, and two species of Cryptochloris (Meester 1965, 1971). Ten species of murids and cricetids, almost half of the total fauna of 24 species of small rodents, are endemic or almost endemic (Davis 1962). Gerbils are the most typical mice. The region is the habitat of the highly specialized crevice-dwelling rock rat Petromus typicus (Petromyidae), a hare (Bunolagus), and the gemsbok Oryx gazella. Equus zebra is found outside this zone only along the southern Cape mountain ranges. Another group of species found only here and in the adjoining grassland subzone of the southern savanna are the hedgehog, Erinaceus frontalis; two ground squirrels, Xerus; two mongooses; Felis nigripes; Vulpes chama (silver fox); Otocyon; brown hyena; the extinct Equus quagga; springbok; red hartebeest; and black wildebeest (Meester 1965). While the endemic fauna provides evidence for past isolation, some species and genera are shared with the Somali Arid Zone far to the northeast. Those discontinuously distributed in this way include the spring hare, Pedetes; caracal lynx, Felis caracal; aardwolf, Proteles; bat-eared fox, Otocyon; dik-dik, Madoqua; and Oryx. This interesting discontinuous distribution may be explained by assuming that a "drought corridor" linked southwestern with northeastern Africa during dry periods in the Pleistocene (Balinsky 1962). As Kingdon (1971) points out, it need only have consisted of areas of dry acacia bush and savanna rather than really arid communities. The Somali Arid Zone in the horn of Africa has several interesting endemic mammals. They include a hedgehog, two elephant shrews, a ground squirrel and several gerbils, the peculiar maned rat Lophiomys, a hairless bathyergid, Heterocephalus, a ctenodactylid, a zebra, and several bovids, including the long-necked Litocranius and some gazelles. Some have spread into adjacent savanna areas. There is a fairly close relationship with the Sudanese Arid Zone but this has a much less well defined

Present-Day Mammals of Africa

fauna than the others, being mainly a transitional area between the northern savanna and the Sahara. As a harsh environment with a long history of extreme aridity (Cooke 1963) the Sahara supports a very small fauna. Typical mammals are a few species of hedgehogs and shrews, an elephant shrew, a dormouse, a few murines, about 15 gerbils, some dipodids, and ctenodactylids. Hares, hyraxes, two mustelids and two viverrids, several canids and felids, the wild ass, two large bovids, the addax, Addax nasomaculatus, the oryx, O. dammah, and about five species of gazelles complete the list. While some authors consider the mammals an impoverished Ethiopian fauna, there are regions in the Sahara in which the fauna has been described as transitional to that of the Near East (see Bigalke 1972). There are few data on biomasses in arid areas. However, Bourliere (1965) gives estimates of 0.3 to 190 kg per sq km (0.003 to 1.9 kg per ha) for stony desert and 4 to 17 kg per sq km (0.04 to 0.17 kg per ha) for sandy desert in the Sahara.

Ethiopian Highlands The large isolated block of mountainous country that dominates Ethiopia is an important center of endemism for mammals. The fauna includes unique forms such as the rodent Muriculus; the hamadryas and gelada baboons, Papio hamadryas and P. gelada; the Simenian fox, Canis simensis; a genet, Genetta abyssinica; the mountain nyala, Tragelaphus buxtoni; and the Ethiopian ibex, Capra walie.

Southwest Cape This small zone at the southern end of the continent is famous for the distinctive Cape flora. A mediterranean climate with winter rains characterizes the western part, while to the east, summer rains become progressively more important. Mountains isolate the region except along the west coast, where it merges into the southwest arid biome. The Fynbos (macchia) vegetation is dominated by sclerophyllous shrubs, grasslike Restionaceae, and Cyperaceae and geophytes. Grass itself is unimportant. Soils tend to be poor and acid and there appears to be little food for mammals. The fauna is accordingly poor in species and in individuals and there is much sharing with adjacent parts of the Southwest Arid and Southern Savanna Zones. The region has, however, been poorly collected and much must still be learned. There are not many insectivores—an elephant shrew, about five species of golden moles including one that is endemic, and three shrews. Bats comprise a fruit bat and about nine species of Micro-

Bigalke

chiroptera, of which two vespertilionids are endemic. The small mammal fauna is dominated by Muridae and Cricetidae. Davis (1962) shows that 21 species are recorded. Three of them are endemic— Praomys verreauxi, Acomys subspinosus, and Tatera afra—and six are represented by distinct subspecies or isolated relic populations of savanna species. The remainder simply encroach into the southwest Cape. The porcupine, one or two dormice, and three species of Bathyergidae complete the list of rodents. The dune mole-rat Bathyergus suillus is endemic. The baboon is the only primate and Orycteropus occurs. Carnivores are surprisingly well represented by all families, the Hyaenidae only being extinct. Three canids, three Mustelidae, the aardwolf, two genets, two mongooses, leopard, caracal wild cat, and ? serval are recorded. Originally lion, elephant, black rhinoceros, mountain zebra, and hippopotamus were present. The dassie Procavia capensis is common. Artiodactyls consist mainly of small browsing species, namely Cape grysbok, Raphicerus melanotis, steenbok, grey duiker, klipspringer, and the vaal ribbok Pelea. The extinct bloubok Hippotragus leucophaeus and the rare bontebok Damaliscus d. dorcas are endemic to the region. Eland, red hartebeest, and buffalo were common in historic times.

Mammals of Madagascar Madagascar is almost 250 mi from the African continent and is separated from it by an old, deep channel. Millot (1952) suggests that the island was colonized by a few ancestral African mammals early in its history when the channel was temporarily narrowed. From these a fascinating, typically insular fauna has evolved by adaptive radiation. There are only 57 genera and about 94 species of mammals (including three introduced forms) belonging to the insectivores, primates, rodents, carnivores, and bats. The only ungulates are the bushpig Potamochoerus, believed to have been introduced from the mainland in historic times, and feral deer. A hippopotamus the size of Choeropsis is a common Pleistocene fossil. Two species of Suncus are the only shrews. The Tenrecidae are the most important insectivores. There are two subfamilies, about 13 genera, and 29 species of these unique creatures. Many of them are hedgehoglike but there are fossorial species that resemble moles (Oryzorictes) and small shrewish forms. Madagascar is famous for its lemurs, of which there are three endemic families. The Lemuridae are quite large and most are arboreal and mainly

Present-Day Mammals of Africa

11

herbivorous. A subfamily contains the lesser lemurs, which resemble galagos and squirrels and take insects. Sluggish, leaf-eating, monkeylike species are placed in the family Indridae. The lemursquirrel or aye-aye,Daubentonia, is a peculiar insectivorous creature classified in its own family. The characteristic rodents are an endemic cricetid subfamily, the Nesomyinae. Like lemurs and tenrecs, they vary in form and size and have radiated to fill many niches. The only carnivores are a number of endemic and singular viverrids. Two genera are classified with the Asiatic palm civets (Hemigalinae). Bizarre and poorly known small mongooses of the endemic subfamily Galidiinae, with about seven species, correspond to the Herpestinae of the mainland while the large primitive catlike fossa Cryptoprocta (Cryptoproctinae) is a nocturnal forest dweller. Bats include both Micro- and Megachiroptera of six families. There are about 20 species but few are endemic.

Prehistoric Extinctions The unique richness of the African mammal fauna is due to many factors. Most of the important ones have already been mentioned. They are the size of the continent, the large area within the tropics, the diversity of physiography and vegetation, the replication of arid zones, and the opportunities for faunal exchange with adjacent land masses and for speciation within the continent. How a fauna shaped by these influences has managed to survive virtually intact to modern times is a fascinating topic. It is well known that many large mammals, particularly herbivores, rather mysteriously became extinct in Eurasia and North America toward the end of the Pleistocene, about 12,000 to 10,000 years ago, leaving these areas with greatly impoverished faunas. South America, too, suffered a drastic reduction, but extinction was only moderate in Australia, in both cases during the period 20,000 to 10,000 years B.C. (Keast 1972). Africa is often said to have escaped these massive extinctions and to have a "pre-Pleistocene fauna." It did, however, suffer losses. Cooke (1972) points out that the Suidae and Bovidae in particular were greatly diminished. Martin (1966) writes, "It seems almost unbelievable, but the African plains game presently contain only about 60 per cent of the genera of large mammals encountered in the hand-axe faunas." Klein (1974) has recently drawn attention to significant megafaunal extinctions during the terminal Pleistocene in the southwest Cape. Simpson (1966, quoted in Keast 1972) does not, however, re-

12

Bigalke

gard extinction at this time in Africa as significantly greater than what might be expected from a normal turnover rate. It certainly seems moderate by comparison with other regions. How is the persistence of the African large mammal fauna to be explained? To summarize current views, Axelrod (1967) sees it as an expression of a continuing equable climate, stemming from the continent's transequatorial position. Martin (1966) has argued in favor of "overkill" by early man with newly developed weapons, the African fauna suffering less than that of other regions because it had evolved in the presence of man. Leakey (1966) advances cogent arguments against this view, asking among other things why more numerous hunters living after the extinction phase should have had such a small effect on the surviving fauna. While the causes of extinction have yet to be determined and their small impact on the African fauna to be adequately explained, I agree with Klein (1974) that "the coincidence between extinctions and the relatively rapid and dramatic climatic change that characterized the end of the Pleistocene is too strong to ignore."

Recent Faunal History From the end of the Pleistocene until comparatively recently the fauna appears to have suffered no major changes in composition, distribution, and abundance. Human influence first became marked in North Africa. The Egyptians probably captured and tamed elephants in early dynastic times (Carrington 1958) as well as domesticating geese, mongooses, the cat, and probably also gazelles, addax, oryx, and others (Zeuner 1963, Bigalke 1964). Their artists already used ivory in about 6000 B.C. (Sikes 1971). Reckless hunting by Roman colonists, capture for military use and for entertainment in Rome, and the gradual development of the Mediterranean coastline led to the extermination of elephants north of the Sahara. It seems likely that this process was completed in the first few centuries after the birth of Christ. Other large mammals were also gradually eliminated, the lion, for example, disappearing in the nineteenth century (Dorst 1970). Commercial elephant hunting was a feature of the African scene for centuries before the rise of European influence. The slave and ivory trades were closely linked and Arab traders distributed muzzleloaders, thereby increasing the drain on elephant populations considerably. Large amounts of ivory were used in Europe for hundreds of years (Sikes 1971).

Present-Day Mammals of Africa

European settlers at the southern end of the continent decimated game herds during the eighteenth and nineteenth centuries and exterminated the bluebuck, Hippotragus leucophaeus, and also the quagga, Equus quagga. Commercial hunting and hunting for sport, both in southern Africa and farther north, became more efficient as firearms improved during the nineteenth century. Coupled with improved communications, the development of the continent, and increasing populations, hunting took a heavy toll of the teeming herds of wild animals so frequently described in the literature of the day. Elephants in particular had seriously declined virtually everywhere by the late nineteenth century when colonial administrations began to introduce firm control measures. The rinderpest panzootic was another important influence. It killed millions of cattle and probably many more millions of wild ungulates throughout the continent at the turn of the century and is responsible for major current anomalies in the distribution of wildlife (De Vos and Lambrechts 1971). A t the beginning of the present century the ranges of many large mammals had been reduced and remaining populations had declined. Sidney (1965), Van der Merwe (1962), Dorst (1970), and others have written on this process. However, many parts of Africa south of the Sahara were still rich in wildlife until the Second World War and game can even now be found in large numbers outside reserves in countries such as Chad, Sudan, Congo Republic, Tanzania, Kenya, Uganda, Zambia, Angola, Rhodesia, and Botswana. Probably the first game reserve to be proclaimed in Africa was the Pongola Reserve in the Transvaal, established in 1894 (Bigalke 1966). This was followed by Hluhluwe in Natal in 1897, the Sabi Game Reserve, Transvaal, in 1898 (subsequently, in 1926, it became the Kruger National Park), and, shortly after 1900, by various sanctuaries in East Africa. In 1925 the first African national park, the Albert, was created in the Congo and was the first step in a series that provided the eastern Congo with the best parks on the continent (Dorst 1970). These were managed as "total reserves." Visitors were allowed only in a limited area of the Albert Park and scientific research was actively pursued in all of them. Many other parks and reserves followed, most of the major new ones being created after the Second World War. Today there are few countries in Africa without nature conservation areas of some kind (see Guggisberg 1970 for list). Forest reserves have also played a most important role as sanctuaries for fauna, although they are often overlooked.

Bigalke

Utilization of Wild Mammals Hunting Peoples For all but a fraction of the million years and more during which man and his immediate predecessors have existed, hunting, fishing, and food gathering— the " 'robber' economy of savagery" (Allan 1965)— have been his means of survival. In Africa there are still peoples who live largely or entirely by this means. The Bushmen of the Kalahari have remained surprisingly numerous, most of them living in Botswana and southwestern Africa. The Pygmies of the equatorial forests may be as common. In East Africa there are small numbers of Dorobo and other minor tribes who still live by hunting and honey gathering. The economies of such peoples are finely balanced. They require big areas of land—perhaps 10 sq mi per person or more—but, given these, they can live on their production indefinitely, barring climatic catastrophe. There is evidence for deliberate limitation of their own populations in accordance with the resources available (Allan 1965). Allan is probably right in his conclusion that the hunters are likely to become extinct since their cultural patterns cannot be preserved without preserving also their isolation and their way of life by reserving vast areas in an increasingly land-hungry world, human "game reserves" that support tiny populations condemned by their technology to remain static or to starve. Perhaps such reserves will remain—there is every reason to urge that they should.

Cultivators Cultivating peoples occupy all the sufficiently humid parts of Africa. Practically all of them supplement their diet by gathering wild plants, by fishing, or by hunting. Indeed, in many systems, a significant part of village subsistence is obtained by a full utilization of the surrounding bush, woodland or forest (Allan 1965). In West Africa much game is taken by "professional hunters" who live by selling meat on village and city markets. The smaller mammals of the forests are the most important sources of this "bush meat"; cane rats (Thryonomys), the giant rat, (Cricetomys), monkeys, bats, duikers, royal antelope (Neotragus), and others are commonly taken. Asibey (1971) estimates the annual yield from wild animals in Ghana at over 8,000 tons, valued at more than US$7,000,000. In Nigeria bush meat consumed was valued atfN10.2 million in 1 9 6 5 - 6 6 (Charter 1971). Virtually everywhere this resource is utilized without control and, in the face of habitat destruction accompanying increased cultivation and a ris-

Present-Day Mammals of Africa

13

ing human population, it is unlikely to survive much longer. If, however, an elementary licensing system such as that recently introduced in Botswana (Child 1970) could be instituted and simple management applied while reliable methods of determining yields are worked out, a valuable source of the protein that most cultivators lack could be preserved (Bigalke 1975).

Pastoralists People who live largely or entirely from their domestic animals occupy vast areas of Africa, including most of the north, parts of East Africa, and many other regions down to the southwestern extremity. Some rely also on agriculture to an extent. Many, for example the Masai, do not hunt to any significant degree for subsistence, and it is probably true that most African game survives in regions occupied by pastoralists. While it is pleasing that this is so, the future of the remaining large assemblages of game outside parks and reserves is in jeopardy for this very reason. Schaefer-Kehnert and Brown (1975) have drawn attention to the implications for wildlife of the almost universal population increase among pastoral peoples and their flocks and herds, a wellknown problem. Serious overgrazing has become the rule and governments are in general extremely loath to institute the controls needed to solve the problem. Direct competition between wild and domestic animals in overgrazed areas is likely to become more severe, to the detriment of game. Furthermore, undeveloped areas still rich in game are likely to be developed for pastoralists because this provides the easiest, if temporary, way out for governments unwilling or unable to be tough.

Game Farming The large savanna ungulate biomasses, already mentioned, were first commented on in the 1950s by scientists such as Pearsall, Fraser Darling, Harrison Matthews, and Worthington, who were concerned about the future of African wildlife in states about to become politically independent (Bigalke 1966). The contrast between overgrazed tribal lands and game areas carrying higher biomasses, but in much better condition, led to support for the idea of utilizing wild animal populations for food. No attempt will be made to review the considerable body of literature on the subject. Talbot et al. (1965) summarized available information almost a decade ago. Parker and Graham (1971) have put forward some trenchant criticisms of the popular view that traditional sources of protein are in short

14

Bigalke

supply and that protein starvation could be alleviated in Africa by using the more efficient game populations instead of domestic animals. They believe the case that game is "better" than domestic stock in some areas to be unproven but concede that the diversity of African herbivores might be more productive of man's requirements than domestic species used exclusively. However, unless pastoral tribesmen change their attitudes under the pressure of necessity, they will continue to use domestic livestock even if the unowned and unownable wild animals are more efficient and productive. Tribal game farming is highly unlikely to take root. Organized game cropping, mainly of the biggest species, has been a means of producing considerable quantities of edible, saleable meat in countries such as Uganda and Zambia and may still be so (there are no recent records). Game ranching on a limited scale has proved feasible and profitable in East Africa and Rhodesia and is now quite widespread in South Africa. In many cases, and especially in South Africa, only one or a few wild ungulate species are run in association with domestic animals. Technological problems have been reviewed by Bigalke (1975). In spite of difficulties of grazing management, disturbance caused by sustained harvesting, disease conditions that preclude the marketing or export of fresh meat, and other disadvantages, some wild ungulates will probably be put to greater use in providing meat in the future, at least regionally and in association with domestic species. The greatest promise for productive game ranching per se seems to lie in large areas of savanna with mixed species populations (perhaps including domestic animals) utilized for sport hunting, trophies, and meat production. Whether it will become important probably depends largely on social and political factors rather than on ecological considerations. The ecological and economic prospects are bright enough to warrant the adoption of ranching with game as a respectable form of land use in undeveloped areas. If it is to succeed, more research is essential.

Tourism Tourism based on national parks and reserves and on hunting has become an important industry in many parts of Africa. In 1968 Kenya was reckoned to have earned $30 million from overseas tourists (Thresher 1972) and the figure has since increased. Hunting safaris were worth £1.25 million and photographic tours £1.4 million in the three East African countries during 1966 (Clarke and Mitchell 1968). Child (1970) reports that tourism based on wildlife is worth R1 million per annum in Botswana, a poor country.

Present-Day Mammals of Africa

No doubt more complete and up-to-date statistics can be found. The point at issue is that wild mammals have developed an economic value of some magnitude that is now appreciated in most underdeveloped countries. This is likely to favor the retention of existing parks at least and may help to ensure that more are established. To maintain them, sophisticated scientific management will be needed.

Conclusion Only in Africa can man still see something of the fantastic mammal fauna that populated the earth in the distant past. For centuries the fauna survived untouched and unaltered. In the past 200 years it has suffered decimation. Recently the rate of change has accelerated. Some species—fortunately only a few—have disappeared, while the ranges of most larger forms have been reduced by habitat loss, disease, or hunting. In order to conserve the remaining mammals for their cultural, aesthetic, and future survival value to man, social changes of great magnitude will be called for. They will entail effort in education and in increasing scientific knowledge upon which conservation and management can be based. The review of African mammal evolution in this volume will serve as a valuable reference work of basic importance to the task. References Allan, W. 1965. The African husbandman. Edinburgh: Oliver and Boyd. Asibey, E. O. A. 1971. The present status of wildlife conservation in Ghana. I.V.C.N. Publ. n.s. 22:15-22. Axelrod, D. I. 1967. Quaternary extinctions of large mammals. Univ. Calif. Publ. Geol. Sei. 74:1-42. Balinsky, Β. I. 1962. Patterns of animal distribution on the African continent. Α«n. CapeProv. Mus. 2:299-310. Bigalke, R. 1966. South Africa's first game reserve. Fauna and Flora 17:13-18. Bigalke, R. C. 1964. Can Africa produce new domestic animals? Neu; Scientist 21(374):141-145. 1966. Some thoughts on game farming. Proc. Grassld. Soc. S. Afr. 1:95-102. 1972. The contemporary mammal fauna of Africa. In A. Keast, F. C. Erk, and B. Glass, eds. Evolution, mammals and southern continents. Albany: State University of New York Press. 1975. Technological problems associated with the utilisation of terrestrial wild animals. Proc. Ill World Conference on Animal Production. Sydney: Sydney University Press. Forthcoming. The biogeography and ecology of mammals in southern Africa. In M. J . A. Werger, ed. Biogeography and ecology of southern Africa. Amsterdam: Junk.

Bigalke

Booth, A. H. 1954. The Dahomey Gap and the mammalian fauna of the West African forests. Re v. Zool. Bot. Afr. 50 (3-4):305-314. 1956. The distribution of primates in the Gold Coast. J. W. Afr. Sei. Assoc. 2:122-133. Bourliere, F. 1965. Densities and biomasses of some ungulate populations in eastern Congo and Rwanda, with notes on population structure and lion/ungulate ratios. Zool Afr. 1(1): 199-208. Carrington, R. 1958. Elephants. London: Chatto and Windus. Charter, J. R. 1971. Nigeria's wildlife: a forgotten national asset. I.U.C.N. Publ. n.s. 22:37. Child, G. N. 1970. Wildlife utilization and management in Botswana. Biol. Cons. 3(l):18-22. Clarke, R., and Mitchell, F. 1968. The economic value of hunting and outfitting in East Africa. E. Afr. Agric. For. J. 32(special issue):89-97. Cooke, Η. B. S. 1963. Pleistocene mammal faunas of Africa, with particular reference to southern Africa. In F. C. Howell, and F. Bourliere, eds. African ecology and human evolution. New York: Viking Fund Publications in Anthropology No. 36. 1972. The fossil mammal fauna of Africa. In A. Keast, F. C. Erk, and B. Glass, eds. Evolution, mammals, and southern continents. Albany: State University of New York Press. Corbet, G. B. 1971. The mammals of Africa: an identification manual 1.2 sub family Potamogalinae. Washington: Smithsonian Institution Press. Darlington, P. G. 1957. Zoogeography: the geographical distribution of animals. New York: John Wiley. Davis, D. H. S. 1962. Distribution patterns of South African Muridae, with notes on some of their fossil antecedents. Ann. Cape Prov. Mus. 2:56-76. De Vos, V., and Lambrechts, Μ. C. 1971. Emerging aspects of wildlife diseases in southern Africa. Proc. Symp. Nat. Cons, as a form of Land Use: 97-109. Dorst, J. 1970. Before nature dies. London: Collins. Dorst, J., and Dandelot, P. 1970. A field guide to the larger mammals of Africa. London: Collins. Ellerman, J. R. 1940-41. The families and genera of living rodents, 2 vols. London: British Museum (Natural History). Estes, R. D. 1974. Social organisation of the African Bovidae. In V. Geist and F. Walther, eds. The behavior of ungulates and its relation to management. U.U.C.N. Publ. 24(l):166-205. Ewer, R. F. 1973. The carnivores. London: Weidenfeld and Nicolson. Ferrar, Α. Α., and Walker, Β. H. 1974. An analysis of herbivore/habitat relationships in Kyle National Park, Rhodesia. J. S. Afr. Wildl. Assoc. 4:137-147. Field, C. R. 1968. A comparative study of the food habits of some wild ungulates in the Queen Elizabeth National Park, Uganda. Symp. Zool. Soc. Lond. 21:135-151. Field, C. R., and Laws, R. M. 1970. Studies on ungulate populations in the Queen Elizabeth National Park, Uganda. J. Appl. Ecol. 7:273-294. Fleming, Τ. H. 1975. The role of small mammals in tropi-

Present-Day Mammals of Africa

15

cal ecosystems. In F. B. Golley, K. Petrusewicz, and L. Repzkawski, eds. Small mammals; their productivity and population dynamics. Cambridge: Cambridge University Press. Foster, J. B., and Coe, M. J. 1968. The biomass of game animals in the Nairobi National Park, 1960-66. J. Zool. Lond. 155:413-425. Gwynne, M. D., and Bell, R. Η. V. 1968. Selection of vegetation components by grazing ungulates in the Serengeti National Park. Nature 220:390-393. Guggisberg, C. A. W. 1970. Man and wildlife. London. Evans. Hershkovitz, P. 1972. The recent mammals of the neotropical region. In A. Keast, F. C. Erk, and B. Glass, eds. Evolution, mammals, and southern continents. Albany: State University of New York Press. Hirst, S. M. 1975. Ungulate habitat relationships in a South African woodland!savanna ecosystem. Wildlife Monographs 44. Washington, D.C.: The Wildlife Society. Hofmann, R. R., and Stewart, D. R. M. 1972. Grazer or browser: a classification based on the stomach-structure and feeding habits of East African ruminants. Mammalia 36(2):226-240. Keast, A. 1972. Comparison of contemporary mammal faunas of southern continents. In A. Keast, F. C. Erk, and B. Glass, eds. Evolution, mammals, and southern continents. Albany: State University of New York Press. Kingdon, J. 1971. East african mammals, vol. 1. London; Academic Press. Klein, R. G. 1974. A provisional statement on terminal Pleistocene megafaunal extinctions in the Cape Biotic Zone (Southern Cape Province, South Africa). S. Afr. Archaeol. Soc. Goodwin Series 2:39-45. Lamprey, H. F. 1963. Ecological separation of the large mammal species in the Tarangire Game Reserve, Tanganyika. E. Afr. Wildlife 1:63-92. Leakey, L. S. B. 1966. Africa and Pleistocene overkill? Nature 212:1615-1616. MacFarlane, W. V., and Howard, B. 1972. Comparative water and energy economy of wild and domestic mammals. Symp. Zool. Soc. Lond. (1972) No. 31:261-296. Martin, P. S. 1966. Africa and Pleistocene overkill. Nature 212:339-342. Meester, J. 1962. The distribution of Crocidura Wagler in southern Africa. Ann. Cape Prov. Mus. 2:77-84. 1965. The origins of the southern African mammal fauna. Zool. Afr. 1:87-95. 1971. The mammals of Africa: an identification manual. 1.3 Family Chrysochloridae. Washington: Smithsonian Institution Press. Millot, J. 1952. La faune malagache et le mythe gondwanien. Mem. Inst. Sei. Madagascar, ser. A, VII(l):l-36. Misonne, X. 1963. Les Rongeurs du Ruwenzori et des regions voisines. Exploration du Pare National Albert (Deuxieme Serie), Fase. 14. Bruxelles: Institut des Pares Nationaux du Congo et du Rwanda. Parker, I. S. C., and Graham, A. D. 1971. The ecological and economic basis for game ranching in Africa. In E. Duffey, and A. S. Watt, eds. The scientific management

16

Bigalke

of animal and plant communities for conservation. 11th Symp. Brit. Ecol. Soc., Oxford: Blackwell. Petrides, G. Α., and Swank, W. G. 1965. Population densities and the range-carrying capacity for large mammals in Queen Elizabeth National Park, Uganda. Zool. Afr. l(l):209-226. Schaefer-Kehnert, W., and Brown, L. H. 1975. Economic and social aspects of animal production in relation to conservation and recreation. Proc. Ill World Conference on Animal Production. Sydney: Sydney University Press. Sclater, W. L. 1896. The geography of mammals. IV. The Ethiopian region. Geogr. J. 7:282-296. Sidney, J. 1965. The past and present distribution of some African ungulates. Trans. Zool. Soc. Lond. 30:1-396. Sikes, S. K. 1971. The natural history of the African elephant. London: Weidenfeld and Nicolson. Simpson, G. G. 1966. Mammalian evolution on the southern continents. Neues. Jb. Geol. Palaeont. Abh. B, 125: 1-18.

Sinclair, A. R. E., and Gwynne, M. D. 1972. Food selection and competition in the East African buffalo (Syncerus caffer Sparrman). E. Afr. Wildlife 10:77-89. Talbot, L. M.; Payne, W. J. Α.; Ledger, H. P.; Verdcourt,

Present-Day Mammals of Africa

L. D.; and Talbot, Μ. H. 1965. The meat production potential of wild animals in Africa. Farnham: Royal Commonwealth Agric. Bureau, Tech. Commun. No. 16. Taylor, C. R.; Spinage, C. Α.; and Lyman, C. P. 1969. Water relations of the Waterhuck, an East African antelope. Am. J. Physiol. 217(2):630-634. Thresher, P. 1972. African national parks and tourism— an interlinked future. Biol. Cons. 4(4):279-284. Van der Merwe, N. J. 1962. The position of nature conservation in South Africa. Koedoe 5:1-122. Vesey-Fitzgerald, D. F. 1960. Grazing succession among East African game animals. J. Mamm. 41:161-172. 1965. The utilisation of natural pastures by wild animals in the Rukwa Valley, Tanganyika. E. Afr. Wildlife 3:38-48. Watson, R. M.; Graham, A. D.; and Parker, I. S. C. 1969. A census of the large mammals of Loliondo controlled area, northern Tanzania. E. Afr. Wildlife 7:43-59. Watson, R. M., and Turner, Μ. I. M. 1965. A count of the large mammals of the Lake Manyara National Park: results and discussion. E. Afr. Wildlife 3:95-98. Zeuner, F. E. 1963. History of domesticated animals. London: Hutchinson.

17

2 Africa: The Physical Setting Η. B. S. Cooke

The Present Environment The great continent of Africa, one-fifth of the land area of the earth, lies astride the equator and extends to very similar latitudes in the north (37°N) and the south (35°S), although the northern portion is twice as large as the southern part. It is a continent of contrasts, both in physical features and climate. Several mountain peaks in the eastern equatorial region rise to elevations greater than 5,000 m above sea level and a few carry permanent glaciers; in Egypt and in the Afar area of Ethiopia, on the other hand, there are depressions whose floors lie more than 100 m below sea level. The tropical forests of western equatorial Africa receive over 1,500 mm of rain a year, with a top figure of 10,000 mm in the mountains of Cameroon; but not far to the north the great Sahara Desert spans the continent from the Atlantic to the Red Sea and continues on into Arabia and the Middle East. Today this desert forms a significant barrier between the Mediterranean coastal region and the region to the south, conveniently dividing the continent into North Africa and sub-Saharan Africa. From a physiographic viewpoint, Africa consists very largely of an elevated interior plateau, generally highest in the east and lowest in the west. The upland is disrupted by several important basins that have received large quantities of sediment derived from the uplifted rims. In sub-Saharan Africa, particularly in the south and on the east side, the elevated interior is commonly separated by a steep slope, or escarpment, from a relatively narrow coastal plain. Some of the desert bordering the Mediterranean is a broad, relatively low-lying area, reflecting its Mesozoic and Cenozoic marine history. Folded mountain chains, so usual in other parts of the world, are confined to the Atlas ranges in the northwest and the Cape ranges at the southern extremity of the continent. The Ethiopian massif and the East African plateau are surmounted by extensive lava flows and volcanic piles and these regions are also cut by the tectonic trenches of the Rift Valley System. The Red Sea Rift serves to separate Africa from Arabia, although the latter has been an integral part of the African block through most of its geological history. Related fracture systems and minor rifting exist in West Africa. Figure 2.1 shows the general relief and major morphological elements of the continent. The varied climates of Africa are reflected in the vegetation patterns (figure 2.2). Environments range from hot and arid in the Sahara and Namib deserts to rainy tropical forests and, because of altitude, to cool mountain communities, including tun-

Cooke

18

Africa: The Physical

LIBYAN

Setting

BASIN

BASIN •CciCHAD B A S I N

CONGO BASI Ν

CUBAΝΟΌ KALAHARI

Figure 2.1

Relief and major morphological elements of Africa, Madagascar, and part of Arabia.

dra in some of the highest mountains. The rainfall is generally highest in summer but the southwest Cape and the northern fringe of the continent receive most of their rain in the winter. These two areas possess a Mediterranean type of vegetation, with low shrubs, evergreen bushes, and belts of forest. In North Africa olives are common but in the

Cape their place is taken by proteas. Both in the Atlas and in the Cape the high ranges prevent the rain from penetrating far into the continent, so that the Mediterranean macchia gives way to semidesert steppe, which borders the more arid desert areas. In the semidesert areas xerophytic shrubs are characteristic, together with tufts of Aristida grass form-

Cooke

19

Africa: The Physical Setting

ο ο «tr

• •• ..»It

No vegetation, rock desert

Desert grasses and shrubs

Semidesert xerophytic shrub steppe

Grassland

Savanna and/or scrub wood

»wn'V» imäB·

Open woodland Mediterranean scrub with Citrus,

Wmik

Olive, Agave

etc.

Undifferentiated river valley vegetation Broadleaf tropical rain forest

Undifferentiated

mountain communities

km 600

Figure 2.2

m

i

General distribution of the main vegetation types in Africa and Madagascar.

ing an incomplete ground cover; there are some thorn trees, especially along the seasonal watercourses. This vegetation becomes scantier as the true desert belt is approached, being limited to occasional shrubs or thorn trees and tough desert grasses; in large areas of the desert, especially the

rock desert and the "sand seas," vegetation of any kind is virtually lacking. A t the moister fringe of the semidesert belts the xerophytic shrubs disappear when the rainfall increases to about 250 mm a year, their decline being accompanied by an increase in the grass until it

20

Cooke

dominates the landscape as dry grassland or steppe, which may be treeless. However, the Aristida grass of the semidesert is replaced by other grasses, notably Eragrostis in the south and Cenchrus and Chrysopogon in the north. The cover remains incomplete and bare soil or sand is evident between the clumps of grass. In places, thorny Acacia and Comiphora form thickets or patches of woodland, or trees may be widely scattered over the landscape. As rainfall increases the grassland gives place to parkland with thorn trees, thorny bushes, and some deciduous trees scattered about or forming a light shady canopy above the grass. Most characteristic of Africa is the extensive savanna and open woodland environment that surrounds the Tropical Forest Zone like a giant horseshoe. It is the typical "game reserve" landscape of Africa, usually a mixture of grass and trees but with a substantial range of variation according to changes in rainfall, seasonality, exposure, soil characteristics, and other factors, so that woodland and savanna, or even grass steppe, may occur within a small area. The grass is typically Hyparrhenia but may be Themeda or Eragrostis. Where the rainfall is below 750 mm, the trees or bushes are mainly thornbearing types (principally Acacia), but broad-leaved deciduous trees also occur and it is the latter that are most abundant where the rainfall approaches 1,000 mm. The moister woodland areas are dominated by a few genera of trees, notably Isoberlinia, Brachystegia, and Julbernardia. Many authorities believe that the climax vegetation of the savanna was originally richer in closed woodland, with relatively little grass, and that repeated fires have brought about the present situation. Within the horseshoe of savanna and woodland lies the tropical forest, largely evergreen but with varying proportions of deciduous trees; the canopy is usually complete and layered, with the trees of the upper level as much as 50 to 60 m high. Toward the margins of the forest proper, the proportion of deciduous trees increases and forest-savanna mosaic forms a transitional zone. The typical tropical forest is essentially confined to low and moderate altitudes, but two other types of evergreen forest also occur. One of these is the rather limited evergreen forest of the Knysna area of the southern Cape, where the winter and summer rainfall regimes overlap. The other is the evergreen forest of the mountains, particularly in the East African and Ethiopian areas, where it normally lies higher than about 1,300 m above sea level. This montane forest differs from the lowland forest in the genera of trees represented and in the lower height of the tallest elements, but both the montane and Knysna forests

Africa: The Physical Setting

contain a very similar spectrum of vegetation types. Forest of this kind, commonly termed "gallery forest," is found in many protected areas at lower altitudes and also extends along many watercourses in regions where the natural rainfall is inadequate to support it. These long ribbons of forest are an important feature of the drier landscapes.

Geological Development With the exception of the Atlas and Cape ranges, Africa lacks the fold mountains of the other continents and has acquired its present form as a result of progressive uplift, gentle warping, volcanic activity, and faulting. Much of the exposed surface consists of Precambrian rocks or of Precambrian rocks hidden by a fairly thin veneer of Phanerozoic sediments (figure 2.3). The Precambrian "basement" consists of extensive areas of folded schists, gneisses, and granite rocks but also comprises thick sequences of sedimentary rocks that have suffered astonishingly little metamorphism and are in many cases gently folded or even still essentially horizontal. Nevertheless, the basement geology is complex and has involved several periods of strong deformation that are recognizable by belts within which radiometric ages tend to be fairly similar. Stabilization of the basement seems to have been progressive, with certain core areas having become rigid more than 2,500 m.y. ago; they are surrounded by younger orogenic belts. The last major metamorphism took place in the final Precambrian or earliest Cambrian (600 to 500 m.y. ago) and in parts of the continent there is no significant break between the Precambrian and the earliest Paleozoic sediments. According to current views, Africa in the early Paleozoic was part of the southern supercontinent of Gondwanaland and both paleomagnetic evidence and the presence of glacial deposits in the western Sahara show that this was then the location of the south pole of the earth. By the Carboniferous the pole was close to the southern tip of Africa and extensive glaciation ushered in the lengthy cycle of deposition of the Karroo System. In the type area in South Africa this unit reaches a maximum thickness of 7,000 m and was deposited in a slowly subsiding basin. Overlying the glacial Dwyka Series are dark shales and grey sandstones, often rich in plant remains of the southern Glossopteris flora and also containing extensive coal deposits. By mid-Permian times slow equatorward movement of Africa had led to warming of the climate so that the Beaufort Series comprises colorful shales and yellowish sandstones indicative of deposition in an environment subject to periodic drying out and flooding. Fossil re-

Cooke

Africa:

The Physical

21

Setting

Precambrian shield Paleozoic and Mesozoic rocks

Tertiary and Quaternary sediments

t'l'd

Mesozoic and Tertiary volcanics

Dunes

Major

Faults

km 200

Figure 2.3

Simplified geological map of Africa and Madagascar.

mains are usually abundant in the Beaufort Series and the complex of reptiles shows increasingly mammallike features in the higher horizons, which are Triassic in age. The overlying Stormberg Series indicates progressive desiccation and the strata include much wind-blown sand of desertic character.

Two of the earliest true mammals so far known, Erythrotherium parringtoni and Megazostrodon rudnerae, come from a late Triassic unit of the Stormberg Series in Lesotho (Crompton 1964; Crompton and Jenkins 1968) and are assigned to the family Morganucodontidae, of which representatives are

22

Cooke

also known from Europe and China (Crompton 1974; see also this volume, Chapter 3). The Karroo cycle ended with the outpouring of up to 1,000 m of basaltic lavas in the early Jurassic, still preserved extensively in the highland of Lesotho. Although the complete succession is present only in the main basin of the Karroo, the system is widely represented by partial sequences elsewhere in southern and eastern Africa as far north as the equator, and there are equivalent beds in Madagascar as well as in other parts of Gondwanaland. In West Africa rocks of continental origin that were deposited at about the same time as the Karroo System are often included under the general term "Continental Intercalate," used by French geologists, and in Egypt and the Sudan they have been included with the "Nubian Sandstone," although both these terms in the strict sense should apply only to rocks of later Mesozoic age. In northern and western Africa the lower and middle Paleozoic are represented by extensive marine sequences related to fluctuations of the ancient Tethys Sea. At the end of the Carboniferous this region was largely emergent so that Permian and Triassic marine beds are restricted to a small part of Tunisia and Algeria. However, Permian marine deposits occur in western Madagascar, indicating the early stages of separation of that fragment of Gondwanaland, and Triassic marine rocks are also found there, as well as in the Sinai Peninsula and in eastern Arabia. The Jurassic is marked by renewed encroachment of the northern Tethys Sea over Morocco, northern Algeria, and Tunisia, but this period is also notable for the development of an arm of the sea across Arabia and Ethiopia and along the eastern side of the present African continent as far south as central Mozambique, as well as western Madagascar. At Tendaguru in Tanzania marine and continental beds interfinger, and from a high Jurassic horizon an edentulous mammalian jaw was found and named Brancatherulum tendagurense by Dietrich (1927); it may be a paurodontid pantothere (Simpson 1928). Evaporites underlie mid-Jurassic marine beds in Tanzania and are presumably associated with the warping and rifting that led to separation of the "eastern" parts of Gondwanaland and early growth of the Indian Ocean. The rifting of the KaribaLuangwa-Lower Zambezi troughs also probably dates back to the Jurassic, but the rifts did not spread, and they became the receptacles for Cretaceous sediments in which dinosaur remains occur. On the west side of Africa the Benue Rift was formed in the late Jurassic or early Cretaceous but also did not spread. At this time rifting began along the

Africa: The Physical

Setting

boundary between Africa and South America and lacustrine and saline deposits were formed in the early Cretaceous before the south Atlantic began to open up. Upper Cretaceous marine beds occur all the way down the Atlantic coast of Africa, but initially with a fauna different from that of the Tethys. North Africa was invaded extensively by the Tethys Sea, which in mid-Cretaceous times broke across the western part of the continent from Algeria to the Gulf of Guinea to link up with the young south Atlantic. Thus by mid-Cretaceous times Africa had for the first time achieved approximately its present borders, still with Arabia as an integral part of the continent, while both the Atlantic and the Indian oceans grew progressively wider. However, the distribution of late Cretaceous dinosaurs seems to demand maintenance of some kind of land link between Europe, South America, Africa, Madagascar, and India that is not easy to reconcile even with most recent interpretations of the palaeomagnetic evidence for the positions of the continents (e.g., Dietz and Holden 1970; Smith and Hallam 1970) but it is in fair agreement with the reconstruction by Smith, Briden, and Drewry (1973). These maps, however, reflect only plate distributions and are not paleogeographic maps in the strict sense. Paleocene marine beds occur in North Africa, including reef formations that became important oil traps. It seems clear that by this time Africa (with Arabia still an integral part) was effectively isolated from the other continents. The borders of the "new" continent were invaded by the sea at many points, with the Tethys covering an extensive area in North Africa, while a large embayment of the Atlantic extended over Nigeria (figure 2.4). According to the paleomagnetic evidence, Africa was at this time still some 15° of latitude south of its present position so that the equator ran from Senegal to the northern tip of Ethiopia. This must have had an effect on the climate, but the implications have yet to be evaluated. The continent was effectively in its present position by the early Miocene. During the Eocene the Tethys withdrew progressively in the north so that by Oligocene times only two narrow belts of marine encroachment remained (figure 2.4). Toward the end of the Oligocene, folding and elevation of the Atlas region led to the complete emergence of northwestern Africa and it is probably in the Oligocene that rifting of the Red Sea trough began and soon led to the almost complete isolation of Arabia by flooding of the trench in mid-Miocene times. In the Miocene the Tethys (Mediterranean) again encroached upon the continental margin, but the invasion was not nearly as extensive as in the

Cooke

Africa: The Physical Setting

23

major basins (figure 2.1) are not so much areas of sagging as areas that have lagged behind in the positive general uplift that has affected the continent since the disruption of Gondwanaland. The Congo Basin was successful in breaching its rising western limb, but the great Cubango-Kalahari Basin is largely one of internal drainage filled with debris that could not be carried away. The Chad Basin is also one of internal drainage containing much accumulated debris. Other small, closed basins have been created in the volcanic piles of the East African -Ethiopian area, but these are the product of more localized deformation associated with the volcanism and the fracture systems of the Great Rift Valley.

Major Fossil Mammal Localities General F i g u r e 2.4 Epicontinental seas in Afro-Arabia during the Eocene (light stipple) and Oligocene (heavy stipple). The dashed line shows the incipient Red Sea Rift. The full horizontal line is the estimated Eocene equator.

Eocene. The principal mammal-bearing deposits in North Africa tend to lie fairly close to the fluctuating Tertiary shorelines. At the end of the Miocene the Mediterranean Basin was temporarily closed off from the Atlantic and evaporation led to substantial shrinking of the water body accompanied by the formation of extensive saline deposits. This phenomenon must have had significant effects on the climate and on animal life in the region. Before the severance of Africa from its neighboring components of Gondwanaland, the continent appears to have been reduced to a landscape of fairly low relief, which L. C. King (1951, 1962) has named the Gondwana surface, still recognizable as patches on top of areas of greatest uplift. With the creation of a new coastline, rejuvenated rivers began the slow process of dissection of the uplifted surface and long stability led to the formation of the most extensive planation, variously called the African surface, midTertiary surface, or Miocene peneplain. Substantial uplift in the Miocene has led to development of two levels of broad flat areas flanking the main river systems, and late Miocene uplift led to incision of the streams in the lower parts of their courses. The total elevation since the Cretaceous is difficult to estimate, but in sub-Saharan Africa post-Gondwana uplift of about 900 to 1,200 m seems to be demanded, with about half of it since the mid-Tertiary. The

The principal localities that have yielded fossil mammals are shown in figure 2.5, although many Pleistocene sites have had to be omitted. There are no Tertiary localities and few Pleistocene sites in the heavily forested equatorial region, although Malembe near the mouth of the Congo borders on it. In general, the more important discoveries have been made in three areas: South Africa, East Africa, and the strip of land bordering the Mediterranean. The latter embodies two regions, the Atlas ranges and adjoining area of northern Algeria and Tunisia (known to the Arabs as the Maghreb) and the northern part of the Libyan and Egyptian deserts. In the Maghreb the fossil occurrences are generally isolated and restricted, especially in the zone of the Alpine folding, but in the desert area to the east are more extensive and are in some degree interrelated by the paleogeography.

Eocene—Oligocene The oldest Cenozoic mammals so far described from Africa are of Eocene age and include some of the earliest marine mammals known. The first discovery was made by Schweinfurth in 1879, when he found remains of the primitive Cetacean Dorudon in the Fayum area of Egypt (Dames 1883, 1894). Schweinfurth's account of the geology was published in 1886, and the area was mapped by Beadnell (1901a, 1905), when he found terrestrial mammals, some of which he described himself (1901b, 1902) while others were published in a series of papers by Andrews, which culminated in his classic monograph (1906). Many expeditions have worked in the area and an excellent historical summary is given

Cooke

24

Africa: The Physical

Setting

Ternifine Mascara 1 Bel Hacel y,Αϊη Boucherit Oued-el-Hammam Lac / Lac Ichkeul Mugharet el'Aliya ^ ^ J J U / - ^ - Δ ΐ , ferryville O u e d el A k r e c h ^ F o u a r a t ^ O • Δ aDjebei M'Dilla Mansaura Sidi Abderrahman^> Bled ed Douarah Beni Mellal Ai'n Hanech' St Arnaud

O LATE PLEISTOCENE Ο MID PLEISTOCENE Δ EARLY PLEISTOCENE *

PLIOCENE

• LATER MIOCENE A EARLIER MIOCENE • OLIGOCENE

-Moghara *\Wadi Natrun

Ο EOCENE

Springbok·*^" Langebaanweg^t ElandsfonteinjC- L> , „ Melkbos® — Nelson Bay MeiKDOS Swartk,jp Figure 2.5

0

200

Principal fossil mammal localities in Africa. The inset shows the East African area at an enlarged scale.

Cooke

by Simons (1968), who led the most recent group from Yale University. Eocene marine beds, consisting largely of massive or well-bedded limestones, sometimes laminated and carrying flint bands and concretions, are extensively developed in Egypt and Libya. Large foraminifera are commonly present. Marls and shales occur occasionally, becoming more frequent in the upper Eocene. Vertebrate remains are extremely rare except in the Fayum, where the Qasr el Sagha Formation consists largely of deltaic and interdeltaic deposits with occasional channel sands, and vertebrate fossils are not uncommon. There are interfingering marine horizons and the invertebrate fauna suggests a late Bartonian age. Several species of cetacean and the sirenian Eotheroides have been found, as well as fish, crocodiles, and the earliest known flightless bird. Plant debris and silicified logs are present and the terrestrial vertebrate remains tend to be associated with them. The land mammals include the primitive "proboscidians" Moeritherium and Barytherium, a hyaenodont and an anthracothere. Overlying the Qasr el Sagha Formation is the Gebel Qatrani Formation, some 200 m thick, consisting of a wide variety of clastic sediments and rare lacustrine limestones. The lower 50 m are rich in fossil wood and most of the terrestrial vertebrates collected before 1961 came from the top of this zone. They are mainly large mammals, including the two Eocene "proboscidean" genera already mentioned as well as two further proboscideans, Palaeomastodon and Phiomia. Hyracoids are abundant and the earlier expeditions found a few rodents and very rare primates. Carnivores and anthracotheres are also present. A much thinner fossil wood zone occurring about 80 m higher in the sequence furnished substantial amounts of material to the Yale expeditions (Simons 1968), and important specimens are also derived from a thin horizon about midway between the Lower and Upper Fossil Wood Zones. The Yale expedition recovered many small mammals including rodents and important primates, largely from the higher horizons. Some 75 m above the Upper Fossil Wood Zone a basaltic lava caps some of the hills, clearly following an erosion interval of unknown duration, and the basalt has a K-Ar date of approximately 27 ± 3 m.y. The rich and varied fauna from the lower zone is generally considered to be early Oligocene in age but it is possible that the upper zone may be middle Oligocene. The sediments are interpreted as representing a channel floodplain complex laid down in an area of low relief under semiarid conditions, but with dense gallery forest

Africa: The Physical Setting

25

along the stream channels and perhaps with some savanna existing between them. This interpretation is consistent with the general aspect of the fauna and would help to account for the presence of the important and varied primate fossils. Eocene marine mammals have also been recorded from the Mokattam Hills near Cairo and from Buel Haderaut in Libya, but no land mammals are associated with them. Elsewhere in Africa sirenians have been reported from Somalia, both in Eocene and Oligocene beds (Savage 1969). A cetacean of mid-Eocene age is recorded from Ameki in Nigeria. Fragmentary remains of Moeritherium were found in a sequence of upper Eocene marine beds at M'Bodione Dadere in Senegal (Gorodiski and Lavocat 1953) and also in continental beds at Gao and at In Tafidet, both in Mali. However, terrestrial mammals of this age are not otherwise known from the sub-Saharan region. The marine Eocene extends westward from Egypt into Libya and southern Tunisia, where it is well developed in the Sirte Basin. Occasional finds of sirenians have been made but the most important area is the Dor el Talha escarpment where terrestrial mammals were first recorded by Arambourg and Magnier (1961) and examined more extensively by Savage (1969, 1971). (Arambourg and Magnier called the locality Gebel Coquin, but the name Dor el Talha is preferred by Savage.) The lowest beds are calcilutites, with abundant oysters and other marine invertebrates. This is followed by 60 m of marls, the lower part of which has sandy channels, bituminous horizons, nodular layers, and gypsum sheets. Land mammals are rare but marine and aquatic mammals, fish, crocodiles, and turtles are abundant, as well as fossil wood and oyster bands. The upper part of the marls is more massive or poorly bedded, becoming thinly bedded at the top. Land mammals are abundant and are associated with fish and reptiles. Capping the marls is 30 m of unfossiliferous sandstone. The oysters in the calcilutites are correlated with those in the middle to late Eocene of Egypt. The lower marls contain sirenians and cetaceans, as well as Moeritherium and Barytherium, as in the Qasr el Sagha Formation of the Fayum and they are probably of equivalent age. Both these genera continue into the upper marls but without aquatic mammals, and this part of the sequence is most probably early Oligocene in age and equivalent to the fauna from the lower quarries in the Gebel Qatrani Formation of the Fayum. In the Oligocene the Sirte Basin was still largely marine but there are two localities in the marginal area that have furnished terrestrial mammals of

26

Cooke

this period. The best fauna, although scrappy, is from Zella, 300 km northwest of Dor el Talha, where an estuarine conglomerate has yielded two of the Fayum proboscideans, a large hyracoid, a carnivore, and an anthracothere, as well as crocodile and turtle remains (Arambourg and Magnier 1961). Farther to the northwest, at Gebel Bou Gobrine in southern Tunisia, is another littoral zone sandstone from which scrappy remains were recovered, including the Fayum proboscidean Phiomia (Arambourg and Burollet 1962). (Arambourg and Magnier give it as "Bon" but the correct Arabic is "Bou.") The fossilbearing horizon lies 100 m below a marine deposit with abundant oysters and pectens of early Miocene (Burdigalian) age and the mammals are certainly Oligocene.

Miocene of North Africa In North Africa the Miocene sea advanced over a surface that had been folded during the late Oligocene, so there are several somewhat distinct basins of deposition. The sites with terrestrial mammals are related to the shoreline and tend to belong to fluviatile or estuarine environments. In Egypt fluviomarine and deltaic beds are exposed over a long distance and at three localities, in the Moghara Oasis and in Siwa and Wadi Faregh, they contain fossil mammals. Moghara has the richest fauna, with the proboscidean Gomphotherium, anthracotheres, perissodactyls, a felid, and a monkey, Prohylobates. The age is somewhat uncertain but is probably early Miocene. Probably the richest and best site in North Africa is that of Gebel Zelten, 200 km northwest of Dor el Talha. It was first recorded by Arambourg and Magnier (1961) and has been worked more extensively by Savage (Savage and White 1965; Savage 1971). The regional setting and stratigraphy are well described by Selley (1968), who shows that the mammal-bearing sediments belonged to a coastal flood plain separating intertidal deposits to the north from the low-lying hinterland of the Sahara to the south. Mixed marine and continental deposits occur both below and above the vertebrate-bearing unit. The rich fauna is regarded by Savage as "Burdigalian" 1 and to indicate a savanna environment. It includes proboscideans, perissodactyls, giraffids, suids, ruminants, carnivores, and creodonts; but sirenians, fish, turtles, crocodiles, and birds have 1 The use of European stage names like Burdigalian may be feasible in the marine associations, but their employment for terrestrial deposits in Africa is hazardous and attention is drawn here to this fact by inserting such terms in quotation marks in this text.

Africa: The Physical Setting

also been found. The whole assemblage of terrestrial mammals is very similar to that from the early Miocene sites in East Africa but the smaller elements and micromammals are lacking at Gebel Zelten. Some 200 km to the northeast of Gebel Zelten is the site of Qasr es Sahabi, where Miocene marine clays with invertebrates and fish occur in a sequence of estuarine, lacustrine, and fluviatile deposits containing well-preserved terrestrial and freshwater vertebrates (Petrocchi 1943). The collection includes a gomphotherelike elephant, Stegotetrabelodon, a mastodont, dinotheres, rhinoceros, an equid, anthracotheres, hippopotamus, suids, bovids, and carnivores. The collection has not been fully described but the age is probably very late Miocene. At El Haserat, 50 km north of Sahabi, an anthracothere has been found and may be of similar age, but this is speculative. In Tunisia, north of Gebel Bou Gobrine, vertebrate fossils have been found at a number of localities, the most important of which is Bled ed Douarah some 40 km west of Gafsa (Robinson and Black 1969). Fossils occur at a number of horizons in the Beglia Formation, to which a "Vindobonian" age is assigned. The Bled ed Douarah fossils come mainly from a fluvio-deltaic complex, which passes northward around Sbeitla into dunes and then into a littoral sand facies at Djebel Cherichira (Biely et al. 1972). The fauna is rich and varied, including mastodonts, Deinotherium, suids, anthracotheres, birds, fish, and, in the upper levels, Hipparion primigenium (Forsten 1972). The birds indicate that both savanna and freshwater stream habitats are represented (Rich 1972). Merycopotamus is found both in the lower and in the upper faunal levels and changes indicate that a time lapse exists between the lower beds, without Hipparion, and the upper levels in which this equine occurs (Black 1972). The caprine bovid Pachytragus is present before Hipparion and is considered by Robinson (1972) to be of African origin. The freshwater fish are strongly African in affinities and contrast markedly with the Eurasian aspect of the modern freshwater fishes of Tunisia (Greenwood 1972). The Beglia Formation usually rests unconformably on earlier rocks, probably because of the Miocene tectonism associated with the Atlas region and its flanks. In the area around Djebel Mrhile and Henchir Beglia there are underlying "Burdigalian" deposits of the Mahmoud Formation that have yielded some fossil remains. On the eastern flank of the Atlas, in the Oued (or Wadi) el Hammam Valley southeast of Oran, marine deposits with plentiful invertebrate fossils assigned to the lower Miocene are overlain uncon-

Cooke

formably by some 400 m of alternating reddened sandstones, sands, shales, and conglomerates, with fossil mammal remains scattered through them; land snails also occur (Arambourg 1951). The continental beds belong to the Bou Hanifia Formation and are covered by marine sandstones and shales with oysters and gastropods. A tuff in the lower part of the Bou Hanifia Formation has been dated radiometrically at 12.18 ± 1.03 m.y. (Ameur, Jaeger, and Michaux 1976) and this provides an important reference point for correlation in the Maghreb. The fauna contains Hipparion, a rhinoceros, hyenas, Orycteropus, bovids, giraffids, rodents, ostrich, and turtle (Arambourg 1951, 1954a; Jaeger, Michaux, and David, 1973). It seems probable that these continental deposits belong to the same general environment as the Beglia Formation of Tunisia but if the absence of Hipparion in the lower part of the latter is real, the basal Beglia may be slightly older than the Bou Hanifia. Sporadic finds of Hipparion and Mastodon have been made elsewhere in Algeria and in Morocco at Gara Ziad, Melka el Ouidane (CampBerteaux), and Tadla Beni Amir, all probably later Miocene in age (Choubert and Ennouchi 1946; Jaeger, Michaux, and David 1973). A very similar fauna, richer than most of the others, occurs at Marceau in Algeria and is also later Miocene in age.

Miocene of Sub-Saharan Africa In sub-Saharan Africa only two Miocene localities have been reported outside the East African region, Malembe and the Namib. A third, Bololo, lies just north of the Congo River 30 km from its mouth and has yielded scrappy mammal remains from a gravel bed of Quaternary age, but the recognizable fragments of molars of Gomphotherium suggest derivation from an unknown Miocene source. On the coast at Malembe, 80 km to the northwest, fossil mammals of mid-Miocene age occur in an estuarine phase of a marine sequence that includes fish, a cetacean, and a sirenian. The terrestrial mammals were presumably washed in and include Gomphotherium, chalicothere, carnivore, suid, and anthracothere remains (Hooijer 1963). Well to the south, in the Namib Desert area, fluvio-lacustrine and continental beds occur and have provided a limited but interesting fauna (Stromer 1926; Hopwood 1929). Rodents are the most important element and there are also insectivores and hyracoids. The larger elements include a small suid, a hyaenodont carnivore, and two bovids. The age is generally regarded as "Burdigalian." The East African region possesses a substantial number of fossil mammal-bearing Miocene, Plio-

Africa: The Physical Setting

27

cene, and Pleistocene deposits, the most important localities being shown in the inset to figure 2.5. The richness of the region in terrestrial fossils is largely a consequence of the volcanic and tectonic environment, which provided favorable basins as traps for sediment, rapid burial, and calcic or alkaline materials to prevent the more usual leaching and destruction of buried bones. The presence of volcanic materials in close stratigraphic association with fossil-bearing sediments has also provided unusually good radiometric controls for the evaluation of the succession. As mentioned earlier, the African continent has experienced considerable uplift, especially in the zone through which runs the complex Rift Valley System. The rifts themselves follow old structural trends in the crust and some of the rifting dates back as far as the late Jurassic or early Cretaceous. By Oligocene times extensive planation had led to the exposure, in the East Africa region, of an erosion surface consisting dominantly of ancient granitic and metamorphic rocks. The continental divide probably lay in the region of the present Eastern Rift and the general altitude of the erosion surface may have been only in the order of 500 m above sea level, with residual masses rising above it (Cooke 1958). Saggerson and Baker (1965) showed the existence of a swell in the basement along the Eastern Rift and estimate that the residual mass of the Cherengani Hills has been uplifted as much as 1,700 m since the time of development of the subvolcanic surface; this would mean a former altitude of about 1,500 m above sea level for the divide at the time. Bishop and Trendall (1967) have examined the subvolcanic surface in Uganda and shown that a westfacing erosion scarp existed south of Moroto, rising some 600 m in 30 km and continuing beneath Mount Elgon. The drainage was westward from a watershed along the line of the residuals of the Chemorongi and Cherengani hills and there was modest relief in what they term the "Kyoga surface." This surface was tectonically modified by the growth of domes on the sites of some of the subsequent major volcanoes and by the development of local basinal structures as volcanism began in the early Miocene. The history of tectonic events in the Kenya Rift is well outlined by Baker and Wohlenburg (1971; Baker et al. 1971; Baker, Möhr, and Williams 1972) and the lava sequences have been described recently by King and Chapman (1972). Miocene fossils were first found in 1909 at Koru, on the flanks of the old cone of Tinderet, and later at Karungu on the south side of Kisingere (Andrews 1911). The richest sites are on the northwest side of

28

Cooke

Kisingere, on Rusinga Island and Mfwanganu Island, which lie near the entrance to the Kavirondo Gulf of Lake Victoria; on the north side of the gulf are other localities, Maboko and Ombo, that are less closely related to volcanic cones. Two other important localities occur on the flanks of Tinderet— Songhor to the north of Koru and the later Miocene site of Fort Ternan to the east. In Uganda there are two important sites some 250 to 300 km north of the Kavirondo Gulf, Napak, on the flanks of an old volcano, and Moroto, on the subvolcanic basement. Lesser sites in Kenya, such as Muruorot and Loperot, are not obviously related to former cones but are associated with extensive lava flows. The literature is abundant, but the initial history of discoveries is well set out by Le Gros Clark and Leakey (1951). The association of volcanic materials has made possible the determination of many radiometric ages (Bishop, Miller, and Fitch 1969; Van Couvering and Miller 1969), reflected in figure 2.6. The faunas have been listed by Leakey (1967) and Bishop (1967), with good bibliographies to that date. Rusinga Island is by far the richest area, having yielded many thousands of fossils. The geology was first described in some detail by Kent (1944) and amplified by Shackleton (1951), Whitworth (1953), and McCall (1958). Van Couvering and Miller (1969) have made some amendments to the earlier stratigraphic interpretations and provided a convenient summary as well as radiometric dates. The island is underlain by the Kiahera Formation, consisting of brown clastic sediments passing into calcic volcanogenic sediments. Overlying this unit unconformably is the 40-m thick Rusinga Agglomerate, largely made up of blocks of nephelinite lava in a matrix of comminuted lava fragments, but with volcanic grits and silts interbedded locally; its radiometric age is 19.6 m.y. The Hiwegi Formation, 50 to 70 m thick, consists of a wide range of volcanic, clastic, and calcareous sediments, including silty beds that constitute the main fossil-bearing unit lying about 20 to 25 m above the base. The Hiwegi Formation is covered by the Kiangata Agglomerate, rich in nephelinite blocks, which in turn is capped by the Lunene Lava, a melanephelinite with a radiometric age of 16.5 m.y. The major fossil horizon in the Hiwegi is thus interpolated as close to 18.0 m.y., but fossils also occur in the upper Kiahera and in a unit known as the Kulu Beds, which lies in an erosion hollow cut into the Hiwegi Formation but whose exact age is not yet clear. The neighboring island of Mfwanganu has been described by Whitworth (1961) and the formations parallel those of Rusinga, although differing in detail.

Africa: The Physical

Setting

Bishop (1963, 1968) has reviewed the distribution of the major fossil-bearing sites and shown that they reflect primarily conditions suitable for fossilization. He recognizes (1) fully lacustrine conditions predominating, with fish fossils present and a lithology mainly of fine-grained clastic sediments and waterdeposited tuffs; (2) subaerial deposits with tuffs, fossil wood, and land gastropods, with interstratified water-borne sediments, resulting from temporary ponding and flash floods; (3) coarse clastics, including conglomerates, mixed with a variable proportion of fine volcanic ash as a matrix, usually resting on the basement and underlying the volcanic sequence proper. The nonvolcanic sediments of type 3 accumulated in valleys and hollows on the irregular prevolcanic topography, some of which may have been the result of tectonic deformation preceding the volcanicity. The fossil assemblages are generally more or less rolled and fragmented, with small faunal elements lacking, but there are rare associated bones suggesting that carcases had been swept in occasionally. Some of the Napak sites, Moroto, Ombo, and Marewa, are regarded as of this type. Fully lacustrine environments, together with subaerial deposits, occur at Bukwa and at Karungu, Mfwanganu, and Rusinga. The remaining sites with well-preserved specimens owe the good preservation to fine-grained subaerial primary tuffs and there may be paleosols indicative of long periods of quiescence and weathering. These conditions prevailed at some of the Napak sites and at Songhor, Koru, and perhaps also at Maboko. The Moruorot deposits consist largely of subaerial tuffs sandwiched between lava flows. Loperot is a mixed environment with subaerial tuffs, fluviatile grits, and possibly lacustrine sediments beneath the Lower Turkana Basalt. Also somewhat different in character are the Miocene occurrences on the west and southwest sides of Lake Albert, where a thick sequence of fluvio-lacustrine sediments of Miocene to Pleistocene age filled the subsiding rift. The major occurrence of terrestrial mammals is at Karugamania, but some specimens are also found in the lower Nyamavi Beds (Lepersonne 1949; Hooijer 1963, 1970). Although the majority of the East African faunas are traditionally regarded as "Burdigalian" because of the general resemblance to the Burdigalian faunas of Europe, Van Couvering (1972) argues convincingly that the radiometric dates show most of the sites to be, in fact, coeval with the Aquitanian of Europe. Thus the "Rusinga-like fauna" may have provided the ancestral stock for many of the characteristic Burdigalian mammals of Eurasia and North

Cooke

Africa: The Physical

With radiometric control

Olduvai

Correlated by fauna only

III/IV Π

1

1

Shungura

Koobi Fora

I

I Kanapoi; Mursi

Kubi Algi

Lothagam-1 Lukeino Mpesida; Melka el Ouidane

29

Setting

Sidi Abderrahman; Olorgesailie; Cornelia Temifine; Yayo; Peninj; Kromdraai Ain Hanech; Bel Hacel; Swartkrans Ouadi Derdemy; Kouala; Sterkfontein Ain Boucherit; Ain Brimba Hamada Damous; Chiwondo; Kaiso; Hadar Bochianga; Langebaanweg-1 Wadi Natrun Sahabi

Oued Zra Ngorora

Beglia Formation

Bou Hanifia

Wadi el Hammam

Muruyur; Alengerr Fort Tern an

Beni Mellal Maboko Malembe Moghara; Wadi Faregh; Siwa Gebel Zelten

Rusinga; Napak Koru; Songhor Namib

Bukwa; Karungu

Zella Gebel Bou Gobrine

Gebel Qatrani Dor el Talha

distribution of later Cenozoic mammal-bearing deposits in Africa.

30

Cooke

America. On the other hand, the younger East African Miocene fauna at Fort Ternan (close to 14 m.y.) shows a substantial change from the Rusinga-like fauna and includes elements suggestive of Eurasian (especially Asian) immigrants (Gentry 1970). Fort Ternan, formerly considered terminal Miocene, is now middle Miocene on the basis of the revised marine time scale, which places the Miocene-Pliocene boundary between 5.0 and 5.5 m.y. (Berggren 1971; Van Couvering 1972). Until less than a decade ago there was a substantial gap in the East African fossil record between Fort Ternan and the terminal Pliocene-early Pleistocene sites of Kenya, Tanzania, and Ethiopia. Recent mapping in the Kenya Rift north of Lake Hannington has disclosed a long sequence of volcanics, some 3,000 m thick, within which occur a number of sedimentary units of varying thickness (McCall, Baker, and Walsh 1967; Martyn 1967; Bishop and Chapman, 1970; Bishop et al. 1971). On the east side of the Tugen Hills successive faults step down the strata from the crest at about 2,800 m above sea level to the floor of Lake Baringo (1,000 m) while on the west side a dissected dip slope descends into the Kerio Valley. The succession rests upon the basement complex and the lower 1,800 m outcrops in the upper part of the main escarpment, capped by a useful marker, the Kabarnet Trachyte, with an age close to 7.0 m.y. The lower phonolites range back to more than 16 m.y. Twelve sedimentary formations of varying thickness occur at intervals through the lava sequence; five of them are older than the Pliocene boundary, namely the Alengerr, Muruyur (both older than 12 m.y.), Ngorora (>9.0 m.y. ο £ W £ Ü Λ

01
Ο φ Ζ oc

S.S «g Η to cd Ö cd O) ^ £ 2 ο -ö u ^ c 2 ο ö __ •Ο S toφ Μ 'S

03

I g^ fc t> ο IN

Φ ·|

io S £ o

32

Cooke

mammalian fauna (Carney et al. 1971). The fossil-bearing unit is 25 to 30 m thick and also contains stone tools recalling the Oldowan industry, while a younger unit has furnished Acheulian artifacts (Bishop, Pickford, and Hill 1975). The fauna suggests an age equivalent to some part of Bed II at Olduvai. Some 200 km to the north of Lake Baringo in the drainage basin of the Kerio River there are three localities on the southwest side of Lake Turkana (formerly Lake Rudolf) that have yielded important collections of fossil mammals. The oldest fauna comes from Lothagam Hill, where Miocene volcanics, dated at 8.31 m.y., are overlain unconformably by a boulder conglomerate followed by a sedimentary sequence with a total thickness of about 600 m. The beds dip westward at moderately low angles and three major units are recognized (Patterson, Behrensmeyer, and Sill 1970). The lowest, Unit 1, consists of conglomerates, thinly bedded sandstones, silts, and occasional shales or tuffaceous shales, apparently representing fairly continuous and rapid deposition on an aggradational plain, perhaps deltaic; this is the major fossiliferous unit. Unit 2 comprises some 80 m of lacustrine clays and silts, almost devoid of fossil mammals. Unit 3 consists of at least 100 m of medium- to coarse-grained clastic sediments, probably of fluvial origin and containing a small number of mammalian fossils; the top is concealed. The contact between Units 1 and 2 is obscured by an intrusive basaltic sill, about 25 m thick, with an isotopic age of 3.71 m.y. As the sill postdates Unit 3, it provides a limiting age for the entire sequence, and it seems probable that the major fossiliferous unit, Lothagam-1, would be about 6.0 or 7.0 m.y. old and hence terminal Miocene (Maglio 1973). The fauna contains the most primitive known elephants (Maglio 1970, 1973), characteristic suids (Cooke and Ewer 1972), the last brachypotherine rhinoceroses (Hooijer and Patterson 1972), and primitive bovids with some resemblances to the Fort Ternan forms. The scanty fauna from Lothagam-3 resembles that of Kanapoi. The Kanapoi locality lies 60 km south of Lothagam and has more than 70 m of gently dipping clastic sediments, the base of which is not exposed. Most of the fossil material comes from the lower part, which consists of clays, silts, and some sandstones, whereas the upper part is generally coarser. A basaltic lava caps the sediments and has a radiometric age close to 4.0 m.y. The fauna is decidedly different from that of Lothagam-1, with the appearance of many new genera and the disappearance of others (Patterson 1966; Maglio 1970; Cooke and Ewer 1972) but is essentially similar to that of Lothagam-3. At

Africa: The Physical

Setting

Ekora, to the northeast of Kanapoi, another basalt is exposed and is overlain by a thin group of silty to gritty sediments, the top of which is concealed. These Ekora Beds have furnished relatively few fossils and the assemblage is like that of Kanapoi, although there are some differences in the proboscideans represented (Maglio 1970). At the northern end of Lake Turkana there are two areas in which occur long sequences of strata covering a considerable time span, the Omo Basin to the north of the lake and the so-called "East Rudolf" or "East Turkana" region on the northeast side of the lake. In the southern Omo Basin of southwestern Ethiopia an isolated area at the foot of the Nkalabong highlands displays about 140 m of clays, silts, and sands, often mottled and stained by limonite ("Yellow Sands") together with occasional gypsiferous lenses, salt horizons, and concretionary shell beds. The sediments are capped by a basalt dated at 4.18 m.y. and the whole unit is known as the Mursi Formation (Butzer and Thurber 1969). A gravel horizon in the middle of the sediments has furnished a very limited fauna inseparable from that of Kanapoi. A number of isolated exposures of sediments and pyroclastics totaling 200 m in thickness, designated the Usno Formation (de Heinzelin and Brown 1969), are found 25 km to the east, near the confluence of the Usno stream with the Omo River. There is a thin basaltic lava intercalated near the base with a provisional age of 3.3 m.y. and a tripartite tuff 50 to 60 m below the top with a radiometric age of 2.64 m.y. Two localities in the Usno Formation, known respectively as "White Sands" and "Brown Sands," have furnished good fossil assemblages from the sandy beds and these can be correlated by magnetostratigraphy with the lower part of Member Β of the Shungura Formation (Brown, pers. comm.). The most extensive sequence outcrops on the west side of the Omo River for a distance of 60 km, although the major exposures are in the northern 20 km. These are the "Omo Beds," first worked extensively by Arambourg in 1932-33 (Arambourg 1947). They are now formally designated the Shungura Formation (Brown, de Heinzelin, and Howell 1970). The beds are tilted westward at about 10° to 25° and comprise more than 750 m of sands, silts, and clays together with tuffs that are often prominent and distinctive enough to be used as marker horizons. The marker tuffs have been assigned alphabetic designations and the whole sequence divided into members made up of the tuff and the overlying sediments; a Basal Member underlies Tuff A (figure 2.7). Radiometric ages from many of the tuffs have been combined with an excellent paleomagnetic rec-

Cooke

ord (Shuey, Brown, and Croes 1974) to provide a well-controlled time scale ranging from 3.1 m.y. for the Basal Member to 0.8 m.y. for the upper part of Member L (there is no member designated I, so J follows H). Within the members, cyclic lithologic sequences occur, interpreted as resulting from rhythmic fluviatile deposition and the complex elements of a river floodplain. Paleosols indicate temporary pauses in the general subsidence of the basin but the upper part of Member G is largely lacustrine. A rich and progressively changing fauna has been recovered and its stratigraphic distribution was reviewed recently by Coppens and Howell (1974; Howell and Coppens 1974). Stone tools occur in and above Member F. Lying unconformably on the earlier beds is a horizontal unit named the Kibish Formation, comprising about 100 m of mixed fluvial, deltaic, littoral, and lacustrine sediments ranging from the Holocene back to a tentative 130,000 years or more (Butzer, Brown, and Thurber 1969). The East Turkana Plio-Pleistocene deposits extend from the Ethiopian border southward for at least 100 km on the east side of Lake Turkana and the belt varies from 10 to 40 km in width. A Miocene site exists in the region but has not yet been investigated in detail (Harris and Watkins 1974). The oldest Pliocene sediments lie in the southern part of the area and belong to the Kubi Algi Formation (Bowen and Vondra 1973). This consists of coarse clastic sediments lying unconformably on Mio-Pliocene volcanics, becoming finer upward through the 90-m thickness. Mammalian fossils are scarce but are sufficient to establish a general resemblance to that of Kanapoi and its correlatives (Maglio 1972). The upper limit of the Kubi Algi Formation is marked by laminated bentonitic tuffs and claystones, termed the Suregei Tuff Complex. This complex marks the base of the Koobi Fora Formation (Bowen and Vondra 1973), which is divided into lower and upper members at an apparent disconformity on top of a widespread tuffaceous unit known as the KBS Tuff. Because of partial isolation of the northern area around Ileret, the unit above the KBS Tuff in that area has been termed the Ileret member and is capped by the well-developed Chari Tuff. The total thickness of the Koobi Fora Formation is about 200 m and it consists of a complex of clastic sediments, ranging from claystones to conglomerates, together with molluskan limestones and other rare elements. The beds are considered to result from the interplay of fluvial, deltaic, and lacustrine environments. The formation seems in general to have been built by streams flowing from the east and northeast to form a prograding fluvio-deltaic complex, but episodes of lacustrine transgression and regression are ap-

Africa: The Physical Setting

33

parent. Although the tuffs are rarer than in the Omo Basin, they are important as marker horizons and some of them have been dated. The KBS Tuff was originally dated at 2.61 ± .26 m.y. by Fitch and Miller (1970) but the faunal evidence suggests correlation with Member F or G of the Shungura and hence a date closer to 1.8 to 1.9 m.y. (Cooke and Maglio 1972). Determinations by Curtis et al. (1975) suggested that there is more than one KBS Tuff and they report dates of 1.82 ± .04 and 1.60 ± .05 at two localities. Zircons have been dated by fission track methods at 2.44 ± .08 m.y. (Hurford, Gleadow, and Naeser 1976) and Fitch, Hooker, and Miller (1976) have made revised radiometric determinations giving an isochron age of 2.42 m.y. Preliminary paleomagnetic data can be interpreted as supporting this date, but the issue has not been resolved at the present time. The correlation of the faunas above the KBS Tuff are not in dispute and the East Turkana and Omo radiometric dates for this section of the record match very well. In figure 2.7 the radiometric dates have been used for the upper part but the positions of the KBS Tuff and the lower section are based on tentative faunal interpretations. Four faunal zones were originally established by Maglio (1972), but his list now needs some revision. The environment at East Turkana was conducive to excellent preservation and the fauna includes many important hominid fossils (Leakey 1970, 1975, 1976; Day 1975). Stone tools have been found in the KBS Tuff and in the upper part of the succession (Isaac, Harris, and Crader 1975). On the Ileret Ridge a 30-m thick, poorly fossiliferous unit, the Guomde Formation, overlies the Chari Tuff and consists dominantly of lacustrine molluscan limestone and dark laminated sandstones. Resting unconformably on the earlier formations in various parts of the area are thin diatomaceous siltstones termed the Galana Boi Beds, probably of late Pleistocene or Holocene age. In the lower reaches of the Awash River in the central Afar region of Ethiopia a sequence of clays, sands, and gravels about 140 m thick constitutes the Hadar Formation (Taieb et al. 1972, 1976). These deposits have yielded substantial collections of fossils with strong resemblances to the fauna from the basal part of the Shungura Formation and have also proved to be rich in hominid remains (Johanson and Taieb 1976). A basalt flow in the lower half of the succession has been dated radiometrically as about 3.0 m.y., which accords well with the faunal evidence. It is only during the past decade that most of the deposits described from the Baringo-Turkana Basin have become known. Until that time the only major

34

Cooke

sequence known was that of Olduvai Gorge, Tanzania, first described by Reck in 1914 and later by L. S. B. Leakey in 1951 after 20 years of work in the gorge. More intensive study since 1959 has led to new understanding of the fauna and archaeology (L. S. B. Leakey 1965; M. D. Leakey 1971), age of the deposits (Evernden and Curtis 1965; Curtis and Hay 1972), and a wealth of detail regarding the geology (Hay 1963, 1971, 1976). The deposits rest on granites and gneisses and the lowest unit recognized is a welded tuff, the Naabi Ignimbrite. Bed I overlies the ignimbrite and has a thickness in excess of 60 m. The lower part is pyroclastic and is largely covered by basalt, formerly thought to be the base of the sequence, and with a mean age of 1.96 m.y. The upper member of Bed I rests on the almost unweathered lava and comprises 10 to 40 m of tuffs and clays accumulated in and around a persistent saline lake with no outlet, so that the shoreline fluctuated greatly in response to precipitation and evaporation changes. The major fossil-bearing deposits were alternatively invaded by the lake, covered, and then exposed to dry out. The mean age of the tuffs from Bed I is 1.79 m.y. A conspicuous claystone and tuff, "Marker Bed B," terminates Bed I and is overlain by Bed II, the lower part of which is lithologically and faunistically similar to the upper member of Bed I. A zeolitized eolian tuff, designated the Lemuta Member of Bed II (Hay 1971) seems to mark a break in the sedimentation and above it sandstones are common and root casts occur in the claystones; lenticular conglomerates are widespread. Bed III usually consists of reddish-brown volcanic conglomerates, sandstones, and claystones cemented by zeolites and rather poor in fossil material. Bed IV may also be reddened but more commonly comprises grey clays, sandstones, and conglomerates, also poor in fossils. Beds III and IV together have a thickness of 20 to 50 m and represent dominantly a playa situation with stream channel deposits. Paleomagnetic studies show a change from reversed to normal polarity within Bed IV, probably representing the BrunhesMatuyama transition and thus providing a date of close to 0.7 m.y. Bed IV is overlain by the Masek Beds (Hay 1971), consisting of 10 m of wind-worked nephelinite tuffs cemented by calcite and zeolites. As a result of major faulting after the deposition of the Masek Beds active erosion took place in the late Pleistocene and two formations, the Ndutu Beds and the Naisiusu Beds, were laid down on the plain adjacent to the gorge and remnants survive on the flanks and floor of the gorge itself (M. D. Leakey et al. 1972). Immediately south of the Olduvai Gorge exposures along the Vogel River were named the Laetolil

Africa: The Physical

Setting

Beds by Kent (1941) and consist largely of fine tuffs from which a moderate fauna was described by Dietrich (1942, 1950). Recent work has yielded a number of important hominid fossils and has shown that the thickest section is as much as 130 m thick, while the fossiliferous zone has been bracketed by radiometric dates as between 3.59 and 3.77 m.y. old (M. D. Leakey et al. 1976). To the north of Olduvai, on the western side of Lake Natron, Plio-Pleistocene lavas are overlain by a 100-m thick sedimentary sequence known as the Peninj group (Isaac 1965, 1967). The lower unit, the Humbu Formation, consists of sands, clays, and basaltic tuffs with mollusk, fish, and mammal fossils, including a hominid mandible, associated with a lower Acheulian industry. An interbedded flow of olivine basalt has normal polarity but the ages are ambiguous (Isaac and Curtis 1974). The faunal material resembles that of Bed II at Olduvai. The overlying Moinik Formation consists largely of lacustrine clays and sands grading both laterally and upward into tuffaceous shales and trachytic tuffs. Faulting terminated the depositional cycle. In the western rift, where fossil mammal discoveries are rare, the most important finds have been in the Kaiso Formation on the east side of Lake Albert and continuing southward astride the Semliki River and the Kazinga Channel. The thick sequence in this trough includes unfossiliferous grits, conglomerates, sandstones, and occasional clays of the Kisegi Beds (Mohari Formation) passing upward through passage beds into the typical Kaiso Formation. The latter consists dominantly of clays, sandy clays, marls, micaceous silts, and fine sands within which there occur occasional ferruginous horizons from which, or just beneath which, virtually all the aquatic and vertebrate remains have come, as well as most of the Mollusca (Bishop 1965). The fossil mammals are derived primarily from two zones in the lower part of the Kaiso Formation and from a zone near the top of the formation, the faunas being termed the "earlier Kaiso" and the "later Kaiso" assemblages, respectively (Cooke and Coryndon 1970). The "earlier" assemblage suggests an age perhaps a little earlier than the typical Kanapoi fauna, whereas the "later" assemblage matches well with that of the Shungura Formation between Members C and G. The well-known Elephas recki occurs at Kaiso Village as part of the "later Kaiso" assemblage and belongs to the early part of Maglio's (1970) Stage 2. A gravel deposit at Kikagati on the Kagera River east of Lake Kivu (but not related to the Kaiso Formation) has yielded as its only fossil three molars of the earliest stage of E. recki. Below Homa Mountain, on the south side of the

Cooke

Kavirondo Gulf of Lake Victoria, Plio-Pleistocene sediments are exposed in a series of steep gullies. The oldest unit is the Kanam Beds (Kent 1941) consisting mainly of light brown lacustrine clays with interbedded fine tuffs and occasional gravels, grading laterally into shallow water and fluviatile sands and gravels. Fossil mammal remains have come from three areas, Kanam East, West, and Central, and also from a site called Kokkoth. Much of the fauna resembles that of the Kanapoi assemblage and suggests a similar age, but the collections also include more advanced elements that conflict with this view and indicate an age equivalent to the middle part of the Shungura Formation (ca 2.0 m.y.). It is possible that the collections sample different horizons and more work is required to resolve the problem. The Kanam Beds were faulted and eroded before the deposition of a younger unit of fluviolacustrine sediments known as the Rawe Beds (Kent 1941) comprising laminated shaly clays and fissile siltstones and sandstones, some of which have yielded fish fossils. Mammalian remains were found at several horizons and the fauna suggests broad equivalence to Bed II at Olduvai. To the northeast of Kanam another group of greenish tuffs and yellowish clays with limestone bands is known as the Kanjera Beds (Kent 1941). The fauna suggests that it is equivalent to the upper part of the Olduvai sequence, perhaps to Beds III and IV, and this is confirmed by the Acheulian hand axes from Kanjera. Two later Pleistocene lake terraces have been developed across the Kanjera Beds. In the eastern (Kenya) rift there are a number of sites of middle to upper Pleistocene age that are important for their archaeological remains rather than their faunas. One of the most remarkable is at Olorgesailie, to the north of Lake Magadi, where hand axes litter the surface in profusion; it has been well described by Isaac (1966). The fauna is limited but resembles that of Kanjera and the top of the Olduvai sequence. Hand axes also occur in diatomites at Kariandusi in the Nakuru-Naivasha Basin and there is a complex of former lake and fluviatile sediments in this basin, originally described by Nilsson (1929) and by Leakey and Solomon (1931) but recently reconsidered by Isaac, Merrick, and Nelson (1972). Many other archaeological sites in East Africa are discussed by Cole (1963).

Pliocene-Pleistocene of North Africa In Egypt the Pliocene is scantily represented and the only major fauna has come from the Wadi Natrun. Blanckenhorn (1901) described the section near Garet el Muluk, where some 25 to 30 m of flu-

Africa: The Physical Setting

35

vial and lacustrine sands and clays yielded fossils from three horizons, and assessed its age rather vaguely as Messinian to Pontian. The fauna, described mainly by Andrews (1902), Stromer (1913, 1914, 1920), and Tobien (1936), includes elements suggestive of an early Pliocene or even a terminal Miocene age, but others that seem more advanced; the site clearly needs reinvestigation. A few other sites in Egypt have furnished rare fossils, the most important of which is Anancus osiris from the Mena House, near Cairo (Arambourg 1946). In the Maghreb region a number of late Pliocene or early Pleistocene sites occur but it is difficult to place them with certainty either in relation to the European or to the sub-Saharan sites. A key locality in this regard is Oued Fouarat (or Αϊη el Arriss), 30 km northeast of Rabat, where marine sands with a Tortonian fauna are overlain by sands and gravels containing marine mollusks assigned to the "Moghrebian" stage, originally regarded by Choubert (1953) as equivalent to the Calabrian of the Mediterranean but now believed to be slightly earlier (Biberson 1964). Gravels in this deposit have yielded remains of the proboscideans Anancus osiris and Mammuthus africanavus, regarded by Arambourg (1970) as very characteristic of the North African "Villafranchian" fauna. At Oued Akrech, south of Rabat, a similar marine setting has furnished M. africanavus, together with Hipparion libycum, a large bovid, and a rhinoceros perhaps representing an early Ceratotherium (Arambourg and Choubert 1965). Substantially later is the middle Pleistocene complex at Sidi Abderrahman, near Casablanca, where marine gravels are covered by a calcareous dune that was attacked by an advancing sea, leaving against the eroded dune, and in karstic caves, a succession of both marine and terrestrial deposits. Acheulian tools are plentiful and correspond with those from the upper part of the Olduvai sequence. Human remains have been found but the associated fauna is generally poor; both Loxodonta atlantica and Elephas iolensis occur. Other archaeological sites and faunas are reviewed by Biberson (1961, 1967). In Algeria the most important area lies 6 km northwest of St. Arnaud on the Constantine Plateau, where there is a series of marine beds of late Miocene age overlain by lacustrine and fluvio-lacustrine deposits (the St. Arnaud Lake Beds) in which fossil mammals occur at two horizons (Arambourg 1970). The lower horizon is now named Αϊη Boucherit and has a fauna resembling that of Fouarat except that it includes Equus numidicus and Hippopotamus, so may be a little younger. The higher level, Αϊη Hanech, is cut off from the main sequence

36

Cooke

by a fault and has a later "Villafranchian" fauna, including Mammuthus meridionalis and other elements that resemble the general faunas of the middle Shungura or lower Olduvai deposits. A third site in this area is at the cemetery of St. Arnaud, where sands overlying fluvio-lacustrine beds have yielded a limited fauna of Pliocene aspect, including a small equine, Hipparion sitifense, that also occurs at Mascara (southeast of Oran), Αϊη el Hadj Baba (near Constantine), and in the earlier Kaiso assemblage in East Africa. Near Oran, in the Chelif Valley, marine beds of Astian and Calabrian type are overlain by dunes and middle Pleistocene terrestrial deposits. A few teeth of Hipparion libycum have been found in the supposed Calabrian. A little way to the east, at Djebel Bel Hacel, marine beds of Astian type are overlain by lacustrine marls and limestones and by fluvial conglomerates, all folded into an anticlinal structure. From these deposits some mammals have been found, including Mammuthus meridionalis, Equus, white rhinoceros, and several bovids, the whole fauna resembling that of Αϊη Hanech and regarded as "upper Villafranchian" (Arambourg 1970). Important younger localities are Lake Karär, to the southwest of Oran, and Palikao (or Ternifine), southwest of Mascara, both of which have yielded faunas resembling those of upper Bed II at Olduvai, although Ternifine is the earlier of the two North African sites (Arambourg 1954b, 1962). In Tunisia the most important site is Garet (or Lac) Ichkeul, west of Bizerta, where Mio-Pliocene marine beds are overlain by a boulder bed and then by sands, grits, and shales (Arambourg 1970). The whole series is quite steeply tilted. The lower sands have furnished an excellent fauna resembling that of Fouarat and Αϊη Boucherit. There is also a good pollen spectrum suggestive of some preglacial cooling (Arenes and Depape 1953). About 100 km to the southeast of Garet Ichkeul, or 45 km southeast of the city of Tunis, is a large valley known as Hamada Damous. In it are substantial exposures of estuarine marls and sands, with a dip of 7° to 9°, and vertebrate fossils occur at several levels; there are also interstratified oyster beds and correlation with Mediterranean stages may later become possible. The fauna (Coppens 1971a, b) includes Anancus osiris in the basal section and Mammuthus africanavus up to the middle of the sequence. The typical Kanapoi suid, Nyanzachoerus jaegeri (Coppens 1971 = N.plicatus Cooke and Ewer 1972) provides an interesting link with the East African assemblages. Another site in the same region, Djebel Mallah, midway between Garet Ichkeul and Hamada Damous, yields a similar fauna. There is also a third locality, Sidi Bou

Africa: The Physical Setting

Koufa, 25 km west of Hamada Damous. Coppens (1971b) correlates the lower part of the Hamada Damous section with Garet Ichkeul, the middle part with Αϊη Brimba, and the upper part with Sidi Bou Koufa, still regarding the latter as upper "Villafranchian" despite the absence of Mammuthus africanavus. In southern Tunisia, some 5 km north of the oasis of Mannsoura, is the site of Αϊη Brimba (Arambourg 1970). It lies at the foot of the tilted Cretaceous limestones that form the Gebel Tebaga and the fossils come from a complex of pink breccias, red shales, and greenish marls. The fauna is very similar to that of Garet Ichkeul. Further to the south, in the Djourab depression northeast of Lake Chad, there are a number of localities that have yielded fossil mammals, apparently covering the whole range from the Pliocene to late Pleistocene; but the stratigraphic relations are not clear and the relative ages are assessed on the faunas (Coppens 1967). A broadly "Villafranchian" fauna was first reported from the area near Koro Toro (Abadie, Barbeau, and Coppens 1959; Coppens 1960). The oldest faunal assemblage, typified by Bochianga, Atoumanga, and Kolinga I, comes from a very fine white sand, often consolidated, underlain by clays, and overlain by diatomite. The fauna is poor but includes Anancus osiris and Primelephas korotorensis, suggesting an age roughly similar to Lothagam-1 in Kenya. A younger fauna, from Ouadi Derdemy and Koulä, comes from a greenish sandy clay or sometimes from an underlying clay. Anancus is present, together with Primelephas (at Koulä) and also both a progressive Mammuthus africanavus and Elephas recki (at Ouadi Derdemy). Although Primelephas is not otherwise known in this association, the occurrence of Elephas recki suggests an age equivalent to the lower part of the Shungura Formation in Ethiopia, and the progressive Mammuthus africanavus indicates an age younger than the Fouarat/Αϊη Brimba assemblage but older than Αϊη Hanech and Bel Hacel. This serves as a useful link between North Africa and sub-Saharan Africa. The large suid Notochoerus is also present, along with a hipparionid, a small hexaprotodont hippopotamus, and a sivathere. A younger fauna occurs at Yayo, typically with Loxodonta atlantica and a hippopotamus resembling the living form, but smaller, and it is roughly similar to the Ternifine assemblage; a problematical hominid, Tchadanthropus uxoris, also comes from Yayo (Coppens 1965). Younger levels are exposed around Ounianga Kebir 400 km northeast of Koro Toro, where the advanced Elephas iolensis has come from one horizon and Loxodonta africana from a higher level.

Cooke

Later Pleistocene fossils occur at many other sites in the Chad Basin, indicating the existence of typical savanna elements and even forest in this area, at least intermittently, during most of the Pleistocene (Franz 1967). Many sites throughout the Sahara have faunas indicative of more humid conditions at times during the Pleistocene (Monod 1963).

Pliocene-Pleistocene of Southern Africa In the Karonga district of northern Malawi, west of Lake Malawi, the Chiwondo Beds comprise a series of pale greenish to brown sandstones and marls with occasional calcareous horizons and vertebrate fossils that resemble elements of the East African Plio-Pleistocene assemblages (Clark, Stephens, and Coryndon 1966). The remains include crocodile and fish, cercopithecoids, a small giraffe, several bovids, hippopotamus, the typical Kanapoi suid Nyanzachoerus jaegeri, a Notochoerus (Mawby 1970), and three primitive proboscideans (Maglio 1970). The age may be a little younger than Kanapoi but is at least as old as the lower part of the Shungura Formation. Overlying the Chiwondo Beds are gravels and sands, termed the Chitimwe Beds, containing "Middle Stone Age" tools and also a fine elephant butchery site (Clark and Haynes 1970). The oldest Pliocene fauna so far found in South Africa comes from phosphate quarries at Langebaanweg, 100 km north of Cape Town (Hendey 1970, 1973, 1976a, b). The major collections are from Ε Quarry, where the phosphate-bearing Varswater Formation overlies a clay of unknown age (Tankard 1975). The 20-m thick formation is divided into four units, the lowest of which (Bed 1) is a beach gravel containing phosphate rock pebbles and yielding marine invertebrates and vertebrates; Bed 2 is a medium-grade sand, mainly nonphosphatic, and contains numerous fossil vertebrates, most of them terrestrial. Bed 3a is a medium-grade phosphate sand of limited lateral extent with plentiful vertebrates, almost exclusively terrestrial. Bed 3b is a somewhat coarser phosphatic sand with few fossils. Overlying the Varswater Formation is from 2 to 40 m of unfossiliferous sand, largely eolian in character. The rich fauna, which has recently been listed by Hendey (1973), represents a remarkable crosssection of vertebrates, including sharks, skates, rays and fish, tortoises, penguins and many other birds, a monachine seal, and numerous terrestrial mammals. Carnivores are more common than usual and include the Eurasiatic Percrocuta and Dinofelis as well as the only remains of bear known south of the Sahara (Agriotherium). There are also several bovids of Eurasiatic aspect, like those of Lothagam

Africa: The Physical Setting

37

and Kanapoi, with which there are distinct faunal resemblances. Hendey (1976b) has also described the first fossil peccary to be found in Africa. The large suid Nyanzachoerus is represented by a variety close to the typical Kanapoi species N. pattersoni. The rhinoceros Ceratotherium praecox is present and is known from Lothagam-1, the Chemeron Formation, the Mpesida Beds, and the Mursi Formation in East Africa (Hooijer 1972, 1973). An Hipparion is found and has been referred to the Kaiso species H. albertense; it is not "Stylohipparion." A gomphothere is present and also Mammuthus subplanifrons, considered by Maglio and Hendey (1970) to be more primitive than the typical Vaal River material. The faunal resemblances thus suggest an age close to that of Kanapoi (i.e., 4.0 m.y.), but Langebaanweg might well be slightly older (Hendey 1970). Interest also attaches to the presence of remains of several small mammals. The overall environment involves interaction between a marine and fluviatile situation and the faunal assemblage implies grassland and riverine woodland. Curiously, both hippopotamus and crocodile are absent. It is possible, although speculative, that an assemblage of material from Klein Zee, near the coast a short distance south of the Orange River (Stromer 1932; Patterson 1965) may be of roughly comparable age. Because of the fossil hominid remains that they contain, great interest attaches to the so-called "cave breccias" of the Transvaal and northern Cape Province. The first discovery was of Australopithecus africanus in the Buxton Lime Quarry at Taung, Cape Province, 130 km north of Kimberley (Dart 1925); but the specimen is that of a child and the first adult was found a decade later at Sterkfontein, 50 km west of Johannesburg, Transvaal (Broom 1936). Further discoveries were made at Kromdraai, 2 km east of Sterkfontein, in 1938, and at Swartkrans, 2 km west of Sterkfontein, in 1947. A complex of small pockets of varying ages occurs at Bolt's Farm, 1 km south of Swartkrans. Another important locality is Makapansgat, in the northern Transvaal, 250 km northeast of Johannesburg. The geology of the Taung deposit is essentially different from that of the Transvaal cave breccias and has been described in some detail by Peabody (1954). The Harts River, a tributary of the Vaal River, flows on one side of a broad valley bounded on the west by a steep escarpment, 30 to 100 m high, formed of Precambrian dolomite limestones. Deltalike aprons have been built out as the result of evaporation of lime-charged water seeping from the cliff. Peabody recognized two major and two minor calc tufas at the Buxton Quarries, each partially eroded before the

38

Cooke

deposition of the next, partly overlapping, carapace of secondary limestone. Fissures and other openings developed in the tufas at various stages and were filled by sandy material, sometimes containing bone or stone implements, and then firmly cemented into a breccia. The Australopithecus skull and the rather scanty fossil material found close to it came from fissure fillings within the oldest (Thabaseek) tufa carapace and Peabody considered the fillings to predate the second (Norlim) carapace. After restudy of the site and the breccia, Butzer (1974b) suggests that the skull was penecontemporary with the second (Norlim) stage of tufa deposition. The Transvaal cave breccias comprise cemented deposits laid down in subsurface caverns or enlarged fissures t h a t were formed by solution in gently dipping Precambrian dolomitic limestones. The process of formation has been considered fully by Brain (1958). While a cavern is cut off from the surface but clear of the ground water table, deposition of flowstone and dipstone occurs, sometimes contaminated by insoluble products resulting from solution of the parent limestone; the result is a white, grey, or banded travertine. As soon as an opening to the surface develops, external debris enters at a slow rate and contaminates the secondary limestone deposit, producing a "Phase 1" breccia t h a t may be rich in bones. With enlargement of the surface opening, external soil is introduced and the deposit soon changes to a clastic one with a calcareous cement, or "Phase 2" breccia. Although not as rich in bone as some of the Phase 1 breccias, it is the brownish or pinkish Phase 2 deposits that have been the main source of the tens of thousands of fossils recovered from these sites. The upper parts of the cave filling are often full of fragments or blocks of dolomitic limestone resulting from collapse of the edge of the opening or of the cave roof itself. Subsequent to their primary filling and cementation, the breccias themselves may be subject to erosion and cavities or fissures in them may be filled by later material. Undermining of the floor may result in collapse during filling, so the depositional history can be complex and difficult to interpret. The early breccia deposits have now lost most or all of their original roofs and the firmly cemented pinkish-brown breccia, originally formed at depth, is now exposed on the surface. The age of any particular deposit is dependent on the date of opening to the surface, which determines the start of its period of sampling of the environment. Correlation between sites and relative dating is dependent at present solely on the faunal material. Approximately 170 mammalian species are repre-

Africa: The Physical Setting

sented in the cave breccias and there are a few reptiles and bird remains. Various analyses of the faunas have shown t h a t the assemblages from Makapansgat and Sterkfontein have much in common and differ from the assemblages from Swartkrans and the Kromdraai "A" faunal site 2 (Ewer 1957; Wells 1962; Ewer and Cooke 1964). These are regarded as representing two faunal "stages" (Wells 1962), now termed the "Sterkfontein Faunal Span" and "Swartkrans Faunal Span" respectively (Cooke 1967). The fauna of the Sterkfontein Extension Site, where stone implements occur, is placed in the Swartkrans Faunal Span (Wells 1962; Cooke 1967); Vrba (1974) confirms the separation and also shows t h a t a still younger breccia occurs at Sterkfontein. Wells (1969) has argued that the Kromdraai "A" fauna should be made an intermediate unit or be placed in the Vaal-Cornelia Faunal Span; this is accepted in the accompanying correlation chart (table 2.1). (The proposal by Hendey [1974] for establishing mammal "ages" for southern Africa is considered premature.) The position of Taung in relation to the Transvaal sites is difficult to assess both on account of the scantiness of the Taung fauna and the existence of some species at Taung t h a t are not present at the other sites, probably because of ecological differences. Of the 20 species at Taung that do occur at other sites, more than half are present at Makapansgat and Sterkfontein and less t h a n half at Swartkrans. Taung is thus tentatively placed in an intermediate position and it seems very improbable that it is younger than Swartkrans. Correlation between the South African cave breccias and the East African sequence is even more difficult, but at present it seems highly probable that both Sterkfontein and Makapansgat are older than Olduvai Bed I and Swartkrans is roughly coeval with Bed I and at least part of Bed II. Further data are required before correlation can be effective. The Vaal River drains a large area in the southern Transvaal and Orange Free State, together with a small area in the northeastern Cape Province. Terrace gravels, at altitudes up to 90 m above the river and at distances up to 10 km from it, have been worked for diamonds for more than a century (du Toit 1907). The area was studied in some detail by Söhnge, Visser, and van Riet Lowe (1937), who dis2 The Kromdraai australopithecine skull came from a nearby site (Kromdraai "B") with a scanty fauna, not necessarily of the same age. At Swartkrans there are two breccias, an earlier "pink" breccia and a secondary "brown" breccia that may be considerably later (Brain, pers. comm.). It is possible that the analyses on available identifications include, in error, some of the later elements.

Cooke

Table 2.1

Africa:

The Physical

Setting

39

Tentative grouping of main Plio-Pleistocene Deposits in Southern Africa.

Faunal Unit 'Recent''

Florisbad faunal span

Cornelia faunal span

Vaal River Basin

Cave deposits

Open sites

Central Africa

Minor erosion; soil formation Alluviation of flood terraces Dissection of earlier fill

Numerous caves

Surface

Nachikufu, etc.

Wonderwerk, etc.

Vlakkraal

Vertizol development Silt and loam deposition Alluviation of tributary sands and gravels Silts in Vaal floodplain Dissection of earlier fill

Mumbwa cave

Florisbad Cave of Hearths

Calcereous paleosol

Hopefield

Alluvium in valleys

Cornelia

Low-level terrace gravels "Younger Gravels"

Kromdraai

Bedrock dissection Swartkrans faunal span

Local calcification Alluviation and reworking of high-level "Older Gravels"

Sterkfontein faunal span

Aggradation of original high-level "Older Gravels" in three stages

Langebaanweg faunal span

Bedrock dissection of major valleys ?Upper Miocene/Pliocene

Broken Hill cave Twin Rivers breccia Chelmer

Younger terrace gravels

X! 0)

ft

ε

Swartkrans Sterkfontein Extension Sterkfontein Makapansgat

tinguished three sets of deposits, designated "older gravels," "younger gravels," and "youngest gravels," which they regarded as reflecting climatic changes. Their conclusions have been discussed or criticized by other workers but the basic stratigraphy is sound and the whole situation has been reconsidered recently by Butzer et al. (1973), whose interpretation is generally followed here (table 2.1). The older gravels lie essentially on three erosional platforms, the highest of which is approximately 90 m above present river level. The original gravels were rich in boulders of diabase but these were decomposed and the weathered products often removed by extensive eluviation that has resulted in considerable redistribution of the resistant elements as "potato gravels" with an enriched diamond content (Cooke 1947). Although only one fossil has ever been recorded as coming from these gravels, there are a number of archaic elements in the younger gravels that are believed to have been derived from them.

tUD Ο c υ 3 e « Ε ^ 53

Older terrace gravels

Ο m

Oldest terrace gravels Langebaan phosphate sands

Chiwondo Beds

Following bedrock dissection of more than 10 m, the younger gravels were deposited on the flanks of the present river and three units can be recognized, styled from the base upward, as A, B, and C. Nearly all the fossil material comes from these deposits (Cooke 1949; Wells 1964), principally from units Β and C, and the gravels also yield middle to late Acheulian tools in considerable abundance. The fill of younger gravels was then dissected and subsequently covered disconformably by "calcified sands," the age of which is not at present clear. Some of the fossils have come from the base of the sands. Tributary valleys are filled by or have flanking exposures of a series of alluvial silts, sands, and gravels to which the name Riverton Formation has been given (Butzer et al. 1973). Middle Stone Age tools occur in these beds and Later Stone Age tools on the surface. Fossil remains are rare. Within the Vaal River Basin are a number of other important deposits of which the oldest are the

40

Cooke

Cornelia Beds, first described by Van Hoepen in 1930 and recently considered in detail by Butzer (1974a). The deposits lie in the basin of the Skoonspruit, a tributary of the Vaal River and situated some 50 km southwest of Standerton. The total thickness of the type site is approximately 15 to 20 m and consists of clays, silts, and loam overlying a thin basal gravel or rubble layer from which have come weathered flakes and crude hand axes. The major fossiliferous unit, 1 to 8 m thick, lies upon this basal part of the sequence and is covered by an almost equal thickness of barren clays and silts, together forming Member 1; a second unit (Member 2) was deposited after an intervening erosion interval and is barren. The artifacts are generally crude and the assemblage is comparable with industries of upper Acheulian affinities from Olduvai Gorge, Olorgesailie, and other sites (Clark 1974). The fauna is equivalent to part of the younger gravels assemblage but also has remarkable links with that from the upper part (Beds III/IV) of the Olduvai sequence (Cooke 1974). This fauna also resembles that from Elandsfontein (Hopefield) in the southwestern Cape Province, where an old deposit of swampy character was exposed by the local removal of the sandy cover by wind action (Singer 1954; Mabbutt 1956; Butzer 1973). Acheulian tools occur with the fossils at Hopefield and include a hominid jaw and partial cranium (Singer and Wymer 1968). Faunal assemblages clearly younger than that of Cornelia have been found at many sites in southern Africa and include only a few elements that are now extinct. Notable are the thermal springs at Florisbad and Vlakkraal in which Middle Stone Age artifacts occur together with fossil bones and teeth. The Florisbad site yielded a human skull and the lower peaty layer is older than 48,000 years B.P. The Vlakkraal fauna (Wells, Cooke, and Malan 1941) is very similar to that of Florisbad (see Cooke 1963) and forms the basis for a "Florisbad-Vlakkraal Faunal Span" following a "Vaal-Cornelia Faunal Span" (Cooke 1967). A deposit at Chelmer in southern Rhodesia is also placed here and it is probable that the Broken Hill site in Zambia is of roughly similar age. Numerous cave deposits containing some extinct elements are widespread in association with Middle Stone Age artifacts, whereas the fauna associated with Later Stone Age is typically one of living species. Klein (1974) has summarized the terminal Pleistocene extinctions in the southern Cape Province and this pattern may be characteristic of a wider area. Intercorrelations between South Africa and East Africa and between the latter and North Africa are

Africa: The Physical

Setting

at present possible only in a broad sense and the principal inferences have been shown in figure 2.6. However, studies are advancing rapidly at the present time and more links between the major regions are coming to light, so it should not be many years before a much firmer framework of correlation is established. The overview of mammal taxa presented in this volume is an important step in this direction.

References Abadie, J.; Barbeau, J.; and Coppens, Y. 1959. Une faune de vertebres villafranchiens au Tchad. C. R. Hebd. S0anc. Acad. Sei. 248:3328-3330

Ameur, R. C.; Jaeger, J.-J.; and Michaux, J. 1976. Radiometric age of early Hipparion fauna in northwest Africa. Nature 261:38-39. Andrews, C. W. 1902. Note on a Pliocene vertebrate fauna from the Wadi Natrun, Egypt. Geol. Mag. (4) 9:433-439. 1906. A descriptive

catalogue of the Tertiary

Verte-

brata of the Fayüm, Egypt. London: British Museum. 1911. On a new species of Dinotherium (Dinotherium hobleyi) from British East Africa. Proc. Zool. Soc. London 1911:943-945. Arambourg, C. 1946. Anancus osiris, un mastodonte nouveau du Pliocene inferieur d'Egypte. Bull. Soc. Geol. Fr. (5) 15:479-495. 1947. Contribution a l'etude geologique et paleontologique du bassin du Lac Rodolphe et de la Basse vallee de l'Omo. Deuxieme partie, Paleontologie. Mission Scient. Omo. Geol.-Anthrop.

1:479-495.

1951. Observations sur les couches a Hipparion de la vallee de l'Oued el Hammam (Algerie) et sur l'epoque d'apparition de la faune de vertebres dite "Pontienne." C. R. Hebd. Seanc. Acad. Sei. 232 (26):2464-2466.

1952. Note preliminaire sur quelques elephants fossiles de Berberie. Bull. Mus. Hist. Nat., Paris, Ser. 2,

(24):407-418. 1954a. La faune a Hipparion de l'Oued el Hammam (Algerie). Int. Geol. Congr., 19th Sess. 21:294-302. 1954b. L'hominien fossile de Ternifine (Algerie). C. R. Hebd. Seanc. Acad. Sei. 239 (15):893-895.

1962. Les faunes mammalogique du Pleistocene circummediterranean. Quaternaria 6:97-109. 1970. Les vertebres du Pleistocene de l'Afrique du Nord. Mem. Mus. Natl. Hist. Nat., Ser. 7, 10:1-126. Arambourg, C., and Burollet, P. F. 1962. Restes de vertebres oligocenes en Tunisie centrale. C. R. Somm. Seanc. Geol. Fr. 2:42-43.

Arambourg, C., and Choubert, G. 1965. Les faunes de mammiferes de l'etage Moghrebien du Maroc occidental. Notes Serv. Geol. Maroc 185:29-33.

Arambourg, C., and Magnier, P. 1961. Gisements de vertebres dans la bassin tertiäre de Syrte (Libye). C. R. Hebd. Seanc. Acad. Sei. 252 (8):1181-1183.

Arenes, J., and Depape, G. 1953. Etude paleobotanique. In C. Arambourg. Contribution a l'etude des flores fossiles

Cooke

quaternaire de l'Afrique du Nord. Arch. Mus. Natl. Hist. Nat. Paris, Ser. 7, 2:7-85. Baker, Β. H.; Möhr, P. Α.; and Williams, L. A. J. 1972. Geology of the eastern Rift System. Geol. Soc. Amer., Spec. Pap. 136:1-67. Baker, Β. H.; Williams, L. A. J.; Miller, J. Α.; and Fitch, F. J. 1971. Sequence and geochronology of the Kenya Rift volcanics. Tectonophysics 11:191-215. Baker, Β. H., and Wohlenburg, J. 1971. Structure and evolution of the Kenya Rift Valley. Nature 229:538-542. Beadnell, H. J. L. 1901a. Decouvertes geologique recentes dans la vallee du Nil et le desert Libyen. VIII Int. Geol. Congr., Paris 2:839-866. 1901b. The Fayüm depression: a preliminary notice of the geology of a district in Egypt containing a new palaeogene vertebrate fauna. Geol. Mag. (4) 8:540-546. 1902. A preliminary note on Arsinoitherium zitteli Beadn. from the Upper Eocene strata of Egypt. Cairo: Survey Dept., Public Works Ministry. 1905. The topography and geology of the Fayüm Province of Egypt. Cairo: Survey Dept., Public Works Ministry. Berggren, W. A. 1971. Tertiary boundaries and correlations. In Β. M. Funnell, and W. R. Riedel, eds. The micropalaeontology of oceans. Cambridge: Cambridge University Press, pp. 693-809. Biberson, P. 1961. Le cadre paleogeographique de la prehistoire du Maroc atlantique. Publ. Serv. Antiq. Maroc 16:1-235. 1964. La place des hommes du Paleolithique Marocain dans la Chronologie du Pleistocene atlantique. Anthropologie 68 (5-6):475-526. 1967. Some aspects of the lower Palaeolithic of northwest Africa. In W. W. Bishop, and J. D. Clark, eds. Background to evolution in Africa. Chicago: University of Chicago Press, pp. 447-475. Biely, Α.; Rakus, M.; Robinson, P.; and Salaj, J. 1972. Essai de correlation des formations Miocene au sud de la Dorsale Tunisienne. Notes Serv. Geol. Tunis 38 (7):7392. Bishop, W. W. 1963. The later Tertiary and Pleistocene in eastern equatorial Africa. In F. C. Howell, and F. Bourliere, eds. African ecology and human evolution. Chicago: University of Chicago Press, pp. 246-275. 1965. Quaternary geology and geomorphology in the Albertine Rift Valley, Uganda. Geol. Soc. Am. Spec. Pap. 84:293-321. 1967. The later Tertiary in East Africa—volcanics, sediments, and faunal inventory. In W. W. Bishop, and J. D. Clark, eds. Background to evolution in Africa. Chicago: University of Chicago Press, pp. 3154. 1968. The evolution of fossil environments in East Africa. Trans. Leic. Lit. Phil. Soc. 62:22-44. 1972. Stratigraphic succession "versus" calibration in East Africa. In W. W. Bishop, and J. A. Miller, eds. Calibration ofhominoid evolution. Edinburgh: Scottish Academic Press, pp. 219-246. Bishop, W. W. and Chapman, G. R. 1970. Early Pliocene

Africa: The Physical Setting

41

sediments and fossils from the northern Kenya Rift Valley. Nature 226:914-918. Bishop, W. W.; Chapman, G. R.; Hill, Α.; and Miller, J. A. 1971. A succession of Cainozoic vertebrate assemblages from the northern Kenya Rift Valley. Nature 233:389394. Bishop, W. W.; Miller, J. Α.; and Fitch, F. J. 1969. New potassium-argon age determinations relevant to the Miocene fossil mammalian sequence in East Africa. Am. J. Sei. 267 (6):669-699. Bishop, W. W., and Pickford, Μ. H. L. 1975. Geology, fauna and palaeoenvironments of the Ngorora Formation, Kenya Rift Valley. Nature 254:185-192. Bishop, W. W.; Pickford, Μ. H. L.; and Hill, A. 1975. New evidence regarding the Quaternary geology, archaeology and hominids of Chesowanja, Kenya. Nature 258:204-208. Bishop, W. W., and Trendall, A. F. 1967. Erosion surfaces, tectonics and volcanic activity in Uganda. Quart. J. Geol. Soc. London 122:385-420. Black, C. C. 1972. A new species ofMerycopotamus (Artiodactyla: Anthracotheriidae) from the late Miocene of Tunisia. Notes Serv. Geol. Tunis 37:5-39. Blanckenhom, M. 1901. Neues zur Geologie und Paläontologie Aegyptens—IV. Das Pliozän—und Quartärzeitalter in Aegypten ausschliesslich des Rothen Meergebietes. Zeits. Deutsch. Geol. Ges. 53:307-502. Bowen, B. E., and Vondra, C. F. 1973. Stratigraphical relationships of the Plio-Pleistocene deposits, East Rudolf, Kenya. Nature 242:391-393. Brain, C. K. 1958. The Transvaal ape-man-bearing cave deposits. Transv. Mus. Mem. 11:1-131. Brock, Α., and Isaac, G. LL. 1974. Palaeomagnetic stratigraphy and chronology of hominid-bearing sediments east of Lake Rudolf, Kenya. Nature 247:334-348. Broom, R. 1936. A new fossil anthropoid skull from South Africa. Nature 138:486-488. Brown, F. H.; Heinzelin, J. de; and Howell, F. C. 1970. Pliocene/Pleistocene formations in the lower Omo Basin, southern Ethiopia. Quaternaria 13:247-268. Butzer, K. W. 1973. Re-evaluation of the geology of the Elandsfontein (Hopefield) site, southwestern Cape, South Africa. S. Afr. J. Sei. 69:234-238. 1974a. Geology of the Cornelia Beds, northwestern Orange Free State. Nas. Mus. Bloemfontein Mem. 9:7-32. 1974b. Paleoecology of South African australopithecines: Taung revisited. Curr. Anthrop. 15:367382; 413-416. Butzer, K.; Brown, F. H.; and Thurber, D. L. 1969. Horizontal sediments of the lower Omo Valley: The Kibish Formation. Quaternaria 11:15-29. Butzer, K.; Helgren, D. M.; Fock, G. J.; and Stuckenrath, R. 1973. Alluvial terraces of the lower Vaal River, South Africa: a reappraisal and reinvestigation. J. Geol. 81:341-362. Butzer, K. W., and Thurber, D. L. 1969. Some late Cenozoic sedimentary formations of the lower Omo Basin. Nature 222:1132-1143.

42

Cooke

Carney, J.; Hill, Α.; Miller, J. Α.; and Walker, A. 1971. Late australopithecine from Baringo District, Kenya. Nature 230:509-514. Choubert, G. 1953. Les rapports entre les formations marines et continentale quaternaries. Actes 4 Internat. Congr. INQUA (Rome-Pisa):576-590. Choubert, G., and Ennouchi, E. 1946. Premieres preuves paleontologique de la presence du Pontien au Maroc. C. R. Soc. Geol. Fr. 1946:207-208. Clark, J. D. 1974. The stone artefacts from Cornelia, O.F.S., South Africa. Nas. Mus. Bloemfontein Mem. 9: 33-61. Clark, J. D., and Haynes, C. V. 1970. An elephant butchery site at Mwanganda's village, Karonga, Malawi, and its relevance for Palaeolithic archaeology. World Archaeol. I (3):390-411. Clark, J. D.; Stephens, Ε. Α.; and Coryndon, S. C. 1966. Pleistocene fossiliferous lake beds of the Malawi (Nyasa) Rift: a preliminary report. Am. Anthrop. 68 (2):46-87. Cole, S. 1963. The Prehistory of East Africa. New York: Macmillan and Co. Cooke, Η. B. S. 1947. The development of the Vaal River and its deposits. Trans. Geol. Soc. S. Afr. 49:243-262. 1949. Fossil mammals of the Vaal River deposits. Mem. Geol. Surv. S. Afr. 35 (3):1-117. 1958. Observations relating to Quaternary environments in East and southern Africa. Trans. Geol. Soc. S. Afr. 60 (Annexure):l-73. 1963. Pleistocene mammal faunas of Africa, with particular reference to southern Africa. In F. C. Howell, and F. Bourliere, eds. African ecology and human evolution. New York: Viking Fund Publications in Anthrop. No. 36:65-116. 1967. The Pleistocene sequence in South Africa and problems of correlation. In W. W. Bishop, and J. D. Clark, eds. Background to evolution in Africa. Chicago: University of Chicago Press. Pp. 175-184. 1974. The fossil mammals of Cornelia, O.F.S., South Africa. Nas. Mus., Bloemfontein Mem. 9:63-84. Cooke, Η. B. S., and Coryndon, S. C. 1970. Pleistocene mammals from the Kaiso Formation and other related deposits in Uganda. In L. S. B. Leakey and R. J. G. Savage, eds., Fossil Vertebrates of Africa, Edinburgh: Academic Press, vol. 2, pp. 107-224. Cooke, Η. B. S., and Ewer, R. F. 1972. Fossil Suidae from Kanapoi and Lothagam, northwestern Kenya. Bull. Mus. Comp. Zool., Harvard 143 (3): 149-295. Cooke, Η. B. S., and Maglio, V. J. 1972. Plio-Pleistocene stratigraphy in East Africa in relation to proboscidean and suid evolution. In W. W. Bishop, and J. A. Miller, eds. Calibration ofhominoid evolution. Edinburgh: Scottish Academic Press, pp. 303-329. Coppens, Y. 1960. Le Quaternaire fossilifere de Koro-Toro (Tchad). Resultats d'une premiere mission. C. R. Hebd. Seanc. Acad. Sei. 251 (21):2385-2386. 1965. L'hominien du Tchad. C. R. Hebd. Seanc. Acad. Sei. 260:2869-2871. 1967. Les faunes de vertebres Quaternaires du Tchad. In W. W. Bishop, and J. D. Clark, eds. Back-

Africa: The Physical

Setting

ground to evolution in Africa. Chicago: University of Chicago Press, pp. 89-99. 1971a. Une nouvelle espece de suide du Villafranchien de Tunisie, Nyanzachoerus jaegeri sp. nov. C. R. Hebd. Seanc. Acad. Sei.Ser. D, 272 (26):32643267. 1971b. Les vertebres Villafranchiens de Tunisie: gisements nouveaux, signification. C. R. Hebd. Seanc. Acad. Sei., Ser. D, 273 (l):51-54. Coppens, Y., and Howell, F. C. 1974. Les faunes mammiferes fossiles des formations Plio-Pleistocene de l'Omo en Ethiopie (Proboscidea, Perissodactyla, Artiodactyla). C. R. Hebd. Seanc. Acad. Sei. Ser. D, 278:2275-2278. Crompton, A. W. 1964. A preliminary description of a new mammal from the Upper Triassic of South Africa. Proc. Zool. Soc. Lond. 142:441-454. 1974. The dentitions and relationships of the southern African Triassic mammals, Erythrotherium parringtoni and Megazostrodon rudnerae. Bull. Brit. Mus. Nat. Hist. (Geol.) 24 (7):400-437. Crompton, A. W., and Jenkins, F. A. 1968. Molar occlusion in late Triassic mammals. Biol. Rev. 43:427-458. Curtis, G. H.; Drake, T.; Cerling, Τ. E.; Cerling, B. L.; and Hampel, J. H. 1975. Age of KBS Tuff in Koobi Fora Formation, East Rudolf, Kenya. Nature 258:395-398. Curtis, G. H., and Hay, R. L. 1972. Further geological studies and potassium-argon dating at Olduvai Gorge and Ngorongoro Crater. In W. W. Bishop, and J. A. Miller, eds. Calibration of hominoid evolution. Edinburgh: Scottish Academic Press, pp. 289-302. Dames, W. 1883. Ueber eine Tertiäre Wirbeltier Fauna von der westlichen Insel des Birket el Qurun in Fayüm (Aegypten). S. B. Preuss. Akad. Wiss., Berlin 1883:129153. 1894. Ueber Zeuglodonten aus Aegypten und die Beziehungen der Archaeoceten zu den übrigen Cetaceen. Paläont. Abh., Jena (N.F.) 1 (5):l-36. Dart, R. A. 1925. Australopithecus africanus: the man-ape of South Africa. Nature 115:195-199. Day, Μ. Η. 1975. Hominid postcranial remains from the East Rudolf succession. In Y. Coppens, F. C. Howell, G. LL. Isaac, and R. E. F. Leakey, eds. Earliest man and environments in the Lake Rudolf Basin. Chicago: University of Chicago Press, pp. 507-521. Dietrich, W. O. 1927. Brancatherulum n.g.—ein Proplacentalier aus dem obersten J u r a des Tendaguru in Deutsch-Ostafrika. Cbl. Min. Geol. Paläont. 1927B:423426. 1942. Ältestquartäre Säugetiere aus der südlichen Serengeti, Deutsch-Ostafrika. Paläontogr. 94A:43-133. 1950. Fossile Antilopen und Rinder Äquatorial afrikas (Material der Kohl-Larsen'schen Expeditionen). Paläontogr. 99A:l-62. Dietz, R. S., and Holden, J. C. 1970. Reconstruction of Pangaea: breakup and dispersion of continents, Permian to Recent. J. Geophys. Res. 75:4939-4956. du Toit, A. L. 1907. Geological survey of the eastern part of Griqualand West. Geol. Comm. Cape of Good Hope: 11th Ann. Rpt. (1906):87-176. Evernden, J. F., and Curtis, G. H. 1965. The potassium-

Cooke

argon dating of late Cenozoic rocks in East Africa and Italy. Curr. Anthrop. 6 (4):343-364 (with comments pp. 364-385). Ewer, R. F. 1957. Faunal evidence on the dating of the Australopithecinae. Proc. 3rd Pan-Afr. Congr. Prehist. (Livingstone, 1955):135-142. Ewer, R. F., and Cooke, Η. B. S. 1964. The Pleistocene mammals of southern Africa. In D. H. S. Davis, ed. Ecological studies in southern Africa. Monographiae Biologicae 14:35-48. Fitch, F. J.; Hooker, P. J.; and Miller, J. A. 1976. 40 Ar/ 39 Ar dating of the KBS Tuff in Koobi Fora Formation, East Rudolf, Kenya. Nature 263:740-744. Fitch, F. J., and Miller, J . A. 1970. Radiometric age determinations of Lake Rudolf artefact sites. Nature 226: 226-228. Forsten, A. M. 1972. Hipparion primigenium from southern Tunisia. Notes Serv. Geol. Maroc 5 (l):7-28. Franz, H. 1967. On the stratigraphy and evolution of climate in the Chad Basin during the Quaternary. In W. W. Bishop, and J. D. Clark, eds.Background to evolution in Africa. Chicago: University of Chicago Press, pp. 273-282. Fuchs, V. Ε. 1950. Pleistocene events in the Baringo Basin. Geol. Mag. 87 (3): 148-174. Gentry, A. W. 1970. The Bovidae (Mammalia) of the Fort Ternan fossil fauna. Fossil Vertebrates of Africa 2: 243-323. Gorodiski, Α., and Lavocat, R. 1953. Premiere decouverte de Mammiferes dans le Tertiare (Lutetien) du Senegal. C. R. Somm. Geol. Fr. 15:314-316. Greenwood, P. H. 1972. Fish fossils from the late Miocene of Tunisia. Notes Serv. Geol. Tunis 37:41-72. Harris, J. M., and Watkins, R. 1974. New early Miocene vertebrate locality near Lake Rudolf, Kenya. Nature 252:576-577. Hay, R. L. 1963. Stratigraphy of Beds I through IV, Olduvai Gorge, Tanganyika. Science 139:829-833. 1971. Geologic background of Beds I and II: stratigraphic summary. In M. D. Leakey, Olduvai Gorge, volume 3: Excavation in Beds I and II, 1960-1963. Cambridge: Cambridge University Press. 1976. Geology of the Olduvai Gorge. Berkeley: University of California Press. Heinzelin, J. de, and Brown, F. H. 1969. Some early Pleistocene deposits of the lower Omo Valley: the Usno Formation. Quaternaria 11:32-46. Hendey, Q. B. 1970. The age of the fossiliferous deposits at Langebaanweg, Cape Province. Ann. S. Afr. Mus. 56 (3):119-131. 1973. Fossil occurrences at Langebaanweg, Cape Province. Nature 244:13-14. 1974. Faunal dating of the late Cenozoic of southern Africa, with special reference to the carnivores. Quatern. Res. 4:149-161. 1976a. The Pliocene fossil occurrences at "E" Quarry, Langebaanweg, South Africa. Ann. S. Afr. Mus. 69 (9):215-247. 1976b. Fossil peccary from the Pliocene of South Africa. Science 192:787-789.

Africa: The Physical Setting

43

Hoepen, E. C. N. van 1930. Fossiele perde van Cornelia, O.V.S. Paleont. Navors. Nas. Mus. Bloemfontein 2 (2):124. Hooijer, D. A. 1963. Miocene Mammalia of Congo. Ann. Mus. R. Congo Beige, Ser. 8, 46:1-77. 1970. Miocene Mammalia of Congo, a correction. Ann. Mus. R. Congo Beige, Ser. 8, 67:163-167. 1972. A late Pliocene rhinoceros from Langebaanweg, Cape Province. Ann. S. Afr. Mus. 59 (9):151-191. 1973. Additional Miocene to Pleistocene rhinoceroses of Africa. Zool. Med. Leiden 46:149-177. Hooijer, D. Α., and Patterson, B. 1972. Rhinoceroses from the Pliocene of northwestern Kenya. Bull. Mus. Comp. Zool., Harvard 144 (l):l-26. Hopwood, A. T. 1929. New and little known mammals from the Miocene of Africa. Am. Mus. Novit. 344:1-9. Howell, F. C., and Coppens, Y. 1974. Les faunes des mammiferes fossiles des formations Plio-Pleistocene de l'Omo en Ethiopie (Tubulidentata, Hyracoidea, Lagomorpha, Rodentia, Chiroptera, Insectivora, Carnivora, Primates). C. R. Hebd. Seanc. Acad. Sei. Ser. D, 278:24212424. Hurford, A. J.; Gleadow, A. J. W.; and Naeser, C. W. 1976. Fission-track dating of pumice from the KBS Tuff, East Rudolf, Kenya. Nature 263:738-740. Isaac, G. LL. 1965. The stratigraphy of the Peninj Beds and the provenance of the Natron Australopithecine mandible. Quaternaria 7:101-130. 1966. The geological history of the Olorgesailie area. Adas 5 Congr. Panafr. Prehist. de Ε studio del Cuaternario 2:125-134. 1967. The stratigraphy of the Peninj Group—early middle Pleistocene formations west of Lake Natron, Tanzania. In W. W. Bishop, and J. D. Clark, eds. Background to evolution in Africa. Chicago: University of Chicago Press, pp. 229-258. Isaac, G. LL., and Curtis, G. H. 1974. Age of early Acheulian industries from the Peninj Group, Tanzania. Nature 249:624-627. Isaac, G. LL.; Harris, J. W. K.; and Crader, D. 1975. Archaeological evidence from the Koobi Fora Formation. In Y. Coppens, F. C. Howell, G. LL. Isaac, and R. E. F. Leakey, eds. Earliest man and environments in the Lake Rudolf Basin. Chicago: University of Chicago Press, pp. 533-551. Isaac, G. LL.; Merrick, Η. V.; and Nelson, C. M. 1972. Stratigraphic and archaeological studies in the Lake Nakuru basin. In Ε. M. van Zinderen Bakker, ed. Palaeoecology of Africa. Cape Town: Balkema. vol. 6, pp. 225-232. Jaeger, J.-J.; Michaux, J.; and David, B. 1973. Biochronologie du Miocene moyen et du superieur continental du Maghreb. C. R. Hebd. Seanc. Acad. Sei. Ser. D, 277: 2477-2480. Johanson, D. C., and Taieb, M. 1976. Plio-Pleistocene hominid discoveries in Hadar, Ethiopia. Nature 260: 293-297. Kent, P. E. 1941. The Pleistocene beds of Kanam and Kanjera, Kavirondo, Kenya. Geol. Mag. 79:117-132. 1944. The Miocene beds of Kavirondo, Kenya. Quart. J. Geol. Soc. Lond. 100:85-116.

44

Cooke

King, B. C., and Chapman, G. R. 1972. Volcanism of the Kenya Rift Valley. Phil. Trans. Roy. Soc. London (A) 271:185-208. King, L. C. 1951. South African Scenery. Edinburgh: Oliver and Boyd. 1962. The morphology of the earth. Edinburgh: Oliver and Boyd. Klein, R. G. 1974. Provisional statement on terminal Pleistocene mammalian extinctions in the Cape biotic zone (Southern Cape Province, South Africa). S. Afr. Archaeol. Soc., Goodwin Series 2:39-45. Leakey, L. S. B. 1951. Olduvai Gorge. Cambridge: Cambridge University Press. 1965. Olduvai Gorge 1951-1961. Volume V.Fauna and background. Cambridge: Cambridge University Press. 1967. Notes on the mammalian faunas from the Miocene and Pleistocene of East Africa. In W. W. Bishop, and J. D. Clark, eds. Background to evolution in Africa. Chicago: University of Chicago Press, pp. 7-29. Leakey, L. S. B., and Solomon, J. D. 1931. East African archaeology. Nature 124:9. Leakey, M.; Tobias, P. V.; Martyn, J. M.; and Leakey, R. E. F. 1969. An Acheulean industry with prepared core technique and the discovery of a contemporary hominid mandible at Lake Baringo, Kenya. Proc. Prehist. Soc. Lond. 3:48-76. Leakey, M. D. 1971. Olduvai Gorge, Volume 3: Excavations in Beds I and II, 1960-1963. Cambridge: Cambridge University Press. Leakey, M. D.; Hay, R. L.; Curtis, G. H.; Drake, R. E.; Jackes, Μ. K.; and White, T. D. 1976. Fossil hominids from the Laetolil Beds. Nature 262:460-466. Leakey, M. D.; Hay, R. L.; Thurber, D. L.; Protsch, R.; and Berger, R. 1972. Stratigraphy, archaeology, and age of the Ndutu and Naisiusu Beds, Olduvai Gorge, Tanzania. World Archaeol. 3 (3):328-341. Leakey, R. E. F. 1970. Fauna and artefacts from a new Plio-Pleistocene locality near Lake Rudolf in Kenya. Nature 226:223-224. 1975. An overview of the East Rudolf Hominidae. In Y. Coppens, F. C. Howell, G. LI. Isaac, and R. E. F. Leakey, eds. Earliest man and environments in the Lake Rudolf Basin. Chicago: University of Chicago Press, pp. 476-483. 1976. New hominid fossils from the Koobi Fora Formation in northern Kenya. Nature 261:574-576. Le Gros Clark, W. E„ and Leakey, L. S. B. 1951. The Miocene Hominoidea of East Africa. Fossil Mammals of Africa, 1. British Museum (Nat. Hist.), London. Lepersonne, J. 1949. Le fosse tectonique Lac Albert-Semliki-Lac Edouard. Resume des observations geologique effectuees en 1938-1939-1940. Ann. Soc. Geol. Belg. 77:M 1-92. Mabbutt, J. A. 1956. The physiographical surface geology of the Hopefield fossil site. Tran. Roy. Soc. S. Afr. 35:2158. Maglio, V. J. 1970. Early Elephantidae of Africa and a tentative correlation of African Plio-Pleistocene deposits. Nature 225:328-332.

Africa: The Physical Setting

1972. Vertebrate faunas and chronology of hominid-bearing sediments east of Lake Rudolf, Kenya. Nature 239:379-385. 1973. Origin and evolution of the Elephantidae. Trans. Am. Phil. Soc., N.S. 63:1-149. Maglio, V. J., and Hendey, Q. B. 1970. New evidence relating to the supposed stegolophodont ancestry of the Elephantidae. S. Afr. Archaeol. Bull. 25:85-87. Martyn, J. M. 1967. Pleistocene deposits and new fossil localities in Kenya (with a note on the hominid remains by P. V. Tobias). Nature 215:476-479. 1969. Notes on the geology of the Kapthurin Beds. In M. Leakey et al. 1969, Proc. Prehist. Soc. Lond. 3:4876. Mawby, J. E. 1970. Fossil vertebrates from northern Malawi: preliminary report. Quaternaria 13 (2):319-324. McCall, G. J. H. 1958. Geology of the Gwasi area. Geol. Surv. Kenya, Report 45:1-88. McCall, G. J. H.; Baker, Β. H.; and Walsh, J. 1967. Late Tertiary and Quaternary sediments of the Kenya Rift Valley. In W. W. Bishop, and J. D. Clark, eds. Background to evolution in Africa. Chicago: University of Chicago Press, pp. 191-220. Monod, T. 1963. The late Tertiary and Pleistocene in the Sahara and adjacent southerly regions. In F. C. Howell, and F. Bourliere, eds. African ecology and human evolution. New York: Viking Fund Publications in Anthropology. 36:117-229. Nilsson, E. 1929. Preliminary report on the Quaternary geology of Mount Elgon and some parts of the Rift Valley. Geologiska Foren. Forhand. Stockholm 51:253-261. Patterson, B. 1965. The fossil elephant shrews (family Macroscelididae) Bull. Mus. Comp. Zool., Harvard 133 (6):295-335. 1966. A new locality for early Pleistocene fossils in northwestern Kenya. Nature 212:577-578. Patterson, B.; Behrensmeyer, A. K.; and Sill, W. D. 1970. Geology and faunal correlations of a new Pliocene locality in northwestern Kenya. Nature 226:918-921. Peabody, F. E. 1954. Travertines and cave deposits of the Kaap escarpment of South Africa and the type locality of Australopithecus africanus Dart. Bull. Geol. Soc. Am. 65:671-706. Petrocchi, C. 1943. Sahabi, eine neue Seite in der Geschieht der Erde. N. Jb. Min. Geol. Paläont. 1943 B:l-9. Pickford, M. 1975. Late Miocene sediments and fossils from the northern Kenya Rift Valley. Nature 256:279284. Rich, P. V. 1972. A fossil avifauna from the upper Miocene Beglia Formation of Tunisia. Notes Serv. Geol. Tunis 35:29-66. Robinson, P. 1972. Pachytragus solignaci, a new species of Caprine Bovid from the late Miocene Beglia Formation of Tunisia. Notes Serv. Geol. Tunis 37:73-94. Robinson, P., and Black, C. C. 1969. Note preliminaire sur les vertebres fossiles du Vindobonien (formation Beglia) du Bled Douarah, Gouvernerat de Gafsa, Tunisie. Notes Serv. Geol. Tunis 31:67-70. Saggerson, E. P., and Baker, Β. H. 1965. Post-Jurassic erosion surfaces in eastern Kenya and their deformation in

Cooke

relation to rift structure. Quart. J. Geol. Soc. Lond. 121:51-68. Savage, R. J. G. 1969. Early Tertiary mammal locality in southern Libya. Proc. Geol. Soc. Lond., No. 1657:167171. 1971. Review of the fossil mammals of Libya. In Symposium on the Geology of Libya, University of Libya, pp. 215-225. Savage, R. J. G., and White, Μ. E. 1965. Two mammal faunas from the early Tertiary of central Libya. Proc. Geol. Soc. Lond., No. 1623:89-91. Schweinfurth, G. 1886. Reise in das Depressionsgebiet im Umkreise des Fayüm. Ζ. Ges. Erdk., Berlin 21:96-149. Selley, R. C. 1968. Near-shore marine and continental sediments of the Sirte basin, Libya. Quart. J. Geol. Soc. Lond. 124:419-460. Shackleton, R. M. 1951. The Kavirondo Rift Valley. Quart. J. Geol. Soc. Lond. 106:345-392. Shuey, R. T.; Brown, F. H.; and Croes, Μ. K. 1974. Magnetostratigraphy of the Shungura Formation, southwestern Ethiopia: fine structure of the lower Matuyama polarity epoch. Earth Planet. Sei. Letters 23:249-260. Simons, E. L. 1968. Early Cenozoic mammalian faunas, Fayüm Province, Egypt. Part I—African Oligocene Mammals: Introduction, History of study, and faunal succession. Peabody Mus. of Nat. Hist. Bull. (Yale University) 28:1-21. Simpson, G. G. 1928. Mesozoic Mammalia XI. Brancatherulum tendagurense Dietrich. Am. J. Sei., Ser. 5 (15): 303-308. Singer, R. 1954. The Saldanha skull from Hopefield, South Africa. Am. J. Phys. Anthrop. 12:345-362. Singer, R., and Wymer, J. 1968. Archaeological investigations at the Saldanha skull site in South Africa. S. Afr. Archaeol. Bull. 23:63-74. Smith, A. G.; Briden, J. C.; and Drewry, C. E. 1973. Phanerozoic world maps. In Organisms and continents through time. Palaeontological Soc., London, Special Papers in Palaeont., no. 12, pp. 1-42. Smith, A. G., and Hallam, A. 1970. The fit of the southern continents. Nature 225:139-144. Söhnge, P. G.; Visser, D. J. L.; and van Riet Lowe, C. 1937. The geology and archaeology of the Vaal River Basin. Union S. Afr. Geol. Suru. Mem. 39 (1 and 2): 1-184. Stromer, Ε. 1913. Mitteilungen über die Wirbeltierreste aus dem Mittelpliozän des Natrontales (Aegypten). Zeits. Deutsch. Geol. Ges. 65:350-372. 1914. Mitteilungen über die Wirbeltierreste aus dem Mittelpliozän des Natrontales (Aegypten). 3. Artiodactyla:Bunodontia:Flusspferd. Zeits. Deutsch. Geol. Ges. 66:1-33. 1920. Mitteilungen über die Wirbeltierreste aus dem Mittelpliozän des Natrontales (Aegypten). 5. Nachtrag zu 1. Affen. 6. Nachtrag zu 2. Raubtiere. S. B. Bayer. Akad. Wiss. 1920:345-370. 1926. Reste land- und süsswasserbewohnender Wirbeltiere aus den Diamantenfeldern DeutschSüdwestafrikas. In E. Kaiser. Die Diamantenwüste Südwestafrikas. Berlin: D. Reimer, pp. 107-153. 1932. Reste süsswasser- und landbewohnender

Africa: The Physical Setting

45

Wirbeltiere aus den Diamantfeldern Klein-Namaqualandes (Südwestafrika). S. B. Bayer. Akad. Wiss. 1931:17-47. Taieb, M.; Coppens, Y.; Johanson, D. C.; and Kalb, J. 1972. Depots sedimentaire et faunes du Plio-Pleistocene de la basse vallee de l'Awash (Afar Central, Ethiopie). C. R. Hebd. Seanc. Acad. Sei., Ser. D. 1275:819-822. Taieb, M.; Johanson, D. C.; Coppens, Y.; and Aronson, J. L. 1976. Geological and Palaeontological background of Hadar hominid site, Ethiopia. Nature 260:289-293. Tankard, A. J. 1975. Varswater Formation of the Langebaanweg-Saldanha area, Cape Province. Trans. Geol. Soc. S. Afr. 77:265-283. Tobias, P. V. 1967. see Martyn, J. M. 1967. Tobien, H. 1936. Mitteilungen über Wirbeltierreste aus dem Mittelpliozän des Natrontales (Aegypten). 7. Artiodactyla: Bunodontia: Suidae. Zeits. Deutsch. Geol. Ges., Berlin 88:42-53. Van Couvering, J. A. 1972. Radiometrie calibration of the European Neogene. In W. W. Bishop, and J. A. Miller, eds. Calibration of hominoid evolution. Edinburgh: Scottish Academic Press, pp. 247-271. Van Couvering, J. Α., and Miller, J. A. 1969. Miocene stratigraphy and age determinations, Rusinga Island, Kenya. Nature 221:628-632. Vrba, E. 1974. Chronological and ecological implications of the fossil Bovidae of the Sterkfontein Australopithecine site. Nature 250:19-23. Wells, L. H. 1962. Pleistocene faunas and the distribution of mammals in southern Africa. Ann. Cape Prov. Mus. Nat. Hist. 2:37-40. 1964. The Vaal River "Younger Gravels" faunal assemblage: a revised list. S. Afr. J. Sei. 60:91-93. 1969. Faunal subdivision of the Quaternary in southern Africa. S. Afr. Archaeol. Bull. 24:93-96. Wells, L. H.; Cooke, Η. B. S.; and Malan, B. D. 1941. The associated fauna and culture of the Vlakkraal thermal springs, O.F.S. Trans. Roy. Soc. S. Afr. 29:203-233. Whitworth, T. 1953. A contribution to the geology of Rusinga Island, Kenya. Quart. J. Geol. Soc. Lond. 109:7596. 1961. The geology of Mfwanganu Island, western Kenya. Overseas Geol. Mineral Resources (Gt. Brit.) 8 (2):150-190. Wood, A. E. 1968. Early Cenozoic mammalian faunas, Fayüm Province, Egypt. Part II—The African Oligocene Rodentia. Peabody Mus. Nat. Hist. Bull. (Yale University) 28:23-102.

Addendum The earliest Tertiary land fauna so far known in Africa has recently been found in Morocco. It is clearly of Paleocene age and has yielded a small collection of microvertebrates, including two different Palaeoryctids, possible Miacid and Provivverine carnivores and other creodont/ carnivore remains. Capetta, H.; Jaeger, J.-J.; Sabatier, M.; Sige, B.; Sudre, J.; and Vianey-Liaud, M. 1978. Decouverte dans le Paleocene du Maroc des plus anciens Mammiferes eutherien d'Afrique. Geobios 11 (2): 257-263.

46

3 Mesozoic Mammals A. W. Crompton and F. A. Jenkins, Jr.

Only three mammals have been positively identified from Mesozoic rocks in Africa. Two of these, Megazostrodon rudnerae (Crompton and Jenkins 1968) and Erythrotherium parringtoni (Crompton 1964), are from the late Triassic, and the remaining specimen, Brancatherulum tendagurense (Dietrich 1927), is from the late Jurassic. The Triassic specimens are important because they are among the oldest known mammals and because they are nearly complete skeletons. They provide critical information on the origin and development of several characteristic features of mammals such as a threeboned middle ear, diphyodonty, precise dental occlusion, regional differentiation in the vertebral column (notably in the anterior cervical and middorsal regions), erect or semierect limb posture, and specialization of the limb joints. The Jurassic specimen consists of a poorly preserved, edentulous jaw. Its value is limited to indicating that mammals were present on the African continent during late Jurassic times.

Megazostrodon and Erythrotherium Classification These two genera are generally considered as nontherian mammals (figure 3.1), a group t h a t has recently been reclassified by Hopson (1970). For convenience, part of the classification is repeated here. Class: Mammalia Subclass: Prototheria Infraclass: Eotheria Order: Triconodonta Family: Morganucodontidae Genera included in this family are Megazostrodon, Erythrotherium, Eozostrodon (= Morganucodon) Order: Docodonta Infraclass: Ornithodelphia Order: Monotremata Infraclass: Allotheria Order: Multituberculata

Kermack, Mussett, and Rigney (1973) also have recently classified nontherian mammals, preferring the term Atheria to Prototheria. We are not in agreement with all aspects of their classification, i.e., the definition of the subclass Atheria, the referral of Megazostrodon to a family Sinoconodontidae and the inclusion of this family in a new suborder Morganucodonta, and the creation of a new suborder Eutriconodonta for both triconodontids and amphilestids. For this reason, we will use Hopson's classification until some of these issues can be discussed in greater detail.

Crompton and Jenkins

Mesozoic Mammals

TRIASSIC

47

CRETACEOUS

MULTITUBERCULATA MONOTREMATA TRICONODONTA /

j-l Er/throtherium* - „ • Megazostrodon*

2 Proboino

-

C = > DOCODONTA

AMPHILESTIDAE

Kuehneotherium

AMPHITHERIIDAE

SYMMETRODOIMTA

=>

EUTHERIA 8 METATHERIA PAURQDONTIDAE 1 x—Brancotheru/um* DRYOL'ESTIDAE

Figure 3.1 Interrelationships of Mesozoic mammals.

Locality Data Erythrotherium parringtoni was discovered in 1962 in the upper Red Beds of the Stormberg Series (about 300 ft below the bottom of the Cave Sandstone) near the village of Tsekong, 4 mi southeast of Mafeteng in Lesotho (Crompton 1964). Megazostrodon rudnerae was discovered in 1966 in the middle of the Red Bed sequence (400 ft below the Cave Sandstone), 3 km southeast of the trading store at Fort Hartley in the Quthing District of southern Lesotho (Crompton and Jenkins 1968). Although Megazostrodon appears to have come from a lower horizon than Erythrotherium, the relative age of the specimens is indeterminate because the Red Beds tend to increase in thickness in a southerly direction and thus a lower horizon in the south may correspond to a higher one in the north. The Red Beds and overlying Cave Sandstone are

generally considered to be Norian and/or Rhaetic (Cox 1973). The division between Cave Sandstone and Red Beds does not correspond to the division between the Norian and Rhaetic, and the age of the transition between these beds appears to differ from locality to locality. If the middle Red Beds of southern Africa are identified as being of Norian age, then the two southern African Triassic mammals are older than the Rhaetic mammals of Europe (Hillaby 1967). However, because the European, Chinese, and southern African Triassic mammals are so similar and because the relative ages of the localities on the three continents cannot be accurately determined at present, it is perhaps safer simply to refer to the southern African mammals as being of late Triassic age. Of the skulls of the Erythrotherium and Megazos-

48

Crompton and Jenkins

trodon only the dentitions have been described (Crompton 1964, 1974; Crompton and Jenkins 1968; Hopson and Crompton 1969). Jenkins and Parrington (1976) surveyed the postcranial skeletons of Megazostrodon and Erythrotherium, together with the dissociated remains of the closely related Eozostrodon (= Morganucodon) from the Rhaetic fissure fillings of Britain. On the basis of these materials, which they concluded were structurally similar, they presented a skeletal reconstruction of a Triassic morganucodontid (figure 3.2).

Dentition The skull of Erythrotherium is partially disarticulated and represents a juvenile. The dental formula (figure 3.3) appears to have been I I* C j- PM j Μ

3

Unerupted and partially

formed second and fourth incisors are present and the last lower deciduous molar is still in place. The molars and premolars are of the typical morganucodontid pattern. The lower molar consists^ of four cusps in a row (from front to back, b a c d) and a well-developed cingulum supporting several cusps, including an enlarged cusp g (Kühneocone). The upper molars also have four cusps in a row (from

Mesozoic Mammals

front to back,B^ A C^ D), and external and internal cingula with cuspules. The lower jaw lacks a pronounced angle and is more slender than that of Eozostrodon. A well-developed dentary condyle is present. A groove on the medial side of the dentary, containing the remnants of the postdentary bones, is evidence that both the reptilian and mammalian jaw articulations functioned alongside one another (as in other morganucodontids; Kermack, Mussett, and Rigney 1973). Consequently, these forms lacked a typically mammalian three-boned middle ear (Crompton and Parker 1978) which was isolated from the lower jaw. In Megazostrodon (figure 3.4), incisors and canines are missing but the premolars and molars are exceptionally well preserved. The number and structure of the teeth, as well as the wear patterns on the molars, indicate that the animal was a young adult. The dental formula was I ? C ? PM f Μ The 5 4 molars are, on the whole, similar to those of Erythrotherium and Eozostrodon except that the cingula and cingular cusps, especially those on the outer surface of the uppers and the inner surface of the lowers, are far larger than in the other morganucodontids. In the upper molars the three principal cusps, B, A, and C, form a wide-angled triangle

Figure 3.2 Reconstruction of Triassic triconodont, based upon Megazostrodon, Eozostrodon, and Erythrotherium.

Crompton and Jenkins

Mesozoic Mammals

49

Figure 3.3 Internal view of the lower dentition and external view of the upper dentition of Erythrotherium parringtoni. rather than being arranged in a straight line. The molar occlusal relationships of Erythrotherium and Megazostrodon are slightly different from that of Eozostrodon. In the African morganucodontids the main cusp of the lower (ä) occluded against and slightly in front of the internal surface of cusp Β of the upper molar; the main upper molar cusp (A) occluded against the external surface of cusp d of the lower. Moreover, the last upper premolar is relatively small. In the Chinese and European morganucodontids the last upper premolar is always larger than the first molar, and cusp (ä) of the lower molar occludes between cusp (B) and cusp (A) of the opposing uppers. Thus, the size of the last upper premolar and the pattern of molar occlusion differentiate the African from the Eurasian forms. On the basis of skull size and tooth structure the morganucodontids appear to have been small insectivores with a premolar and molar dentition designed for puncturing and shearing. Diphyodonty and fixed relationships between teeth appear to have arisen in the advanced mammallike reptiles immediately ancestral to these late Triassic mammals. Hopson (1973) has argued that a diphyodont dentition may have been related to the origin of mammary glands and suckling behavior. This mode of feeding permits most skull growth to take place before the first set of milk teeth erupt. The limited amount of growth that takes place after this stage is only sufficient to accommodate a single replacement

of milk teeth by permanent incisors, canines, and premolars as well as the addition of permanent molars.

Brain Size Jerison (1973) has published a formula that expresses the relationship between brain and body size in vertebrates, viz., Ε = KP2/3 (where Ε and Ρ are brain and body weight in centimeter-gram units [grams or milliliters] and K, a proportionality constant). An average value for Κ for living mammals is 0.12 and for lower vertebrates, 0.007. There are very few reliable data on brain/body ratios in mammallike reptiles. Part of the problem is the difficulty of determining how much of the braincase was occupied by the brain itself. One of the most mammallike of the cynodonts, Probainognathus, has a maximum brain volume of about 1.5 ml (=1.5 g). The total body length of this form must have been about 30 to 35 cm, with a body weight between 1,000 and 1,500 g. If Probainognathus were a typical reptile, the expected brain size would have been between 0.7 and 0.9 ml, whereas if it were an average mammal it would have been between 12 and 16 ml. The brain size of Probainognathus falls well within the range of "average" reptiles. Jerison (1973) has estimated that the brain volume of a late Jurassic triconodont was about onehalf that of an "average" mammal. Recently dis-

50

Crompton and Jenkins

Mesozoic Mammals

PM,"

I

1 mm

1

Figure 3.4 Dentition of Megazostrodon rudnerae. (A) External view of upper dentition, (B and C) internal and external views of lower dentition, (D) crown view lower PM5 through M3 and upper PM5 through M2.

covered early Cretaceous triconodonts appear to have had slightly larger brains than the Jurassic forms. As the brain size of an advanced mammallike reptile is less than one-eighth of that of an average mammal, there is a marked contrast in brain size between this form and the Jurassic and Cretaceous triconodonts. Increase in brain size in early mammals is accompanied by a reorganization of the structure of the braincase. In advanced mammallike reptiles the cranial cavity anterior to the pituitary fossa nar-

rows rapidly in vertical dimension. The olfactory bulbs and tracts were contained in this narrow, cylindrical space formed between the frontals above and the orbitosphenoids below and laterally. A deep, vertical space, presumably filled by a cartilaginous interorbital septum, lies between the dorsally situated orbitosphenoids and the horizontal palatal bones. In mammals the expansion of the forebrain is accompanied by a ventral migration of the orbitosphenoid so that it becomes part of the basicranial axis in line with the basioccipital and basisphenoid;

Crompton and Jenkins

concomitantly, the frontals are drawn down to form a major part of the lateral wall of the braincase. This has already taken place in Cretaceous triconodonts. The individual bones comprising the braincase in the Triassic triconodont Megazostrodon are dislocated, and it is therefore not possible to measure the intracranial volume accurately. However, the individual elements making up the braincase of morganucodontids appear to be almost identical to those of later triconodonts, an indication that the brain/ body ratio was probably the same as, or close to, that of Jurassic triconodonts. In the transition from advanced mammallike reptiles to early mammals, a sudden and marked enlargement of the brain took place.

Postcranial Skeleton Megazostrodon appears to have been 10 cm in head/body length (or 14 cm from snout to tip of tail) and weighed between 20 and 30 g (Jenkins and Parrington 1976). Eozostrodon and Erythrotherium are about 20% smaller. The postcranial skeleton of early mammals, insofar as known, was structurally advanced over those of advanced cynodonts, i.e., those mammallike reptiles from which the first mammals appear to have arisen. Nevertheless, several primitive features are retained in the morganucodontid skeleton, such as a cynodontlike pectoral girdle that retains both coracoids and lacks a supraspinous fossa, the presence of a condyle on the humerus (rather than a trochlea) for articulation with the ulna, and the presence of an acetabulum composed of three more or less separate bony facets (iliac, ischial, and pubic). Jenkins (1974) reviewed the longstanding question of terrestriality versus arboreality in the origin and evolution of mammals. Although various authors have argued that the earliest mammals were adapted for one or the other habitat, Jenkins pointed out that for many small mammals, at least, the locomotor repertoire required for both habitats is much the same and that from the perspective of a small mammal one set of substrates grades into the other. The primitive musculoskeletal adaptations of mammals are best interpreted in terms of providing a flexible range of postures and movements that are employed in active foraging over spatially complex substrates. Such a locomotor repertoire is required by both arboreal and terrestrial substrates. The earliest mammals, which were very small, might well be expected to have been behaviorally and structurally adapted for the kind of niche occupied today by some tree shrews or tenrecs. Jenkins and Parrington (1976) interpreted vari-

Mesozoic Mammals

51

ous skeletal features of morganucodontids either as adaptations for locomotion on spatially complex and even steep surfaces, such as might be encountered by mammals active in both terrestrial and arboreal environments, or as structural innovations that appear for the first time in mammalian ancestry (for which the adaptive significance is unclear). Among the former are the following features. (1) Small size; the incremental increase in energy consumption for running up steep inclines, as compared with that required for running on a level surface, is far less (relative to body size) in small than large animals (Taylor, Caldwell, and Rowntree 1972). (2) Limb posture; the femur and humerus appear to have been abducted from the sagittal plane, and the elbow and knee flexed, so that the center of gravity was held low with a relatively wide stance. (3) Well-developed ball and socket joints at the shoulder and hip, making possible limb movements over a wide range of excursions. (4) A pelvis of a generalized mammalian pattern (although it should be remembered that at least two lineages of mammallike reptiles, the tritylodontids and ictidosaurs, evolved basically the same structure). (5) A hallux with joint adaptations similar to those in Tupaia (Jenkins 1974), and therefore probably capable of independent, "opposable" movements to the same degree. (6) Claws laterally compressed and sharp, with prominent flexor tubercles, an adaptation common to climbing mammals (see Cartmill 1974). (7) Atlanto-occipital and atlanto-axial joints with all the major adaptations (with the exception of synostosis of the atlantal arches and intercentrum) for movement and stability characteristic of later mammals. (8) Enlargement of the midcervical vertebral canal. This is evidence of an increased size of the cervical intumescence of the spinal cord, correlated with the greater neuromuscular control of the forelimb characteristic of mammals. (9) Distinctive structural modifications in the mid-dorsal vertebrae, notably in zygapophyseal orientation ("diaphragmatic vertebra") and spinal process orientation ("anticlinal vertebra"), evidence of regional specialization of the epaxial musculature by means of which mammals typically localize flexion-extension movements in the mid-dorsal region. Extension and flexion are important elements in locomotion in small mammals (Jenkins 1974).

Biology Following the general consensus that early mammals were probably capable of maintaining a constant body temperature, Jerison (1973) suggested

52

Crompton and Jenkins

that homeothermy enabled the first mammals to invade a nocturnal niche that in late Triassic times did not include reptiles. Crompton, Taylor, and Jagger (1978) surmised that early mammals were probably similar to modern-day hedgehogs and Madagascan tenrecs in maintaining a low body temperature (25 to 30°C) but possessing a reptilian metabolic rate three to five times lower than most living mammals. Not only is a reptilian metabolic rate economical and sufficient to sustain a constant low body temperature, but it is also suitable for a nocturnal existence. High ambient temperatures and solar radiation, coupled with diurnal activity and a low preferred body temperature, would have required inordinate amounts of evaporative heat loss to maintain a low body temperature. For this reason hedgehogs and tenrecs secrete themselves in burrows during the day. The high body temperature of typical mammals probably developed as an adaptation to a diurnal existence. Jerison (1973) pointed out that the invasion of a nocturnal niche by previously diurnal reptiles would have made extraordinary demands not only on their temperature regulation capacities but also on their sensory systems. Diurnal reptiles rely primarily on vision whereas most nocturnal mammals depend on developed auditory and olfactory senses. The dramatic enlargement of the olfactory bulb and forebrain in early mammals supports Jerison's view that these forms processed considerably more olfactory information than their mammallike reptile antecedents. Jerison (1973) suggested that the expansion of brains in early mammals was also, in part, related to greater neuronal complexity for auditory discrimination. Comparing mammals to reptiles, he pointed out, there is at least a tenfold increase in the neurons directly related to the auditory system at the thalamic level and up to a hundredfold increase in neurons at the cortical level. No complete endocranial casts of Mesozoic mammals are currently available, and therefore it is not possible to document a marked increase in the brain areas processing auditory information in the earliest mammals as it is in the case of the enlarged olfactory bulbs. However, there is evidence of a marked and sudden enlargement of the cochlear region of the inner ear in early mammals. The ability of mammals to discriminate between frequencies covering a wide range depends in part on the length of the basilar membrane and associated organ of Corti. In mammals the cochlea is housed in the petrous portion of the petrosal bone. The ventral surface of the cochlea housing, or the promontory, is a marked feature on the ventral side of the skull in front of the fenestra ovalis in all early mammals.

Mesozoic Mammals

This is in marked contrast to advanced mammallike reptiles where the position and size of the cochlea is not represented on the external surface of the bones surrounding the inner ear. Triassic mammals are much smaller than their known cynodont antecedents and it can perhaps be argued that because the cochlea is a sense organ it may not have undergone the same degree of reduction as the rest of the skull during the cynodont-mammal transition. However, a Jurassic survivor of the mammallike reptiles, Oligokyphus (which weighed about 1,000 g), is considerably smaller than its late Triassic antecedents (which weighed up to an estimated 40 kg) and shows no evidence of a relative increase in the size of the cochlear housing. An important feature of the Triassic mammals is the acquisition of a new mammalian jaw joint between the squamosal and dentary. This joint exists alongside the old reptilian joint between the quadrate (= incus) and articular (= malleus). The dentary condyle is far larger than the old reptilian jaw joint and these bones could therefore be used solely to conduct vibrations to the inner ear that were received by the tympanic membrane. In the earliest mammals the enlarged cochlea and the auditory ossicles (freed from a jaw support function) are evidence of improved auditory acuity. This interpretation supports Jerison's (1973) hypothesis that an improvement in the sense of hearing would have been expected if the early mammals had invaded a nocturnal niche. The features of the postcranial skeleton are not inconsistent with the suggestion that early mammals were nocturnal insectivores. The diversity within late Triassic morganucodontid mammals is limited, but they had a rather wide distribution—China, southern Africa, and Europe. They occur in great numbers in the British fissure fillings. Rapid dispersal into what was probably a worldwide distribution and limited diversity accords with the view that early mammals invaded a previously vacant nocturnal niche.

Relationships It is generally accepted that the morganucodontids (figure 1.1) are the stock that gave rise to the Triconodonta of the late Jurassic and Cretaceous and to the Docodonta of the Jurassic (Hopson and Crompton 1969; Mills 1971; Parrington 1971; Crompton and Jenkins 1973). Little is known of the origin and relationship of the amphilestids from the late Jurassic; traditionally classified as a subfamily of triconodontids, they were given separate family status by Kühne (1958). It is even questionable whether they are triconodonts, for the occlusion of

Crompton and Jenkins

their postcanine teeth is different from that of most morganucodontids, typical triconodonts, and the therian mammals. The amphilestids probably arose from the same stock that gave rise to the morganucodontids and the enigmatic late Triassic haramiyids, the latter known only from isolated teeth. The haramiyids are possibly related to the largest group of prototherian mammals, the multituberculates (Hahn 1973). Kielan-Jaworowska (1970) has suggested that a relationship exists between monotremes and multituberculates. A third group of Triassic mammals, the Kuehneotheriidae, appear to have given rise to the therian mammals (Kermack, Kermack, and Mussett 1968) but are far less common in the British deposits and have not as yet been found outside Europe. The interrelationship of the late Triassic mammals has not yet been clearly established. The morganucodontids and kuehneotheriids share a suite of osteological features: a diphyodont dentition with well-differentiated premolars and molars, a fixed relationship between upper and lower molars, a transverse component to jaw movement during dental occlusion, and a dentary-squamosal articulation. These features, which clearly differentiate these two groups of early mammals from all the mammallike reptiles, may also have been present in other Triassic mammals. Some workers (Hopson and Crompton 1969; Parrington 1971; Crompton and Jenkins 1973) have claimed that these features indicate close relationship between morganucodontids and kuehneotheriids. One or several of these osteological features may have evolved independently in other mammallike reptile lineages, e.g., the dentary-squamosal articulation in the ictidosaurs and tritylodontids (Crompton 1972), but the combination of these features probably arose only in the group that gave rise to both groups of Triassic mammals. When the kuehneotheriids and haramiyids are better known, it may be possible to increase this list to include features now only known in the morganucodontids, such as expansion of the forebrain and an enlarged cochlea. Other workers (Kermack 1967; Mills 1971), however, have suggested that the differences in molar structure and occlusion not only between the morganucodontids and kuehneotheriids but also between the northern (Europe and China) and southern (Lesotho) morganucodontids are so great that they probably arose from different groups of mammallike reptiles. If this is correct, it implies that at least the two main groups of Triassic mammals have independent lineages extending well back into the Triassic and possibly even into the upper Permian. In this case many of the features considered diagnos-

Mesozoic Mammals

53

tic of mammals must have arisen independently at least twice, and even more if the haramiyids and amphilestids are not closely related to either the morganucodontids or kuehneotheriids. Only further study of available material and the discovery of new material will resolve this problem.

Associated Fauna The upper Red Bed vertebrate fauna is dominated by small and large prosauropods (Charig, Attridge, and Crompton 1965). Less numerous elements are the advanced mammallike reptiles, tritylodonts (Ginsberg 1962), ictidosaurs (Crompton 1958), primitive crocodiles (Nash 1975), and a few thecodonts (Walker 1972). This fauna is typical of late Triassic terrestrial deposits of other continents. Morganucodontid mammals (Kermack, Mussett, and Rigney 1973), tritylodonts (Young 1947), prosauropod dinosaurs (Young 1951), and primitive ornithischian dinosaurs (Simmons 1965) are known from the Lufeng Beds of China. The Los Colorados Formation of Argentina (Bonaparte 1971) contains a similar array of tritylodontids, ictidosaurs, prosauropods, saurischian dinosaurs, primitive crocodiles, and thecodonts. Tritylodonts, saurischian and ornithischian dinosaurs, primitive crocodiles, and thecodonts are also known from the North American late Triassic deposits (Romer 1971). Prosauropod and other saurischian dinosaurs, morganucodontids, tritylodontids, and thecodonts are known in the late Triassic beds of Europe (Robinson 1957, 1971). In many cases the representatives of these various groups on different continents are almost identical. For example, among the mammals, Erythrotherium from southern Africa, Eozostrodon (= Morganucodon) watsoni from Wales, and Eozostrodon (= Morganucodon) oehleri from southern China are so similar that Kermack, Mussett, and Rigney (1973) have placed them all in the same genus. The bipedal prosauropods, such as Thecodontosaurus from Europe, Massospondylus from southern Africa, and Yaleosaurus from North America, are as difficult to tell apart as are the melanorosaurid prosauropods such as Melanorosaurus from southern Africa, Riojasaurus from South America, and Sinosaurus from China. The tritylodontids (Tritylodon from southern Africa, Bienotherium from China, and the undescribed tritylodontid from the Kayenta Formation of North America) are also very similar. The distribution of the late Triassic faunas and the close similarity between individual genera on what are now widely separated continents suggest that barriers that now exist between the continents were not present in late Triassic times.

54

Crompton and Jenkins

Mesozoic Mammals

Brancatherulum tendagurense

References

Classification

Bonaparte, J. F., 1971. Los tetrapodos del sector superior de la Formacion Los Colorados, La Rioja, Argentina. (Triäsico Superior) Op. Lill. 22:1-183. Branca, W., 1916. Ein Säugetier?-Unterkiefer aus den Tendaguru-Schichten Wissens. Ergeb. TendaguruExpedition 1909-12. Arch. Biontol. 4:137-40. Cartmill, M., 1974. Pads and claws in arboreal locomotion. In Jenkins, F. Α., ed., Primate locomotion, New York: Academic Press, pp. 45-83. Charig, A. J.; Attridge, J.; and Crompton, A. W., 1965. The origin of the sauropods and the classification of the Saurischia. Proc. Linn. Soc. Lond. 176:197-221. Clemens, W. Α., and Mills, J. R. E., 1971. Review of Peramus tenuirostris Owen (Eupantotheria, Mammalia). Bull. Brit. Mus. Nat. Hist. (Geol.) 20:89-113. Colbert, Ε. H., 1961. Dinosaurs, their discovery and their world. New York: E. P. Dutton. Cox, C. B., 1967. Changes in terrestrial vertebrate faunas during the Mesozoic. In W. B. Harland, et al., eds., The Fossil Record. London: Geological Society, pp. 7789. 1973. Gondwanaland Triassic stratigraphy. An. Acad. Brasil Cienc. 45:115-119. Crompton, A. W., 1958. The cranial morphology of a new genus and species of ictidosaurian. Proc. Zool. Soc. Lond. 230:183-216. 1964. A preliminary description of a new mammal from the upper Triassic of South Africa. Proc. Zool. Soc. Lond. 242:441-452. 1968. In search of the insignificant. Discovery 3: 23-32. 1972. The evolution of the jaw articulation of cynodonts. In K. A. Joysey, and T. Kemp, eds., Studies in vertebrate evolution. Edinburgh: Oliver and Boyd pp. 231-251. 1974. The dentitions and relationships of the southern African Triassic mammals Erythrotherium parringtoni and Megazostrodon rudnerae. Bull. Brit. Mus. Nat. Hist. (Geol.) 24:397-437. Crompton, A. W., and Jenkins, Jr., F. Α., 1968. Molar occlusion in late Triassic mammals. Biol. Rev. 43:427458. 1973. Mammals from reptiles; a review of mammalian origins. Ann. Rev. of Earth and Planetary Sei. 1:131-155. Crompton, A. W., and Parker, P. 1978. Evolution of the mammalian masticatory system. American Scientist 66: 192-201. Crompton, A. W.; Taylor, C. R.; and Jagger, J. Α., 1978. Evolution of homeothermy in mammals. Nature 272: 333-336. Dietrich, W. O., 1927.Brancatherulum n.g., ein proplacentalier aus dem obersten Jura des Tendaguru in DeutschOstafrika: Centralbl. Min. Geol. Pal. 2927(B):423-426. Ginsburg, L., 1962. Likhoelia ellenbergeri, Tritylodonte du Trias Superieur du Basutoland (Afrique du Sud). Ann. Paleont. 48:179-194. Hahn, G., 1973. Neue Zähne von Haramiyiden aus der

The classification of Brancatherulum tendagurense is difficult because the fossil consists only of a poorly preserved lower jaw. Subclass: Theria Order: Pantotheria Family: Paurodontidae

The specimen is a damaged and edentulous right jaw and was first described by Branca in 1916. It was referred to the Pantotheria by Hennig (1919), fully described and named by Dietrich (1927), and further reviewed and discussed by Simpson (1928). The only feature of diagnostic importance is a pantotherelike angle; the number of postcanine teeth (between six and eight) is relatively small. For these reasons it is usually referred to the Paurodontidae. (See Simpson 1928 for an account of the similarities between Brancatherulum and Peramus from the British Purbeckian deposits.) However, Kühne (1968) noted similarities in the back of the jaws of Amphitherium, Brancatherulum, Peramus, Archaeotrigon, and a specimen from the Kimmeridgian of Portugal that he refers to the genus Peramus; but he did not go so far as to suggest that these forms (which are usually placed in two separate families) should be placed in a single taxonomic unit. Clemens and Mills (1971) have recently published a review of the structure of the teeth and jaws of Peramus and they conclude that the phylogenetic relationships of Brancatherulum remain obscure.

Age and Associated Fauna The Tendaguru deposits have also yielded a diverse and abundant dinosaurian fauna. These have been dated as being of late Kimmeridgian or early Tithonian age. The dinosaurs are similar to those of the Morrison Formation of North America, particularly the smaller saurischians, the heavy sauropods, and the stegosaurs (Colbert 1961), indicating that late Jurassic vertebrate faunas still had a worldwide distribution. Our knowledge of Mesozoic mammals is based almost entirely upon remains from North America, Europe, and Asia and with the exception of the forms mentioned in this paper none are known from southern continents. It will not be possible to answer many fundamental questions about mammalian evolution, such as the origin and distribution of monotremes or the differentiation of marsupials and placentals from mammals of "therian grade," unless a concerted effort is made to search for Mesozoic mammals in Africa and other southern continents.

Crompton and Jenkins

Deutschen Ober-Trias und ihre beziehungen zu den Multituberculaten. Palaeontogr. 142:1-15. Hennig, E., 1919. Die entstehung des Säugerzahns und die Paläontologie. Naturwiss. Wochenshrift, N. F. 18 (No. 51):l-7. Hillaby, J., 1967. Clues to evolution. New Scientist 34:163. Hopson, J. Α., 1970. The classification of nontherian mammals. J. Mamm. 51:1-9. 1973. Endothermy, small size, and the origin of mammalian reproduction. American Naturalist 107: 446-452. Hopson, J. Α., and Crompton, A. W., 1969. Origin of mammals. In T. Dobzhansky; Μ. K. Hecht; and Steere, W. C., eds., Evolutionary biology 3:15-72. Jenkins, F. Α., Jr., 1974. Tree shrew locomotion and the origins of primate arborealism. In Jenkins, F. A. Jr., ed., Primate locomotion. New York: Academic Press, pp. 85-115. Jenkins, F. Α., Jr., and Parrington, F. R., 1976. The postcranial skeletons of the Triassic mammals—Eozostrodon, Megazostrodon and Erythrotherium. Phil. Trans. Roy. Soc. Lond. {B)273:387-431. Jerison, H. J., 1973. Evolution of the brain and intelligence. New York: Academic Press. Kermack, D. M.; Kermack, Κ. Α.; and Mussett, F., 1968. The Welsh pantothere Kuehneotherium praecursoris. J. Linn. Soc. Lond. 47:407-423. Kermack, Κ. Α., 1967. The interrelation of early mammals. J. Linn. Soc. Lond. 47:241-249. Kermack, Κ. Α.; Mussett, F.; and Rigney, H. W., 1973. The lower jaw of Morganucodon. J. Linn. Soc. Lond. 53:87175. Kielan-Jaworowska, Z., 1970. Unknown structures in multituberculate skull. Nature 226:974-976. Kühne, W. G., 1968. Kimmeridge mammals and their bearing on the phylogeny of the Mammalia. In Ε. T. Drake, ed., Evolution and environment. New Haven: Yale University Press, pp. 109-123.

Mesozoic Mammals

55

Mills, J. R. E., 1971. The dentition of Morganucodon. In D. M. Kermack, and K. A. Kermack, eds., Early Mammals. London: Academic Press, pp. 29-62. Nash, D. S., 1975. The morphology and relationships of a crocodilian, Orthosuchus strombergi, from the upper Triassic of Lesotho. Ann. S. Afr. Mus. 67:227-329. Patterson, B., and Olson, E. C., 1961. A Tricondontid mammal from the Triassic of Yunnan. Internat. Colloq. on the evolution of mammals. Kon. Vlaamse Acad. Wetensch. Lett. Sch. Künsten Belgie, part I: 129-191. Parrington, F. R., 1971. On the upper Triassic Mammals. Phil. Trans. Roy. Soc. Lond. Ser. Β 261:231-272. Robinson, P., 1957. The Mesozoic fissures of the Bristol Channel area and their vertebrate faunas. J. Linn. Soc. Lond. 43:260-279. 1971. A problem of faunal replacement on Permo-Triassic continents. Palaeontology 14:131-153. Romer, A. S., 1969. Cynodont reptile with incipient mammalian jaw articulation. Science 166: 881-882. 1971. Tetrapod vertebrates and Gondwanaland. Proc. Int. Union Geol. Sei., Gondwana Symposium. Pap. No. 2:111-124. Simmons, D. J., 1965. The non-therapsid reptiles of the Lufeng Basin, Yunnan, China. Fieldiana: Geology 15: 1-93. Simpson, G. G., 1928. Mesozoic Mammalia XI Brancatherulum tendagurense. Dietrich. Am. J. Sei. 15:303308. Taylor, C. R.; Caldwell, S. L.; and Rowntree, V. J., 1972. Running up and down hills: some consequences of size. Science 178:1096-1097. Walker, A. D., 1972. New light on the origin of birds and crocodiles. Nature 237:257-263. Young, C. C., 1947. Mammal-like reptiles from Lufeng, Yunnan, China. Proc. Zool. Soc. Lond. 117:537-97. 1951. The Lufeng saurischian fauna in China. Paleont. Sin. 134:1-96.

56

4 Insectivora and Chiroptera Percy M. Butler

Several groups constitute the nonrodent micromammalian fauna. Insectivora is here used in a broad sense to denote an artificial assemblage of early offshoots from the eutherian stem (Butler 1972). The fossil history of the insectivores and bats is much less well known than that of most groups of larger mammals and their phylogeny in Africa has not yet passed the conjectural stage. At present they form a significant element of the fauna (21 genera of insectivores and 43 genera of bats), and this has probably been the case throughout the Tertiary. However, only a few localities have so far yielded remains of small mammals; the known distribution of the families of insectivores and bats is summarized in table 4.1.

Order Insectivora Family Ptolemaiidae RECORDED SPECIES: Ptolemaia lyonsi Osborn 1908; Schlosser 1923; Van Valen 1966; Coombs 1971; Simons and Gingerich 1974. Early Oligocene, Fayum. Qarunavus meyeri Simons and Gingerich 1974; Schlosser 1910, 1911; Matthew 1918; Butler 1969. Early Oligocene, Fayum. Kelba quadreemae Savage 1975. Early Miocene, Rusinga.

The relationships of these very imperfectly known forms are still uncertain, even at the ordinal level. Ptolemaia lyonsi was based on a single mandible with worn teeth; a second specimen has recently been described by Simons and Gingerich (1974). A pair of mandibles with milk teeth and unworn M t and M2 were referred to the same species by Schlosser (1910, 1911), but Matthew (1918) doubted whether they belonged to the same genus. They have now been named Qaranavus meyeri by Simons and Gingerich (1974). Kelba quadreemae was described from two isolated upper molars by Savage (1965), who regarded it as an arctocyonid. A complete upper dentition is now available and will soon be described by R. J. G. Savage. Van Valen (1966) reviewed the various opinions on the relationships of Ptolemaia, which he placed near the Pantolestidae, a view first tentatively expressed by Schlosser (1923). In 1967 he included Kelba in the Ptolemaiidae without giving reasons, but presumably because of a general pantolestid similarity in the upper molar pattern (see also R. J. G. Savage, chapter 11, for further discussion); he placed the Ptolemaiidae next to the Pantolestidae in his superfamily Tupaioidea. Butler (1969) noted a

Butler

Table 4.1

Insectivora and

Chiroptera

57

Known distribution in time of the families of insectivores and bats on the African continent.

Family Ptolemaiidae Macroscelididae Erinaceidae Soricidae Tenrecidae Chrysochloridae Pteropodidae

Vampyravus

Rhinopomatidae Emballonuridae Megadermatidae Rhinolophidae Hipposideridae Nycteridae Myzopodidae Vespertilionidae Molossidae

Early Oligocene: Egypt

Africa

Mio-Pliocene: Morocco

East Africa

Africa

PlioceneEarly Pleistocene: S. Africa

—

—

—

—

—

—

—

X

—

—

X

—

—

X X X

Pliocene

Early Miocene East Africa

s.w.

s.w.

Early Pleistocene: Olduvai

Recent

X X X ? X X X

—

X X

X

—

X X X

—

—

—

—

—

—

—

—

—

—

X

—

—

—

—

—

—

—

X X X X X X

X

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

X

—

X X — — — -

—

X X

—

—

—

—

X

—

X

—

X X X

—

—

—

—

—

—

—

—

X

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

X

—

—

—

—

—

—

— —

X X

number of similarities in the teeth between the juvenile jaw of Qarunavus and the macroscelidid Rhynchocyon. Until more is known of the structure of these animals it seems hardly justifiable to classify them other than as Eutheria insertae sedis. Simons and Gingerich (1974) suggest a possible relationship of Ptolemaia to the Tubulidentata. Family Macroscelididae RECORDED SPECIES: Metoldobotes stromeri Schlosser 1910, 1911; Matthew 1910, 1915; Patterson 1965. Early Oligocene, Fayum. Rhynchocyon clarki Butler and Hopwood 1957; Butler 1969. Early Miocene, Rusinga, Songhor. Rhynchocyon rusingae Butler 1969. Early Miocene, Rusinga, Songhor. Macroscelididae indet. 1. Butler 1969. Early Miocene, Rusinga. Macroscelididae indet. 2. Butler 1969. Early Miocene, Rusinga. Myohyrax oswaldi Andrews 1914; Stromer 1926; Hopwood 1929; Whitworth 1954; Patterson 1965. Synonym: M. doederleini Stromer 1926. Early Miocene, Karungu, Koru, Rusinga, Namib. Protypotheroides beetzi Stromer 1922, 1926; Hopwood 1929; Whitworth 1954; Patterson 1965. Synonym: Myohyrax osborni Hopwood 1929. Early Miocene, Namib. Palaeothentoides africanus Stromer 1932; Butler and Hopwood 1957; Patterson 1965. Pliocene, Klein Zee. Elephantulus fuscus leakeyi Butler and Greenwood 1976, 1965. Synonym: Nasilio sp. Late Pliocene to early Pleistocene, Olduvai, Makapansgat and probably other

X X X

X X X X X X

—

X X

cave breccias in the Transvaal, where it has been recorded as Elephantulus (Nasilio) cf brachyrhynchus. Elephantulus broomi Corbet and Hanks 1968; Broom 1937, 1938; Butler and Greenwood 1976. Synonym: Elephantomys langi Broom 1937. Late Pliocene to early Pleistocene, S. Africa (Schurveberg, Makapansgat), Olduvai. Most of the specimens referred to Elephantulus langi by De Graaff (1960) appear to be E. antiquus. Elephantulus antiquus Broom 1948; Butler and Greenwood 1976. Late Pliocene to early Pleistocene, Bolt's Farm, Makapansgat, and other S. African cave breccias. cf Elephantulus antiquus Butler and Greenwood 1976. Early Pleistocene, Olduvai. Macroscelides proboscideus vagans Butler and Greenwood 1976. Late Pliocene, Makapansgat, Taung. Mylomygale spiersi Broom 1948; De Graaff 1960; Patterson 1965. Late Pliocene, Taung, Sterkfontein. This family, confined to Africa, occupies an isolated position among eutherian mammals, in recognition of which a separate order Macroscelidea has been proposed for it (Butler 1956b, Patterson 1965; ranked as a suborder of the Insectivora by Van Valen 1967). McKenna (1975) regards the Asiatic Paleocene-Oligocene family Anagalidae as macroscelidean, and believes that the Macroscelididae originated in Asia in or before the early Oligocene. He considers that the order Macroscelidea is related to the order Lagomorpha; Szalay (1977) goes further and includes the Macroscelidea in the Lagomorpha as a suborder. The fossil history of the Macroscelididae in Africa, though fragmentary, is better known than that of

58

Butler

other African "insectivore" families; it has been reviewed by Patterson (1965). The oldest known species, Metoldobotes stromeri from the Fayum, was originally interpreted as a mixodectid but was recognized as macroscelidid by Patterson (1965). It is known only from a single specimen, a mandibular ramus of which the anterior tip is missing. The dental formula, as interpreted by Patterson, is Ι3?01Ρ4Μ2. The anterior part of the dentition, from P3 forward, is abbreviated in comparison to the length of M ^ , a resemblance to the living Macroscelides that is more likely to be due to parallel evolution than to direct relationship. There is an enlarged, procumbent incisor (I3 or I2), which was apparently not the most anterior tooth. The canine is a small tooth with a single blunt cusp. The cheek teeth are bunodont and low-crowned as in Miocene Rhynchocyoninae, and M3 is absent as in most macroscelidids. Despite its early date, Metoldobotes is already too specialized to stand near the ancestry of any of the later members of the family. Patterson (1965) very tentatively referred it to the Macroscelidinae, a subfamily otherwise known only from the Pliocene and Pleistocene, but this carries little conviction. Metoldobotes seems best interpreted as an extinct offshoot from an Eocene macroscelidid stock, indicative of a radiation that had already taken place by early Oligocene time. The fragment of a femur, tentatively referred to Metoldobotes by Schlosser (1911), seems to have little resemblance to living Macroscelididae, and the identification must be regarded as very improbable. By the Miocene the Macroscelididae had diversified into at least two subfamilies, of which the Rhynchocyoninae still exist. Two Miocene species have been placed in the existing genus Rhynchocyon (Butler and Hopwood 1957, Butler 1969), but these show primitive characters that might justify at least a subgeneric separation. R. clarki is known from the anterior part of a skull, and a number of teeth and jaw fragments have been referred to this species mainly on the basis of size. The anterior dentition is unknown beyond P1 and P2. R. rusingae is known only from jaw fragments and an isolated M1. The skull of R. clarki shows the characteristic rhynchocyonine flattening of the face, with a broad unperforated palate and long infraorbital canal, but it is primitive in some respects. The crowns of the teeth are less elevated than in living species, and the anterior hypoconid crest (cristid obliqua) meets the trigonid midway between the protoconid and the metaconid, instead of joining the metaconid; in these respects there is a resemblance to Metoldobotes. R. rusingae is larger than R. clarki, resembling the

Insectivora and Chiroptera

living species in size; it also agrees with living forms in the presence of a protostylid on the posterior ridge of P2, P3, and dP3, and it may be near the direct ancestry of the two living species of Rhynchocyon. The Myohyracinae appear to have specialized for a herbivorous diet, and were formerly regarded as hyracoid (e.g., Whitworth 1954); they were recognized as macroscelidid by Patterson (1965). Myohyrax oswaldi is known from the almost complete skull and mandible, some tarsals, vertebrae, and fragments of long bones and is thus the most completely known fossil species of the family. Protypotheroides beetzi is represented only by teeth and jaws. The skull of Myohyrax resembles those of Macroscelidinae rather than Rhynchocyoninae in the narrow facial region. The mandible is much deeper below the cheek teeth and more expanded in the region of the angle than in other members of the family, implying a greater development of the masseter musculature. The first two incisors in each jaw are enlarged and procumbent (perhaps a resemblance to Metoldobotes), and their enamel is confined to the labial surface. The cheek teeth are prismatic as in Macroscelidinae, but they have crown cement which does not occur in that subfamily. Wear of the upper cheek teeth produces fossettes, which are subdivided as in some species of Εlephantulus iE- antiquus, E. edwardi). Ρ and P3 are fully molariform, P2 and P2 partly so. M2 is similar in size to M1, and M2 to Mj. Third molars are present as small teeth in both jaws. Thus the chewing area has been considerably expanded. Protypotheroides is larger and somewhat differently specialized. Its cheek teeth lack cement, but a fossette develops in the trigonids of P 4 -M 2 , due to a deepening of the valley between the paraconid and the metaconid. A second fossette in the talonid of these teeth separates the entoconid from the rather strongly developed hypoconulid. Two problematic specimens (Butler 1969) may indicate other lines of evolution in the Miocene. One of them, from Rusinga, consists of two associated mandibles of a juvenile animal with milk molars. DP3 has a metaconid, displaced lingually, and thus differs from Rhynchocyon which has a protostylid in a more buccal position on this tooth, but no metaconid. In Macroscelidinae the metaconid of dP3 is indistinct or absent, but there is a metastylid situated more posteriorly and connected to the rudimentary hypoconid by the equivalent of the cristid obliqua. The unerupted Mi of the Rusinga specimen can be seen in vertical cross-section; it is more high-crowned than in Rhynchocyon. Though tentatively referred to the Myohyracinae, this specimen may represent a hitherto unknown group.

Butler

The second problematic specimen, from Koru, consists of a fragment of mandible with unworn P2 and P3. P2 has a protostylid as in some specimens of Rhynchocyon, but it also possesses a minute metaconid in a posterolingual position. On P3 the protostylid is absent, but the metaconid is much better developed. The Macroscelidinae form the largest subfamily at present, but they are unknown in the fossil record before the Pliocene (unless Metoldobotes is included). The taxonomy of the living species has been revised by Corbet and Hanks (1968), who recognize three genera, Petrodromus (one species), Elephantulus (including Nasilio) (nine species), and Macroscelides (one species). A tenth species of Elephantulus, E. fuscus, has since been distinguished by Corbet (1973). The Macroscelidinae cannot be derived from either of the Miocene subfamilies, though they perhaps show rather more resemblance to the Myohyracinae than to the Rhynchocyoninae; presumably their ancestry lies in a third group, at present unknown. By the time that the subfamily appears in the fossil record, the main radiation must have taken place, for Macroscelides and Elephantulus were well differentiated at Makapansgat, and a third, now extinct genus, Palaeothentoides, is known from Klein Zee. The most recent study of Pliocene-Pleistocene Macroscelidinae is by Butler and Greenwood (1976), who reported on specimens from Olduvai Bed I and from Makapansgat. The commonest species from Olduvai possesses a third molar in the lower jaw and is close to the living Elephantulus (Nasilio) fuscus, a species with a limited distribution around the southern end of Lake Nyasa. It occurs also at Makapansgat, and possibly at other Transvaal sites from which "Nasilio" has been reported but not examined in detail. However, the main species at Makapansgat, and probably in other Transvaal cave breccias, is E. antiquus, a distinctive species with some resemblance to E. rufescens, from East Africa, and to E. edwardi, from South Africa. Less common, but found both at Olduvai and a number of Transvaal localities, is E. broomi (= E. langi) a form which seems to resemble quite closely the living E. intufi. Macroscelides is represented at Makapansgat by a form only subspecifically different from the living M. proboscideus; the genus occurs at Taung, but apparently not at Olduvai. Vegetation, and therefore rainfall, seems to be a major factor limiting the distribution of the species of Macroscelidinae; at the extremes are Petrodromus, an inhabitant of forests and probably in a number of respects the most primitive surviving form, and Macroscelides, which is confined to areas with rainfall below

Insectivora and Chiroptera

59

100 mm and specialized in many respects. The fossil history of Petrodromus is unknown. M3 is absent in most Macroscelidinae, but it occurs in the Pliocene Palaeothentoides and in three species of Elephantulus (E. brachyrhynchus, E. fuscus, E. fuscipes) that are placed by some authors (including Patterson 1965) in a separate genus, Nasilio. In most other characters the species of Nasilio are very close to Elephantulus intufi, E. rupestris, and E. broomi; indeed lower dentitions referred to E. broomi are indistinguishable from those of the early Pleistocene form of E. fuscus except for the absence of M3. Outside the Macroscelidinae M3 is present only in the Myohyracinae, where it is associated with M3, a tooth that occurs in Nasilio only as a very rare individual abnormality. Whether M3 is a primitive character in the family which has been lost in most lines, or whether it was lost early and then regained in a limited number of lines, is a question that cannot be answered on existing knowledge. Another character of taxonomic interest is the form of P2. In E. intufi, E. rupestris, E. broomi, and the species possessing M3, as well as in Macroscelides, P2 resembles P3 in being a comparatively broad tooth with four roots and well-developed lingual cusps; in the remaining species of Elephantulus, including the late Pliocene-early Pleistocene E. antiquus, P2 is narrower, with three or two roots, and it shows various degrees of reduction of the lingual cusps. Similar differences, though less marked, can be seen in P3. It is not unlikely that the broad type of P2 is the more primitive, and that the narrower type has come about by reduction; even within the species E. (Nasilio) brachyrhynchus there is much individual and regional variation in the size of the protocone of P2, leading in extreme cases to its almost complete absence. Broom (1937) proposed the generic name Elephantomys for E. langi (i.e., E. broomi) and E. intufi because of their "molariform" P2, but subsequently (1938) he realized that the type species of Elephantulus, E. rupestris, also had this character and therefore that Elephantomys was invalid. Although he did not create a name for the remaining species he continued to believe that the character has taxonomic value, and in this belief he was probably correct. Undescribed material from Irhoud-Ocre, Morocco (about 2 m.y.), shows the presence there of a species of Elephantulus, perhaps ancestral to E. rozeti which now inhabits North Africa (J.-J. Jaeger, pers. commun.). There remains Mylomygale, a problematic form from the late Pliocene of the Transvaal, known only by a lower jaw from Taung (Broom 1948) and an iso-

60

Butler

lated P4 from Sterkfontein (De Graaff 1960). It has high-crowned teeth like Macroscelides, but the dentition anterior to P3 is even more abbreviated than in that genus. It agrees with Protypotheroides and differs from Macroscelidinae in the presence of a reentrant fold of enamel between the entoconid and the enlarged hypoconulid. Myxomygale needs restudy. Patterson (1965) proposed for it a new subfamily Myxomygalinae, but it might eventually prove to be a late survivor of the Myohyracinae. From this very incomplete history the following tentative conclusions may be drawn. The Macroscelididae were a component of the Paleogene fauna of Africa, and there is an indication of a major radiation of the group that was already in progress in the Oligocene and continued into the Miocene. An unknown derivative of this radiation gave rise to the Macroscelidinae, which underwent a secondary radiation in the late Tertiary. Rhynchocyon is the only surviving relic of the earlier radiation. Family Erinaceidae RECORDED SPECIES: Galerix africanus Butler 1956a. Early Miocene, Songhor, Rusinga. Lanthanotherium sp., Butler 1969. Early Miocene, Songhor. Amphechinus rusingensis Butler 1956a, 1969. Early Miocene, Rusinga, Songhor. Gymnurechinus leakeyi Butler 1956a, 1969. Early Miocene, Rusinga. Gymnurechinus camptolophus Butler 1956a. Early Miocene, Rusinga. Gymnurechinus songhorensis Butler 1956a, 1969. Early Miocene, Songhor, ?Rusinga. Protechinus salis Lavocat 1961. Mio-Pliocene, Beni Mellal. Erinaceus broomi Butler and Greenwood 1973; Broom 1937, 1948. Synonym: Atelerix major Broom 1937. Early Pleistocene, Bolt's Farm, Olduvai.

Unlike the Macroscelididae, the Erinaceidae are widespread in the lower Tertiary of Europe, Asia, and North America, and they appear in Africa only as immigrants. Whereas in the Miocene of Europe the Echinosoricinae were as numerous and varied as the Erinaceinae, in Africa they are represented by only half a dozen fragments of mandible from Songhor and Rusinga. The specimens have been provisionally placed in the European genera Galerix and Lanthanotherium, but the material is so incomplete that identification can be only tentative; it is not even certain that more than one species is represented. The evi-

Insectivora and Chiroptera

dence for the existence of Galerix in the Mio-Pliocene of Morocco, consisting as it does of an isolated incisor tooth (Lavocat 1961), is very doubtful indeed. Erinaceinae, on the other hand, are common in the early Miocene of East Africa. A species oi Amphechinus (= Palaeoerinaceus) is represented by several specimens from Rusinga and Songhor, including the anterior part of a skull with associated lower jaw. Although comparatively late in date (the genus occurs in the Oligocene of Europe and China), A. rusingensis possesses some primitive characters unknown in the European forms—e.g., the relatively larger upper canines and the two-rooted P2. Most of the Miocene erinaceid material from East Africa has been placed in the genus Gymnurechinus, unknown outside Africa but the most common component of the insectivore fauna of Rusinga and Songhor. Several skulls have been found, and also part of a skeleton (Butler 1956a, 1969). Like Amphechinus, Gymnurechinus is at a level of evolution at which the dentition had already approached the modern erinaceine condition, but the skull and skeleton still retained many echinosoricine features. It has avoided a number of specializations that exclude Amphechinus from the direct ancestry of Erinaceus and thus cannot be derived from Amphechinus; it probably represents an offshoot from the unknown stock that gave rise to Erinaceus. Other derivatives of this stock make their appearance in Europe in the middle ("Erinaceus" sansaniensis) and late Miocene (Mioechinus oeningensis) (Butler 1948, 1956a), but the Oligocene ancestry is unknown. Thus the African Miocene erinaceids probably did not come from Europe. A more likely source is Asia south of the Tethys Sea, but unfortunately no fossil insectivores have been obtained from that region. The African Miocene erinaceids appear to have become extinct: although Gymnurechinus resembles Erinaceus in many ways it is too specialized in others to stand in the direct ancestry of the modern forms. Protechinus salis has been claimed as a possible derivative (Lavocat 1961); it has several primitive characters such as the transverse upper molars, but at the same time it has advanced to the modern erinaceine condition in the facial position of the lacrimal foramen and other features of the antorbital region. It is known only by a maxillary fragment, some incomplete mandibles, the lower dentition, and the upper cheek teeth. It could well be an early representative of the Erinaceus group that had invaded Africa from Europe or Asia in late Miocene time. Postpalerinaceus vireti Crusafont and Villalta (1948) from the early Pliocene of Spain shares some

Butler

specializations with Gymnurechinus and could be a migrant from Africa. At present most African Erinaceidae belong to the genus Erinaeeus, of which they form a subgenus Atelerix; the only exceptions are single species of the predominantly Asiatic genera Hemiechinus and Paraechinus that are probably very late immigrants into northern Africa. Atelerix is represented in the fossil record by E. (A.) broomi (= A. major Broom), known from the anterior part of a skull from Bolt's Farm, Transvaal, and from numerous dental and skeletal remains from Bed I, Olduvai (early Pleistocene). An analysis of its characters (Butler and Greenwood 1973) shows that E. broomi shares many primitive characters with E. (Atelerix) algirus, a Mediterranean species, and with E. europaeus; at the same time it shares some more specialized characters with the existing E. (A.) albiventris, widespread in tropical Africa, and E. (A.) sclateri from Somalia. The evidence is consistent with an invasion from the north of a stock resembling E. europaeus, of which E. algirus is the least modified surviving descendant. Thus there appear to have been at least two invasions of Erinaceidae into Africa, one at the beginning of the Miocene, the result of which subsequently died out, and the other perhaps at the beginning of the Pleistocene. The first was probably from southern Asia, the second from southwestern Asia or Europe. Protechinus may represent another invasion, perhaps late Miocene. Family Soricidae RECORDED SPECIES: Crocidura sp. Butler and Hopwood 1957. Early Miocene, Rusinga (?). "Sorex" dehmi africanus Lavocat 1961. A crocidurine (see Repenning 1967). Mio-Pliocene, Beni Mellal (Morocco). Myosorex robinsoni Meester 1955; Butler and Greenwood 1965. Late Pliocene and early Pleistocene, Makapansgat, Sterkfontein, Bolt's Farm, Swartkrans, Kromdraai, Olduvai. Sylvisorex cf granti Thomas. Butler and Greenwood (unpub.). Early Pleistocene, Olduvai. Sylvisorex sp. Butler and Greenwood 1965 (as Suncus cf lixus (Thomas)). Early Pleistocene, Olduvai. Suncus varilla (Thomas). Meester and Meyer 1972; Meester 1955 (as Suncus sp.). Late Pliocene and early Pleistocene, Makapansgat, Sterkfontein, Bolt's Farm, Sterkfontein Extension, Swartkrans, Kromdraai. Suncus cf varilla Butler and Greenwood 1965 (as S. cf orangiae (Roberts)). Early Pleistocene, Olduvai. Suncus infinitesimus (Heller). Meester and Meyer 1972.

Insectivora and Chiroptera

61

Late Pliocene and early Pleistocene, Sterkfontein, Sterkfontein Extension, Gladysvale, Makapansgat. Suncus cf infinitesimus. Butler and Greenwood 1965 (as Suncus sp. 1). Early Pleistocene, Olduvai. Crocidura taungensis Broom 1948; Meester 1955; De Graaf 1960. Late Pliocene, Taung. Crocidura cf hindei Thomas. Butler and Greenwood 1965. Early Pleistocene, Olduvai. Crocidura cf bicolor Bocage. Davis and Meester (unpub.). Latest Pliocene, Bolt's Farm. Diplomesodon fossorius Repenning 1965, 1967. Late Pliocene, Makapansgat.

Although the shrews are the most numerous insectivores in Africa at the present time, their fossil record in the Tertiary is extremely meager. All African shrews belong to the subfamily Crocidurinae, distinguished according to Repenning (1967) by their unpigmented teeth and by characters of the mandibular condyle and P 4 . Heim de Balsac and Lamotte (1956) noted a number of resemblances of Myosorex to Soricinae and proposed to unite the two subfamilies Soricinae and Crocidurinae. Though Myosorex clearly stands apart from other African shrews, its resemblances to Soricinae may consist of primitive characters lost in other Crocidurinae. Because the present distribution of Crocidurinae is mainly African and Oriental, Repenning (1967) believed that the group had originated south of the Tethys Sea, and entered Europe only in the Miocene. He identified the following European species as crocidurine: "Sorex" pusilliformis Doben-Florin (early Burdigalian), "S." dehmi Viret and Zapfe (late Burdigalian or Vindobonian), Miosorex grivensis (Deperet) (Vindobonian), and Soricella discrepans Doben-Florin (Burdigalian). The first of these has one lower tooth between I, and P4, as in all living crocidurines except Myosorex, in which an additional minute tooth is normally present anterior to P4. Miosorex and Soricella have two intermediate teeth and "Sorex" dehmi has three; reduction of teeth seems to take place in the region immediately anterior to P4. The range of dental formulae shown by these presumably immigrant forms suggest a source area inhabited by a wide variety of crocidurine shrews. It is unlikely that this area was Africa, for despite the discovery in the East African early Miocene of a fair number of specimens of tenrecids and other small mammals of similar size to shrews, only one soricid specimen has been recorded. Even this specimen is doubtful; it is a surface find from Rusinga, a mandibular ramus containing only M2 and a broken M1; but showing the alveoli of the other teeth. Insofar as it can be com-

62

Butler

pared, it closely resembles species of the Crocidura flavescens-C. occidentalis group, which are among the more advanced of living forms. The lack of primitive characters in such an early shrew is remarkable, and until more material is obtained the suspicion must remain that this specimen is of Recent origin. If the Rusinga record is rejected, the oldest African shrew would be "Sorex" dehmi africanus from the Mio-Pliocene of Morocco, regarded by Lavocat (1961) as only subspecifically different from "S." dehmi of the Miocene of Europe. It is known from two mandibular rami. Although its dental formula is very primitive, there is as yet no evidence that this form is close to the ancestry of the numerous living African crocidurines. It does, however, show that the subfamily had reached Africa by the beginning of the Pliocene. Several species are known from the Plio-Pleistocene of Olduvai and the Transvaal, but they are not sufficiently different from living forms to throw much light on the evolution of the group. For this it is necessary to fall back on a comparative study of the Recent species. The genus Myosorex (in which Surdisorex is included as a subgenus) has a wide distribution in tropical Africa, where most of the species are confined to high altitudes (Heim de Balsac and Lamotte 1956, Heim de Balsac 1967, 1968). Two closely related species, however, extend to South Africa (Meester 1958): M. cafer and M. varius, which are believed to represent successive southward migrations during Pleistocene pluvials. Farther north, on the Uluguru Mountains of southern Tanzania, occurs M. geatus which seems to belong to the same group; Heim de Balsac (1967) has associated with it a population from the eastern border of Rhodesia, included by Meester (1958) in M. cafer. M. robinsoni, from the Plio-Pleistocene of the Transvaal, is an early member of this group of species. It has some characters that seem to exclude it from the direct ancestry of M. cafer and M. varius and may represent an earlier migration. A form similar to M. robinsoni is common in Bed I at Olduvai. According to Heim de Balsac (1967) the most primitive living species of the genus is M. schallen, found in forests at low altitude south of Mount Ruwenzori, and an origin of the genus in equatorial Africa seems highly probable. Myosorex has fossorial adaptations and is confined to moist soil in forested areas or river banks. Sylvisorex and Suncus differ from Crocidura in the possession of an additional small tooth in the maxilla anterior to P4 (probably P3), present also in

Insectivora and Chiroptera

most species of Myosorex. The species of Sylvisorex have been reviewed by Heim de Balsac and Lamotte (1957), who point to the frequency of archaic characters in the genus, and regard it as broadly ancestral to Suncus and at least in part to Crocidura. Sylvisorex is restricted to tropical Africa, where five of the eight species are confined to mountains, two live in low hygrophilous forest and only one (S. megalura) extends to savanna. Two species occur in Olduvai Bed I. One of these, known only from the lower jaw and dentition, except I1; seems to be close toS. granti and S. megalura. The other is confined to the lower part of Bed I where it is the commonest shrew; it is known from numerous lower jaws and limb bones, but no upper teeth have been identified. It is a very distinct species, related apparently to S. johnstoni of West Africa, but much larger. Whereas Sylvisorex is confined to Africa, Suncus has a wide distribution in Eurasia, and the greatest diversity of forms occurs in the Oriental Region where the genus probably originated. Some species there—e.g., S. fellowes-gordoni from Ceylon—have primitive characters like Sylvisorex, such as a narrow mandibular condyle and a basined M3 talonid, and they cast some doubt on whether the two genera should be separated. Only six species of Suncus occur in Africa (Meester and Lambrechts 1971), including S. murinus, a commensal that has almost certainly been introduced in historical times. Fossils from Olduvai and the Transvaal cave breccias are close to the living S. varilla (including S. orangiae) and S. infinitesimus (including S. chriseos). Both these species are related to S. etruscus, an inhabitant of the Mediterranean region and southwestern Asia that has extended its range into West Africa south of the Sahara. It seems likely that this group of species entered Africa rather late. Crocidura is by far the largest genus of African shrews, and indeed one of the largest genera of living mammals. The diversity of its species on the continent strongly suggests an African origin, though the genus is widely distributed in Asia and Europe. There is some indication that Crocidura is polyphyletic, having been derived from a number of members of the Sylvisorex-Suncus group by loss of P3 (Heim de Balsac and Lamotte 1957). This tooth reappears as an atavism in several of the species (Meester 1953). Some of the most primitive species of Crocidura, such as C. maurisca and C. bottegi, are very close to Sylvisorex, while more advanced forms such as C. bicolor resemble Suncus. It is remarkable that in the early Pleistocene of Olduvai and the Transvaal Crocidura was less common than Suncus or Myosorex. The principal species at Olduvai, repre-

Butler

sented by about 20 mandibular fragments, seems to be a primitive relation of living species grouped around C. hirta and C. hindei; five additional specimens indicate the presence of three other species, probably Crocidura but too fragmentary to be identifiable. These figures may be contrasted with some 350 specimens from Olduvai referable to the genus Suncus. At Bolt's Farm, Transvaal, Crocidura is represented by a small species resembling C. bicolor (Davis and Meester, unpub.); a maxilla of another species from Taung was named C. taungensis by Broom (1948). Paracrocidura and Scutisorex are specialized forms of uncertain affinities and with no fossil record. Praesorex is perhaps merely a large form of Crocidura (Heim de Balsac and Lamotte 1957). The most remarkable fossil shrew so far reported from Africa is Diplomesodon fossorius Repenning (1965) from Makapansgat. It is known from more than 20 specimens, which show the mandible and the facial part of the skull. The type and only other species of the genus is D. pulchellum (Lichtenstein) from Central Asia, and the discovery of a South African species is of some zoogeographical interest. The two species agree in dental formula (they differ from Crocidura in having lost an antemolar from the maxilla), the short stout Ii, the short talonids of Μχ and M2 and their general robustness, but they differ in some respects. In D. fossorius the ventral part of the mandibular condyle is much wider transversely, exceeding Suncus and the most advanced species of Crocidura, whereas in D. pulchellum it is narrower as in primitive species of Crocidura; the talonid of M3 in D. fossorius is reduced to a cingulum, a greater degree of reduction than occurs in Crocidura, whereas in D. pulchellum, though simplified, it is no more reduced than in many species of Crocidura and Suncus; P4 oiD. fossorius has a well-developed metaconid, absent inZ). pulchellum, and I2 of D. fossorius is less reduced in comparison with P 4 . In view of the prevalence of parallel evolution in the Soricidae, the generic reference of D. fossorius should be treated with great reserve; it could well be a derivative of the Crocidura group paralleling Diplomesodon in adaptations to dry environmental conditions. Some soricid specimens have been obtained from Omo (H. Wesselman, pers. comm.), and also from Bulla Regia, Tunisia (ca 2.5 m.y.), and Irhoud-Ocre, Morocco (ca 2 m.y.) (J.-J. Jaeger, pers. comm.). None of this material has yet been described. This survey underlines the lack of information about the small mammals of Africa in the later Tertiary. The existence of Soricidae in Africa in the early Miocene requires confirmation. Their high de-

Insectivora and Chiroptera

63

gree of specific differentiation, reaching in some cases to the generic level, implies that they have lived in Africa for a period much longer than the Pleistocene, and their presence in the Mio-Pliocene of Morocco confirms this. Some groups, however, might have entered relatively late, such as the Suncus etruscus-S. varilla group. There is an indication that the main source of immigrants was the Oriental Region.

Family Tenrecidae RECORDED SPECIES: Protenrec tricuspis Butler and Hopwood 1957; Butler 1969. Early Miocene, Songhor, Rusinga, Napak. Erythrozootes chamerpes Butler and Hopwood 1957; Butler 1969. Early Miocene, Koru, Songhor, Napak. Geogale aletris Butler and Hopwood 1957. Early Miocene, Rusinga.

The African Tenrecidae now consist of the two genera of the subfamily Potamogalinae: Potamogale, with a single species from the Congo forest region, and Micropotamogale, with two species isolated on widely separated mountains, Nimba and Ruwenzori. The remaining subfamilies, Tenrecinae, Oryzorictinae, and Geogalinae, are confined to Madagascar. The fossil record is restricted to the Miocene of East Africa. Outside Africa, the Tenrecidae have usually been held to be related to the Solenodontidae of the West Indies and the Apternodontidae from the Oligocene of North America; Van Valen (1967) goes so far as to include the Apternodontidae within the Tenrecidae as a subfamily. Butselia, from the early Oligocene of Europe, was placed in the Tenrecoidea by Quinet and Misonne (1965), but additional material has shown that it falls better into the Plesiosoricidae (Butler 1972). Matthew (1913) traced the ancestry of the Tenrecidae to the North American Paleocene Palaeoryctes; on the other hand Butler (1956b) and McDowell (1958) pointed to their resemblance in many nonprimitive characters to other lipotyphlan insectivores, especially the Soricidae, and in a recent review Butler (1972) has argued that zalambdodont molar teeth do not form a satisfactory basis for insectivore classification. Whatever their relationships may be, it is not disputed that the Tenrecidae are an ancient, isolated group. Though unknown before the Miocene, it is very likely that they formed part of the Paleogene African fauna. In the Miocene there were three very distinct genera, implying an earlier radiation (Butler and Hopwood 1957, Butler 1969). Protenrec is

64

Butler

known from the antorbital part of a skull, the upper dentition as far forward as P3, and the mandible and lower dentition as far forward as the canine. Of Erythrozootes most of the skull is known, together with the entire upper dentition and the lower cheek teeth. Geogale aletris is represented by a single specimen consisting of the anterior part of a skull with worn teeth. Protenrec and Erythrozootes agree in that the protocones of the upper cheek teeth are much better developed than in any living tenrecid except Potamogale, and the talonids of the lower teeth are correspondingly comparatively large. However, neither of these Miocene genera approaches the tribosphenic condition as closely as does Potamogale, and if zalambdodonty is taken to be the derivative state, Potamogale must be more primitive. Micropotamogale, on the other hand, is more completely zalambdodont than the fossil forms. The Miocene genera have well-developed lacrimal canals, lost in Potamogalinae perhaps in adaptation to their aquatic habits. Protenrec is uniquely primitive for a tenrecid in the intraorbital position of its lacrimal foramen; it also has a long infraorbital canal like that of erinaceids. Erythrozootes shows another resemblance to erinaceids in the ossification of the medial wall of its alisphenoid canal. Its tympanic cavity and basicranium closely resemble those of the Madagascan Microgale and Geogale and differ from Potamogalinae. Erythrozootes differs from Protenrec in dental formula, having only two upper premolars; it is a much larger animal, with blunter cheek teeth and a rugose cranial surface. The two genera are evidently neither closely related to the living subfamilies nor to each other. They indicate that a greater variety of tenrecids probably existed in Africa in the middle Tertiary, including the unknown ancestors of the living African and Madagascan forms. In this connection Geogale aletris, the third Miocene form, is of interest. It shows several special resemblances to the living G. aurita from Madagascar: the enlarged first upper inciscors, widely separated from each other in the midline; reduction of the teeth anterior to P4; and the infraorbital canal, which is much longer than in other living Tenrecidae. Some differences may be significant, notably a backward prolongation of the palate in G. aletris, and parallelism cannot be ruled out. If the relationship is real, it would follow that Geogale reached Madagascar independently of the ancestor of the other Madagascan tenrecids, for it is too specialized to have given rise to them.

Insectivora and Chiroptera

Family Chrysochloridae RECORDED SPECIES: Prochrysochloris miocaenicus Butler and Hopwood 1957; Butler 1969. Early Miocene, Songhor, Koru. Proamblysomus antiquus Broom 1941. Plio-Pleistocene, Bolt's Farm. Chlorotalpa spelea Broom 1941. Late Pliocene, Sterkfontein. Amblysomus hamiltoni De Graaff 1958 (as Chrysotricha hamiltoni). Late Pliocene, Makapansgat.

Like the tenrecids, the golden moles are unknown before the Miocene, but presumably formed part of the early Tertiary African fauna. Their relationships are very obscure, mainly because of their extreme fossorial specialization. Their adaptations have been paralleled on other continents, notably in Epoiocatherium, a North American pholidote, Necrolestes, a South American marsupial, and Notoryctes, the Australian marsupial mole (Turnbull and Reed 1967). The Chrysochloridae are often grouped with the Tenrecidae, but the two families share very few characters that cannot be found in Soricidae and other Lipotyphla. Although zalambdodont cheek teeth occur in both families, the patterns differ in detail. The Miocene Prochrysochloris is known from the facial part of the skull and the upper and lower dentition. The skull seems to be typically chrysochlorid. The teeth, however, have some features shared with Miocene tenrecids: V-shaped protocones on the upper molars, comparatively large talonids on the lower molars, and a maximum molar emphasis on M 1-2 . Such resemblances cannot be taken as implying anything more than an approach of both families towards the primitive lipotyphlan condition. Classification of the living species has been discussed by several authors. Characters that have been used are the presence or absence of the posterior molars, the degree of reduction of the talonids of the lower cheek teeth, and proportions of the skull. Simonetta (1968) emphasized the size and shape of the head of the malleus (which is often greatly enlarged) and the morphology of the epitympanic recess in which it lies. He grouped the Chrysochloridae into three subfamilies with six genera and showed that loss of the lower molar talonids took place independently within each subfamily. Unfortunately Simonetta's classification cannot be applied to Prochrysochloris, the ear region of which is unknown. It resembles Eremitalpa in having a very broad, short face.

Butler

Insectivora and Chiroptera

Three species have been described from the PlioPleistocene of the Transvaal, but their exact relationships to living species remain to be determined. Chrysochlorids have not been recorded from Olduvai.

Order Chiroptera

The Pattern of Insectivore Evolution in Africa The fossil record of African insectivores is so scanty that only a few broad and tentative conclusions can be drawn. In the Oligocene there appears to have been a fauna containing ptolemaiids and macroscelidids, and almost certainly also tenrecids and chrysochlorids, families that evolved on the continent during its period of isolation. These families persisted into the Miocene, by which time the Macroscelididae and Tenrecidae had undergone an adaptive radiation. The Miocene saw the appearance in Africa of the Erinaceidae, immigrants presumably from southern Asia. Whether the Soricidae entered from Asia at the same time requires confirmation. A considerable faunistic change intervened between the early Miocene, as represented by the deposits in Kenya and Uganda, and the Plio-Pleistocene, as seen at Olduvai and the Transvaal. Ptolemaiidae disappear; Myohyracinae and Rhynchocyoninae are replaced by Macroscelidinae; Miocene erinaceids disappear and Erinaceus has invaded from the north; Soricidae become the dominant insectivores; Tenrecidae disappear; only the Chrysochloridae persist. Apart from Chrysochloridae, the only early Miocene insectivores that have survived into the modern fauna with little change are Rhynchocyon and the Potamogalinae, which are not represented in the Pliocene and Pleistocene record. A principal factor in such a change of fauna is most likely to have been the entry of mammals from the Oriental and Palearctic regions during the later Tertiary, but the lack of paleontological data on the small mammals of Africa and southern Asia of that time prohibits a more detailed analysis: the only direct evidence is the presence of Erinaceidae and Soricidae in the Mio-Pliocene of Morocco. Once the isolation of Africa had been ended, one can envisage a series of invasions. The early Pleistocene species of Erinaceus and Suncus are so closely related to extra-African forms that it is unlikely that they had been on the continent for very long, but the higher level of differentiation of most Soricidae suggests that their ancestors entered Africa much earlier, perhaps during the Miocene.

65

RECORDED SPECIES: Family Pteropodidae Propotto leakeyi Simpson 1967; Walker 1969. Early Miocene, Songhor, probably Rusinga.

Family Microchiroptera insertae sedis Vampyravus orientalis Schlosser 1910, 1911; Savage 1951; Russell and Sige 1970. Synonym: Provampyrus orientalis Schlosser 1911. Early Oligocene, Fayum. Indet. Lavocat 1961. Mio-Pliocene, Beni Mellal.

Family Emballonuridae Taphozous incognita Butler and Hopwood 1957 (as Saccolaimus incognita). Early Miocene, Koru. Indet. Butler 1969. Early Miocene, Rusinga.

Family Megadermatidae Indet. Butler and Hopwood 1957. Early Miocene, Rusinga. Afropterus gigas Lavocat 1961; Russell and Sige 1970. Mio-Pliocene, Beni Mellal. Cardioderma sp. Butler and Greenwood 1965. Early Pleistocene, Olduvai.

Family Rhinolophidae Rhinolophus Pliocene, Beni Rhinolophus Late Pliocene,

ferrumequinum millali Lavocat 1961. MioMellal. cf capensis Lichtenstein, De Graaf 1960. Makapansgat.

Family Hipposideridae Hipposideros sp. Butler 1969. Early Miocene, Songhor. Asellia (?) vetus Lavocat 1961. Mio-Pliocene, Beni Mellal.

Family Myzopodidae Myzopoda sp. Butler and Greenwood (unpub.). Early Pleistocene, Olduvai.

Family Vespertilionidae Indet. Lavocat 1961. Mio-Pliocene, Beni Mellal. Myotis sp. Broom 1948. Plio-Pleistocene, Bolt's Farm. Myotis sp. Butler and Greenwood (unpub.). Early Pleistocene, Olduvai. Cf Nycticeius (Scoteinus) schlieffeni (Peters), Butler and Greenwood 1965. Early Pleistocene, Olduvai. Cf Pipistrellus (Scotozous) rueppelli (Fischer), Butler and Greenwood 1965. Early Pleistocene, Olduvai. Eptesicus cf hottentotus (Smith), Butler and Greenwood (unpub.). Early Pleistocene, Olduvai. Miniopterus cf schreibersi (Kuhl) Butler and Greenwood (unpub.). Early Pleistocene, Olduvai.

66

Butler

Family Molossidae Indet. Lavocat 1961. Mio-Pliocene, Beni Mellal. Indet. (4 spp.). Butler and Greenwood (unpub.). Early Pleistocene, Olduvai.

The paleontological record of the Chiroptera in Africa, as in the world generally, is even more incomplete than that of the Insectivora. Hayman (1967) distinguishes 187 species now living on the African continent, of which 157 are endemic or extend only to southern Arabia or to Madagascar and adjacent islands. Those are placed in 43 genera representing 9 families, of which the Vespertilionidae is much the largest, with 75 species in 14 genera. The next largest are the Molossidae with 31 species in 6 genera, and the Pteropodidae with 27 species in 12 genera. The oldest fossil bat described from Africa is Vampyravus (= Provampyrus) orientalis, a humerus from the Oligocene of the Fayum. Though referred by Schlosser (1911) to the American family Phyllostomatidae, its morphology falls outside the known range of members of that family (Savage 1951) and the relationships of Vampyravus are problematic; Smith (1972) refers it tentatively to the Megadermatidae. Some chiropteran jaw material has recently been collected from the Fayum but not yet described (E. L. Simons, pers. comm.). Four families have been recognized in the early Miocene of East Africa. Of these, the Pteropodidae are represented by Propotto leakeyi, originally described by Simpson (1967) as a lorisid, but recognized as pteropodid by Walker (1969). It is known only from mandibular fragments. It shows some primitive features: the cheek teeth are less spaced than in most modern forms; the molars are very low crowned and provided with small blunt cusps, and the groove that runs along the length of the tooth in modern forms is not developed. If Archaeopteropus, from the Oligocene of Italy, is removed from the Megachiroptera following Russell and Sige (1970), Propotto becomes the oldest known member of the suborder. It does not seem likely, however, that the Pteropodidae are of African origin, as the family reaches its greatest diversity in the Oriental Region: apart from the Epomophorus group, which is confined to Africa, the African Pteropodidae belong to or are related to genera of mainly southern Asiatic distribution. The Emballonuridae are known from the early Miocene by a skull fragment (Butler and Hopwood 1957) and a humerus (Butler 1969). This family has a pantropical distribution and first appears in Europe in the early Oligocene. The Megadermatidae appear in the late Eocene or early Oligocene of Eu-

Insectivora and Chiroptera

rope and are now widely distributed in the Old World tropics. A poorly preserved fragment of mandible from the early Miocene of Rusinga has been referred to the family, and Russell and Sige (1970) have suggested that Afropterus gigas Lavocat (1961), based on some isolated molars from the MioPliocene of Morocco, is also a megadermatid. A maxilla and a mandible from Olduvai appear to represent an extinct species related to the existing Cardioderma cor. The Hipposideridae are represented by a humerus from the Miocene of Songhor (Butler 1969) and by a maxilla and a mandible from the Mio-Pliocene of Morocco (Lavocat 1961). The family goes back to the middle Eocene in Europe, and at present is widespread in the Old World tropics. The Nycteridae, at present confined to tropical Africa except for one Asiatic species, are known from undescribed Pliocene material from Kanapoi (B. Patterson, pers. comm.). The Rhinolophidae, which date from the late Eocene or early Oligocene of Europe, have been recorded in Africa only from the Mio-Pliocene of Morocco (Lavocat 1961) and the late Pliocene of Makapansgat (De Graaff 1960). It is somewhat surprising that the two largest families of living African bats, the Molossidae and the Vespertilionidae, together comprising more than half the existing species, have not been found in the Miocene of East Africa, although they go back to the beginning of the Oligocene in Europe. In view of the small sample (four or five species) of bats so far identified from the African Miocene, the absence of these families may be a matter of chance. Certainly in the Mio-Pliocene of Morocco both families were present, though known only from isolated molar teeth (Lavocat 1961). In the early Pleistocene of Olduvai, four species of Molossidae and five of Vespertilionidae have been distinguished on the basis of mandibles or humeri or both; thus the two families had reached their present-day dominance. A vespertilionid, Myotis sp. has been obtained from Bolt's Farm, Transvaal (Broom 1948). A humerus from Olduvai agrees very closely, except for its larger size, with that of Myzopoda aurita, the only living species in the family Myzopodidae, confined to Madagascar. This family, therefore, formerly existed on the African continent. Myzopoda may be an offshoot from the Vespertilionidae. The Rhinopomatidae, with one genus found in arid regions of southern Asia and northern Africa, are the only living family not represented in the fossil African record. They are probably late immigrants. Because of the mobility conferred on them by their

Butler

powers of flight, the Chiroptera are less useful as paleogeographical indicators than most other orders of mammals. Six of the existing African families are known from the Paleogene of Europe: Emballonuridae, Megadermatidae, Rhinolophidae, Hipposideridae, Molossidae, and Vespertilionidae. Three of these were in Africa in the early Miocene (Emballonuridae, Megadermatidae, Hipposideridae), but whether they had crossed the Tethys Sea at an earlier date or whether they had entered only after a land connection with Eurasia had been formed is unknown. The Pteropodidae probably evolved south of Tethys, but their existing distribution suggests an Asiatic rather than an African source, and it may be supposed that they entered Africa with the Erinaceidae at the beginning of the Miocene. The Molossidae and the Vespertilionidae, which dominate at the present day, appear in the African record, together with the Rhinolophidae, only in the Mio-Pliocene; it is possible that like the Soricidae they entered at a relatively late date, though further discovery may well disprove this. The Nycteridae appear from their existing distribution to have originated in Africa. The peculiar Myzopodidae are probably also of African origin, but they survive only in Madagascar. The Rhinopomatidae may be late immigrants from southern Asia like Paraechinus. Although the evidence is scanty, it does suggest a chiropteran history that parallels that of the Insectivora. Vampyravus is so far the only bat recorded from the Paleogene, when the continent was isolated and endemism must have predominated. Then at the beginning of the Miocene we can imagine a modernization brought about by the introduction of groups from Eurasia, followed by a second phase of modernization later in the Miocene. Faunistic interchange, particularly with the Oriental Region, must have played an important part in the later phases of bat evolution in Africa because many of the genera, and species groups within large genera, extend into southern Asia.

References Andrews, C. W. 1914. On the lower Miocene vertebrates from British East Africa collected by Dr. Felix Oswald. Q. J. Geol. Soc. Lond. 70:163-186. Broom, R. 1937. On some new Pleistocene mammals from limestone caves of the Transvaal. S. Afr. J. Sei. 33:750768. 1938. Note on the premolars of the elephant shrews. Ann. Transv. Mus. 19:251-252. 1941. On two Pleistocene golden moles. Ann. Transv. Mus. 20:215-216.

Insectivora and Chiroptera

67

1948. Some South African Pliocene and Pleistocene mammals. Ann. Transv. Mus. 21:1-38. Butler, P. M. 1948. On the evolution of the skull and teeth in the Erinaceidae, with special reference to fossil material in the British Museum. Proc. Zool. Soc. Lond. 118:446-500. 1956a. Erinaceidae from the Miocene of East Africa. Fossil Mammals of Africa, 11. London: British Museum (Natural History) . 1956b. The skull of Ictops and the classification of the Insectivora. Proc. Zool. Soc. Lond. 126:453-481. 1969. Insectivores and bats from the Miocene of East Africa: new material. In L. S. B. Leakey, ed., Fossil vertebrates of Africa, vol. 1. New York and London: Academic Press, pp. 1-38. 1972. The problem of insectivore classification. In K. A. Joysey and T. S. Kemp, eds., Studies in vertebrate evolution. Edinburgh: Oliver & Boyd, pp. 253-265. Butler, P. M., and M. Greenwood. 1965. Insectivora and Chiroptera. In L. S. B. Leakey, ed., Olduvai Gorge 19511961. Vol. 1: Fauna and background. Cambridge: Cambridge University Press, pp. 13-15. 1973. The early Pleistocene hedgehog from Olduvai, Tanzania. In L. S. B. Leakey, R. J. G. Savage, and S. C. Coryndon, eds.,Fossil vertebrates of Africa, vol. 3. London: Academic Press, pp. 7-42. 1976. Lower Pleistocene elephant-shrews (Macroscelididae) from Olduvai and Makapansgat. In R. J. G. Savage, and S. C. Coryndon, eds., Fossil Vertebrates of Africa, vol. 4. London: Academic Press, pp. 1-56. Early Pleistocene Soricidae and Chiroptera from Olduvai. In preparation. Butler, P. M., and A. T. Hopwood. 1957. Insectivora and Chiroptera from the Miocene rocks of Kenya Colony. Fossil Mammals of Africa 13. London: British Museum (Natural History), 35 pp. Coombs, N. C. 1971. Status of Simidectes (Insectivora, Pantolestoidae) of the late Eocene of North America. Am. Mus. Novit. 2455:1-41. Corbet, G. B. 1973. Family Macroscelididae. In J. Meester, ed., The mammals of Africa: an identification manual. Washington: Smithsonian Institution, part 1.5. Corbet, G. B., and J. Hanks. 1968. A revision of the elephant-shrews, family Macroscelididae. Bull. Br. Mus. Nat. Hist., Zoology 16:47-111. Crusafont Pairo, M., and J. F. de Villaita. 1948. Sur un nouveau Palerinaceus du Pontien d'Espagne. Eclog. Geol. Helv. 40:320-333. Davis, H. D. S., and J. Meester. Report on the microfauna in the University of California collections from the South African cave breccias. In preparation. De Graaff, G. 1958. A new chrysochlorid from Makapansgat. Palaeont. Afr. 5:21-27. 1960. A preliminary investigation of the mammalian microfauna in Pleistocene deposits of caves in the Transvaal system. Palaeont. Afr. 7:59-118. Hayman, R. W. 1967. Chiroptera. In J. Meester, ed., Preliminary identification manual for African mammals. Washington: Smithsonian Institution, part 11.

68

Butler

Heim de Balsac, H. 1967. Faits nouveaux concernant les Myosorex (Soricidae) de l'Afrique Orientale. Mammalia 31:610-628. Heim de Balsac, H., and M. Lamotte. 1957. Evolution et phylogenie des Soricides africains. II. La Lignee Sylvisorex-Suncus-Crocidura. Mammalia 21:15-49. Hopwood, Α. T. 1929. New and little-known mammals from the Miocene of Africa. Am. Mus. Novit. 344:1-9. Lavocat, R. 1961. Le gisement de vertebres miocenes de Beni Mellal (Maroc). fitude systematique de la faune de mammiferes. Notes Mem. Serv. Mines Carte geol. Maroc 155:29-94. McDowell, S. B. 1958. The Greater Antillean insectivores. Bull. Am. Mus. Nat. Hist. 115:115-214. McKenna, M. C. 1975. Toward a phylogenetic classification of the Mammalia. In W. P. Luckett and F. S. Szalay, eds., Phylogeny of the primates: a multidisciplinary approach. New York and London: Plenum, pp. 21-46. Matthew, W. D. 1910. Schlosser on Fayum mammals. A preliminary notice of Dr. Schlosser's studies upon the collection made in the Oligocene of Egypt for the Stuttgart Museum by Herr Markgraf. Am. Nat. 49:429-483. 1913. A zalambdodont insectivore from the middle Eocene. Bull. Am. Mus. Nat. Hist. 32:307-314. 1915. A revision of the Wasatch and Wind River faunas. Part IV—Entelonychia, Primates, Insectivora (part). Bull. Am. Mus. Nat. Hist. 34:429-483. 1918. A revision of the Lower Eocene Wasatch and Wind River faunas. Part V.—Insectivora (continued), Glires, Edentata .Bull. Am. Mus. Nat. Hist. 38:565-657. Meester, J. 1953. The genera of African shrews. Ann. Transv. Mus. 22:205-217. 1955. Fossil shrews of South Africa. Ann. Transv. Mus. 22:271-278. 1958. Variation in the shrew genus Myosorex in southern Africa. J. Mammal. 39:325-339. Meester, J., and A. von W. Lambrechts. 1971. The southern African species of Suncus Ehrenberg (Mammalia: Soricidae.) A««. Transv. Mus. 27: 1-14. Meester, J., and I. J. Meyer. 1972. Fossil Suncus (Mammalia: Soricidae) from southern Africa. Ann. Transv. Mus. 27:269-277. Osborn, H. F. 1908. New fossil mammals from the Fayum Oligocene, Egypt. Bull. Am. Mus. Nat. Hist. 24:265272. Patterson, B. 1965. The fossil elephant shrews (Family Macroscelididae). fiii/Z Mus. Comp. Zool. Harv. 133:295335. Quinet, G. E., and X. Misonne. 1965. Les insectivores zalambdodontes de l'Oligocene inferieur beige. Bull. Inst. Roy. Sei. Nat. Belg. 41:1-15. Repenning, C. A. 1965. An extinct shrew from the early Pleistocene of South Africa. J. Mammal. 46:189-196. 1967. Subfamilies and genera of the Soricidae. U. S. Geol. Surv., Professional Paper, 565. 74 pp. Russell, D. E., and B. Sige. 1970. Revision des chiropteres

Insectivora and Chiroptera

lutetiens de Messel (Hesse, Allemagne). Palaeovert. 3: 83-102. Savage, D. E. 1951. A Miocene phyllostomatid bat from Colombia, South America. Univ. Calif. Pubis. Bull. Dep. Geol. 28:357-366. Savage, R. J. G. 1965. Fossil mammals of Africa:19. The Miocene Carnivora of East Africa. Bull. Br. Mus. Nat. Hist. 10:241-316. Schlosser, Μ. 1910. Über einige fossile Saugetiere aus dem Oligocän von Ägypten. Zool. Anz. 53:500-508. 1911. Beiträge zur Kentniss der Oligozänen Landsäugetiere aus dem Fayum: Aegypten. Beitr. Paläont. Geol. Öst.-Ung. 24:51-167. 1923. Mammalia. In Κ. A. von Zittel, F. Broili, and M. Schlosser, eds., Grundzuge der Paläontologie (Paläozoologie), 4th ed. Munich: R. Oldenbough. Simonetta, Α. M. 1968. A new golden mole from Somalia with an appendix on the taxonomy of the family Chrysochloridae (Mammalia, Insectivora). Monitore Zool. Ital. (n.s.) 2 (supp.):27-55. Simons, E. L., and P. D. Gingerich. 1974. New carnivorous mammals from the Oligocene of Egypt. Ann. Geol. Surv. Egypt 4:157-166. Simpson, G. G. 1967. The Tertiary lorisiform primates of Africa. Bull. Mus. Comp. Zool. Harvard 136:39-62. Smith, J. D. 1976. Chiropteran evolution. In Biology of the Phyllostomatidae. Special publications of the Texas Technical University 1:49-69. Stromer, Ε. 1922. Erste Mitteilung über tertiäre Wirbeltier-Reste aus Deutsch-Südwestafrika. Sber. Bayer. Akad. Wiss. 1921:331-340. 1926. Reste Land- und Süsswasser-bewohnender Wirbeltiere Deutsch-Südwestafrikas. In E. Kaiser, ed., Die Diamantenwüste Südwestafrikas, vol. 2. Berlin: Dietrich Riemer, pp. 107-153. 1932. Palaeothentoides africanus nov. gen., nov. spec., ein erstes Beuteltier aus Afrika. Sber. bayer. Akad. Wiss. 1931:177-190. Szalay, F. S. 1977. Phylogenetic relationships and a classification of the eutherian Mammalia. In Major Patterns in Vertebrate Evolution, Μ. K. Hecht, P. C. Goody, and Β. M. Hecht, eds. NATO Advanced Study Inst., Ser. A 14:315-374. Turnbull, W. D., and C. A. Reed. 1967. Pseudochrysochloris, a specialized burrowing mammal from the early Oligocene of Wyoming. J. Paleont. 41:623-631. Van Valen, L. 1966. Deltatheridia, a new order of mammals. Bull. Am. Mus. Nat. Hist. 132:1-126. 1967. New Paleocene insectivores and insectivore classification. Bull. Am. Mus. Nat. Hist. 135:217-284. Walker, A. 1969. True affinities of Propotto leakeyi Simpson 1967. Nature 223: 647-648. Whitworth, T. 1954. The Miocene hyracoids of East Africa. Fossil Mammals of Africa, No. 7, British Museum (Natural History). 58 pp.

69

5 Rodentia and Lagomorpha Rene Lavocat

The Recent African rodent fauna is exceedingly complex. Some very important elements, for example, the Muroidea, Sciuroidea, and Gliroidea, have worldwide distribution; others, such as the Anomaluroidea, Pedetoidea, Ctenodactyloidea, and Bathyergoidea, are strictly limited to Africa. The Hystricidae and the genera Petromus and Thryonomys have long been considered to share very close morphological similarities with the South American rodents now placed in the Caviomorpha. Some authors even included Thryonomys and Petromus in subfamilies of the Caviomorpha. Transatlantic land bridges were usually assumed to have provided the opportunity for these New World forms to reach Africa in relatively recent times. On the other hand, zoologists have long recognized the importance of land connections between Africa and Asia for understanding the origin of at least part of the African rodent fauna, but the extent of Asiatic influence is by no means perfectly clear.

Distribution of the Recent Fauna The distribution of African rodent genera has been given by Ellerman (1940). Looking at the main pattern of this distribution we see a very sharp distinction between the Ethiopian and Palaearctic African regions. The Bathyergidae, Anomaluridae, Pedetidae, and Thryonomyidae are absent from the Palaearctic area, and the Sciuridae are represented there only by a single genus. Of the Muscardinidae, only two rather different genera occur here, one in each major biogeographic region. On the other hand, the Dipodidae and Ctenodactylidae are typically North African and can be found overlapping the Ethiopian zone only at a very narrow boundary. Only the Hystricidae, a few genera of Murinae, and three genera of Gerbillinae are really common to the two regions. In the Ethiopian region the Murinae and Sciuridae are very abundant, usually with wide generic distribution. Among the Thryonomyidae, Thryonomys occurs through most of the zone, in contrast to Petromus, which is limited to a rather narrow region of Southwest Africa in what seems to be a relict distribution. The distribution given by Ellerman for the Bathyergidae is "From Sudan, Abyssinia and Somaliland, and from the Gold Coast to the Cape" (1940:79). For the Anomaluridae he gives "Africa, Western and Central: from Sierra Leone to Uganda, Tanganyika and Northern Rhodesia" (1940:536). The distribution of the Pedetidae is given as "Central and Southern Africa: from Kenya and Angola to

70

Lavocat

Cape Province" (p. 547). The Graphiurinae are present throughout Africa south of the Sahara. Until recently a great many problems surrounded the African rodent fauna, such as origin, mode and time of immigration, and relationships. For example, which groups are autochthonous? When did the Cricetodontidae and the Muridae arrive in Africa? Must we recognize a relationship between Thryonomys and Petromus on the one hand and the Hystricidae on the other, and if so, what kind? With which of the Muroidea is the genus Mystromys related? What are the relationships between the Malagasy Nesomyidae and the African rodents? What roles have endemism and immigration played in the development of the African rodent fauna? What is the systematic position of the Bathyergidae? Is their infraorbital structure primitive or advanced? How does one explain the morphological similarities between Thryonomys and Petromus and the South American Caviomorpha? It is not possible to answer these questions without recourse to the historical facts. But until recently the history of African rodents was very poorly known. Now, as a result of recent field studies and laboratory research, enough information is available to provide an understanding of at least the main stages of this history, even if we cannot yet perceive all the details.

Sites and Collections Only those sites which have produced fossil rodents and those collections which have been studied are recorded here. Oligocene. The oldest levels from which rodents are known in Africa are the early Oligocene deposits of the Fayum, United Arab Republic. First worked by Andrews and Beadnell in 1901, further collections were made by the American Museum of Natural History. This material was studied by Osborn (1908). Collections in Stuttgart and Munich were studied by Schlosser (1911). New excavations in 1961-67 by Simons produced material now housed in the Yale Peabody Museum and studied by A. E. Wood (1968). Lower Miocene. Among the sites from this level are the Diamond Fields of Southwest Africa (Namib), studied by Stromer (1926). The material is now in the collections of the Munich Museum, although part of it was destroyed during World War II. Additional material in the American Museum was studied by Hopwood (1929). From East Africa, sites from the Lake Victoria region include Karungu, collections from which are

Rodentia and

Lagomorpha

now in the British Museum, and were studied by Andrews (1914); Rusinga, Mfwanganu, Songhor, and Koru, collections of the National Museum of Kenya studied by Maclnnes (1957) and by Lavocat (1967); Napak, Uganda, now in the Geology Department, Bedford College, London, and studied by Lavocat (1967); Bukwa, Uganda, of the Uganda Museum, studied by Lavocat (1967); Kirimun, northern Kenya, specimens in the University of California, Berkeley, and Loperot, collections of the National Museum of Kenya, both studied by Lavocat (1967). Upper Miocene. In East Africa, materials are from Fort Ternan, collections of the National Museum of Kenya, and Nakali, in the University of Madrid, both under study by Lavocat. In North Africa, the localities are Beni Mellal, now in Hautes Etudes, Paris, studied by Lavocat (1961), and by Jaeger; also from Oued Zra, in Morocco, under study by Jaeger. Upper Pliocene. This includes material from Lake Ichkeul in Tunisia, collected and studied by Jaeger (1971a); Makapansgat and Sterkfontein Caves in South Africa, studied by Lavocat (1967b) and de Graaf (1960). Pleistocene. In Morocco, specimens are from Jebel Irhoud, collected and studied by Jaeger (1971b). 01duvai Gorge in Tanzania has produced material of this age, studied briefly by Lavocat (1967b) and later by Jaeger. In South Africa, additional specimens from Swartkrans and Kromdraai have been studied by Lavocat (1967b) and de Graaf (1960). In Algeria, Ternifine is under study by Jaeger.

Definitions The definition of the Hystricomorph, Sciuromorph, and Myomorph structure is found in every handbook related to the rodent's anatomy; it is not so easy to find the definition of sciurognathy and hystricognathy. In a few words, the difference is that in the sciurognath, the posterior half of the mandible, and especially its inferior border, is in the same longitudinal plane as the incisor, whereas in the perfectly hystricognath structure, this posterior half is clearly in a longitudinal plane external to the plane of the incisor. Sometimes, the interpretation can be somewhat dubious, at least at first sight. But on the whole, we have here an excellent anatomical characteristic.

Systematics of African Rodents The revised diagnoses of important taxa established as a consequence of current research are

Lavocat

given below. Table 5.1 lists all recognized genera and gives their distribution by geological age and locality, as well as the origin of holotype material. Order Rodentia Suborder Hystricognathi Tulberg 1899 DIAGNOSIS. Rodents with an infraorbital foramen generally of great size, of the hystricomorph type, but secondarily reduced in one fossorial family. Mandible always hystricognath. Pterygoid fossa open anteriorly, communicating with the orbitotemporal cavity, or secondarily with the endocranial cavity. Maxillary bone nearly excluding palatine from orbitotemporal floor. Middle ear primitively with a very prominent and ventrally free promontory (Diamantomys, Lagostomus, Thryonomys), evolving in some forms toward a condition in which tympanic sheet covers the medial half of the promontory. Teeth of a primarily pentalophodont type. Most members with P4 or dP4 plus three molars, but P 3 persists in several families; in Bathyergoidea from three to six teeth. Infraorder Phiomorpha Lavocat 1962 DIAGNOSIS. Lower jaw always hystricognath. Infraorbital structure primarily hystricomorph, greatly modified and secondarily reduced in some Bathyergoidea. Molar teeth morphologically tetraor pentalophodont; greatly simplified in the Bathyergidae. Generally four cheek teeth, rarely three, sometimes five. DISTRIBUTION. Tropical to warm temperate of Old World. Superfamily Thryonomyoidea Wood 1955 DIAGNOSIS. Superfamily of Phiomorpha with brachydont to hypsodont dentition in which crown is not subdivided into many cusps. Anterior palatine foramina generally well developed. Frontal sinus sometimes developed. Generally four cheek teeth, sometimes five (with persistence of P3) in the upper jaw. Upper incisors not extending far posteriorly. DISTRIBUTION. Oligocene: Fayum; Miocene: East Africa, Namib, Morocco, Chios Island, Chinji of Indian Siwaliks; Recent: Ethiopian biogeographic region. Family Phiomyidae Wood 1955 DIAGNOSIS. Lower molars with from three to five transverse and more or less complete crests; upper molars with four to five crests; cusps still well individualized. Milk teeth with delayed replacement

Rodentia and Lagomorpha

71

in some forms, persistent in others. Molar form highly variable. P 3 or dP 3 present in some forms. DISTRIBUTION. Oligocene: Fayum; Miocene: East Africa, Namib. Family Thryonomyidae Pocock 1922 DIAGNOSIS. Thryonomyoidea with a muzzle of normal proportions; where known, masseter muscle insertion extending far in front of infraorbital foramen, and anterior palatine foramina well developed. Semihypsodont molars with well developed crests. The number of crests reduced in several forms. DISTRIBUTION. Oligocene: Fayum; lower Miocene: East Africa, Namib; upper Miocene: Morocco, East Africa (Fort Ternan), Chinji, Indian Siwaliks; Recent: Ethiopian biogeographic region. Family Diamantomyidae Schaub 1928 DIAGNOSIS. Thryonomyoidea with five upper cheek teeth, molariform, rather hypsodont and with secondary crests well developed. DISTRIBUTION. Oligocene: Fayum; Miocene: East Africa, Namib. Family Kenyamyidae Lavocat 1973 DIAGNOSIS. Thryonomyoidea with short masseteric insertion, four gliriform cheek teeth, with cusps weakly or not at all distinct, long and narrow crests, which may be low or high on a very brachydont crown. DISTRIBUTION. Lower Miocene: East Africa. Family Myophiomyidae Lavocat 1973 DIAGNOSIS. Small Thryonomyoidea. P 3 (or dP3) present at least in Myophiomys. Cheek teeth cricetodontoid with prominent cusps and lower crests. DISTRIBUTION. Oligocene: Fayum; lower Miocene: East Africa, Namib. Superfamily Bathyergoidea Osborn 1910 DIAGNOSIS. Fossorial Phiomorpha; infraorbital foramen primitively large; secondarily reduced in the Bathyergidae, very much so in the recent ones; incisors frequently very lengthened; number of cheek teeth varying, their structure simplified to extremely simplified. DISTRIBUTION. Miocene: East Africa, Namib; Recent: Africa south and east of a line from Togo to Ethiopia. Family Bathyergoididae Lavocat 1973 DIAGNOSIS. Bathyergoidea with rather conservative structure of cheek teeth. Incisors lengthened.

72

Lavocat

Rodentia and

Lagomorpha

Table 5.1 Distribution of the species of rodents in African Tertiary an»d Pleistocene layers. Oligocene

Lower Miocene

Upper Miocene Pliocene

Pleistocene

fa

Pi Ο l-H I 6 3 5, 3 a >• & fa ο. £ -d ft 3 § £>


48,000

-FISH

HOEK

HI »35,900? »40,500 ? KLASIES RIVER

100CA\

150-

US

BROKEN HILL (KABWE)

·» 110,000

200SALDANHA

250-

sigmoid sinus —> jugular bulb); substantial undercutting of lateral part of posterior wall of pyramid in posterior cranial fossae; deep, large middle cranial fossae, with undercut, robust lesser wings of sphenoid, steep undercut anteromedial portion of pyramid behind foramen ovale, and lacking Sylvian crest formation; anterior cranial fossae with high orbital roof, flattened anteriorly, quite deeply set cribriform plate of ethmoid, low crista galli, slender internal frontal crest. Calvarium elevated above upper margin of orbit due to high hafting of calvaria to facial skeleton; continuous supraorbital torus, with raised but inferiorly disposed glabellar area, and supraglabellar depression; relatively narrow interorbital space; orbits semirounded; broad ethmolacrimal articulation; rather low situation of nasion; nasal margin elevated, quite sharply defined, bordered inferolaterally by moderate continuation of canine pillars that pass superiorly into flattened maxillary planum; flattened intercanine subnasal area; usually slight to well-

Hominidae

marked nasal sill, some subnasal guttering, probably no anterior nasal spine; face moderately to substantially prognathous due to projection of subnasal maxilla with anterior dentition forward of frontal plane; persistance of premaxillary suture; robust, laterally flared malozygomatic region, with straight (unnotched) inferior border; malomaxillary area flat, inclined inferiorly; zygomatic pillar situated above M1 or even approximated to P4; palate with steep shelving far anteriorly, lacking a median palatine torus; molar portion of occlusal plane parallel to the Frankfurt Horizontal, the anterior portion slightly superiorly angled; anterior teeth set in a somewhat curved line. Mandible with ramus of moderate height, somewhat posteriorly inclined with strong endocoronoid and moderate endocondyloid crests; body generally robust with substantial lateral and marginal tori; symphysis with flattened anterior surface, posteriorly inclined with broad and moderate to substantial alveolar planum, moderate to marked superior and inferior transverse tori; internal mandibular arch contour may approximate a V shape. Dedicuous dentition with small spatulate dc having thickened basal lingual ridge with attendant depressions; dc with small (or no) mesial and larger distal cusplets, and both buccal and stronger lingual grooves; dm1 longer than broad, strong mesiobuccal angle of crown with mesial accessory cusp, probably with Carabelli's cusp, strong trigon (oblique) crest, and strong buccal grooves; drr^ incompletely molarized with five-cusped, strong protoconid with expanded, sloping buccal face, and inferiorly expanded enamel line, mesial accessory cusplet, deep, lingually situated anterior fovea with ill-developed anterior wall, large centrally situated metaconid, talonid relatively short compared to large trigonid and with less developed cusps (hypoconulid, entoconid); dm2 four-cusped with relatively small hypocone, salient buccal cusps, some even substantial expression of Carabelli complex, ill-developed distal wall of posterior fovea; dm2 with symmetrical Y-5 pattern, sixth cusp present or not, weak, oblique accessory groove distal of anterior fovea, trace of buccal cingulum, little secondary fissuration. Permanent dentition with harmoniously proportioned anterior and postcanine teeth; anterior teeth (crowns) relatively larger, and posterior teeth (roots) relatively less robust than in other species; premolars and molars of moderate size, and not buccolingually broad as in other species; upper incisors moderately shovel-shaped, Ρ with buccal grooves and flattened buccal face; lower incisors with substantial lingual hollowing, no (Ii) or weak fla) lingual grooves, and without lingual tubercle; upper C asymmetrical with pointed, projecting apex, strong vertical curvature, weak lingual grooves and adjacent ridges, except prominent vertical lingual ridge, sometimes with small lingual tubercle; lower C large relative to other species, strongly asymmetrical, strong distal buccal groove, marked lingual (especially distal) grooves and prominent lingual ridge, wide flat gingival eminence, large distal lingual cusplet; P3 sometimes > P4, usually two- but occasionally three-rooted, variably smooth to slightly crenulated enamel, P3 with well-defined buccal grooves, crown

Howell

usually wider at occlusal margin than cervical line, P 4 with ill-developed talon, equally expressed buccal grooves, and lacking distal lingual groove; P 3 < P 4 ; P 3 of fundamental hominid structure, rather markedly asymmetrical, lingual cusp well-defined, set rather mesial, buccal and lingual triangular ridges well expressed, buccal grooves well defined; P 4 with well-defined cusps; only moderate molarization, may be cuspule formation distal to posterior fovea and mesial to anterior fovea, foveae well developed, both buccal grooves well defined. Permanent molars overall smaller in length and breadth dimensions than other species, M3 customarily

« κ, a, ο .5? .C Κ ο« ?« 5δ § ε £ -S S «

3 α ε ο Ό ο

•S2

ο ε ε Ν £ £ „α Ο 9 ο ,Ο „Ο δ ? •s ο -S e Ρ c Ρ S

Pleistocene

Middle Early

Oligocene

j

/

/ Μ

\

Late

9

Ό Okapia sp.» Giraffa Giraffa ijumae·' ~7 ^ \ camelopardalis y \ \ GIRAFFINAE / Sivatherium »-Sivatherium giganteum / maurusium \ \ \ G. pygmaea »Giraffa sivalensis •fi \ \ G. stillei G. gracilis ' 1 V \ » w »Giraffa attica \

X

Pliocene Miocene

Giraffa , camelopardalis/

\\

Samotherium

spV\ \

\

/

\ /

\G. stillei /

»Giraffa punjabiensis »-Giraffokeryx punjabiensis »- Giraffa priscilla »-Samotherium boissieri »Palaeotragus spp

Palaeotragus Samotherium^G. jumae/?Helladotherium »Giraffokeryx germaini africanum / | ( Palaeotragus p r i m a e v u s \

V

Zarara zelteni

^

—

00

Samotherium sp.

\

\

> Prolibytherium magnieri —

?

Eumeryx sp.

Figure 25.9 Relationships of the African Giraffidae, revised on the basis of an original Miocene radiation from which Europe and Asia were subsequently colonized. Arrows suggest possible lineages but not necessarily direct descent; dotted lines separate subfamilies; carets indicate possible migrations from northern Africa into Eurasia; and question marks indicate putative origins or occurrences.

Churcher

lacking. Because the Asian Helladotherium (early to ?middle Pliocene) lacked or possessed only small ossicones, Bramatherium (late Pliocene) had paired upwardly directed frontoparietal ossicones and paired laterally directed parietal ossicones, and Hydaspitherium (late Pliocene) had two ossicones that were fused at their bases into a unit, while in Vishnutherium (late Pliocene) the condition is not recorded, it is unlikely that we can identify in an Asiatic genus the ancestral form from which Sivatherium may have derived. The African sivatheres therefore appear to belong to two indirectly related genera for which ancestral genera are hard to identify. The four-ossiconed Giraffokeryx is now known from Africa from late Miocene deposits at Lothagam, Kenya. There are now no known major palaeotragine forms unique to non-African areas and there are three primitive forms (Zarafa, P. primaevus, and Prolibytherium) known only from Africa. The giraffid radiation thus appears to have begun in Africa. The genus Giraffa is first known in the fossil record from the late Miocene of Africa. So little is known of these early giraffes that their direct origin and characteristics are still uncertain, although it is probable that they derived from a Samotheriumlike ancestor as far as size and the characteristics of the dentition indicate. The Giraffa cf jumae from Baringo Basin, Kanapoi, and Langebaanweg represent the late Miocene to early Pliocene African population. The earliest Asiatic record of Giraffa is G. priscilla (Pilgrim 1910) from the early Pliocene Chinji Zones of the lower Siwaliks of India. Early Pliocene G. attica from the Pontian of Pikermi and Salonika in Greece and in Asiatic Turkey represent the earliest Eurasian Mediterranean populations. The wide geographic distribution of Giraffa by the end of the early Pliocene suggests that the genus had been in existence for some considerable time and had become well adapted to a widespread niche. G. priscilla is founded on a left M3, a fragment of mandible, and a left M2 and a right M3. Matthew concluded that it may represent a primitive giraffine, although he inclined to consider it more probably a palaeomerycine. Colbert (1935) recorded Giraffa punjabiensis from the Dhok Pathan Zone of the Middle Siwaliks, of late middle Pliocene age, and confirmed Pilgrim's (1910) original establishment of G. priscilla. This was a small animal, smaller than G. camelopardalis or G. sivalensis, and possibly similar in size to G. gracilis. Its premolars are narrower and its lower molars bear well-developed cingula and buccal ectostylids. Giraffa sivalensis (syn. Camelopardalis affinis) is known from the Upper Siwaliks of early Pleistocene age. It is slightly smaller

Giraffidae

529

than G. camelopardalis and larger than G. punjabiensis, and both Pilgrim (1911) and Matthew (1929) considered it to be more advanced than the modern G. camelopardalis. Giraffa attica from the early Pliocene of Pikermi, Salonika, and Turkey, is smaller than G. gracilis, although it somewhat resembles that species, and perhaps these two species and G. sivalensis are closely related. Many of the early species, whether Asiatic (G. priscilla or G. punjabiensis), Eurasian (G. attica), or African (G. gracilis, G. stillei, or G. pygmaea), were more lightly built than the Pleistocene to Recent G. camelopardalis or G. sivalensis, although the extinct G. jumae apparently coexisted from late Miocene to late Pleistocene times with these smaller species. The modern African and now extinct late Pleistocene Asiatic giraffes derived from a savanna-dwelling ancestor, apparently resident in the SaharanArabian region during the late Miocene. G. priscilla would represent the Asian stock from which G. punjabiensis, G. sivalensis, and possibly the Chinese Honanotherium may have derived. G. attica would represent the Eurasian population corresponding to G. sivalensis in morphology. G. camelopardalis may derive from the G. jumae or G. gracilis stocks. Unfortunately the variation in size and morphological characters of modern G. camelopardalis is such as to render any conclusions on the limits of variability of the extinct Giraffa populations inconclusive. It is not inconceivable that the G. gracilis and G. jumae specimens represent the lesser and greater limits of size and morphological variations of a single population, the modern descendants of which we call G. camelopardalis. It is impossible on the present state of knowledge to do more than suggest stages in the giraffine lineage. A derivation from a condition of size and morphological complexity as seen in Palaeotragus primaevus through that seen in Samotherium to that of G. gracilis and ending in G. camelopardalis appears possible. Palaeotragus germaini would then represent a parallel or convergent evolution similar to Samotherium and G. stillei, a lineage of smaller giraffids that were probably distinct from the G. gracilis-G. camelopardalis lineage, while G. jumae may have represented the large individuals of that lineage. Excavations presently being carried out in the Siwalik deposits by many workers should throw further light on the chronology and relationships of the Asiatic Giraffidae with each other and with the African species. The three types of giraffid known from Africa probably represent adaptations to three different modes of life. The smaller and more lightly built Pa-

530

Churcher

laeotragus seems to have been a cursorial browser adapted to an open woodland environment, while the modern palaeotragine, Okapia johnstoni, had become progressively adapted to a closed forest environment. Large size might be a hindrance in the latter environment, and this factor may be responsible for Okapia being more or less similar in size to Palaeotragus. The larger Giraffa is adapted to an open environment where neither trees nor broken ground prevents the animal from escaping any potential predator. The long neck and legs allow the animal to browse on foliage that is out of reach of most of the other large herbivores and ensure the animal an adequate supply of food even in times of scarcity. The heavier and also larger sivatheres probably lived in habitats similar to that of giraffes. Meladze (1964) considered that sivatheres were savanna forms while Hamilton (1973) concludes that they were woodland or forest forms that fed on low vegetation or grasses on the woodland floor. The presence of Sivatherium and Giraffa spp. (including'Okapia" stillei) in the Olduvai and Laetolil deposits suggests that they may have shared the same habitat. The short neck of Sivatherium suggests a grazing or low browsing habit of foraging, similar to that of cattle or some of the larger African antelope or buffalo, and an adaptation to open woodland or bush veld. It is possible, therefore, to consider that the palaeotragines were specializing toward life in a thick bush or forest environment, that sivatheres lived in more open woodland or forest in which they could move without obstruction because of their size, and that giraffes occupied the open orchard savanna or bush veld and that all of these forms browsed but that the palaeotragines and sivatheres both ate forbs, herbs, low woody plants, and grasses, while giraffes mainly ate the foliage located at higher levels of the taller woody plants and trees. The variation in shape and the presence or absence of ossicones may reflect different habits or methods for intraspecific combat (figure 25.10). Sparring and fighting in Giraffa (Innis 1958; Dagg and Foster 1976) and Okapia (Walther 1960, 1962) involves lateral display leading to heavy blows of the head and ossicones to the sides of the neck and body of the opponent. Lateral display is observed in tylopods and ruminants to be the early stage in combat and may be considered primitive (Geist 1965). The change from lateral display to a swinging lateral buffet requires little behavioral modification and, in fact, would be enhanced by one opponent retreating. A long neck would help in delivering strong blows with the head and blunt ossicones

Giraffidae

would ensure that the opponent received bruises but was not badly damaged. The elongation of the neck in giraffes could then reflect selective advantages in both feeding and sparring habits and would have become a characteristic of the family by the late Miocene. The flattened and expanded ossicones of the sivatheres may indicate that the display of these ornaments, either laterally or frontally, was more important than their use as bludgeons and that the increased size was a major factor in intimidating the opponent. Frontal presentation may have led to head-to-head contact that developed into a shoving match. A short massive neck supported by a similarly short and massive body would provide the purchase and mass necessary to compete in such a trial of strength, as they would resist compression, be less likely to bend, and require great force to dislodge. In the early Miocene Prolibytherium magnieri the morphology of the cervical vertebrae and head suggests that already the sivathere line had adopted this mode of contest, and the continued success of the line and the increased complexity of the ossicones reached in the Plio-Pleistocene Indian and African sivatheres indicates that the shoving form of intraspecific combat was not abandoned. Some of the sivatheres were without ossicones (e.g., Helladotherium) and in Sivatherium the female may lack ossicones. The genus Indratherium was originally described as separate from Sivatherium but is now generally considered to represent the sexually dimorphic female of Sivatherium (Bohlin 1926; Matthew 1929; Colbert 1935). Both these conditions suggest that the size of the creatures was sufficient to reduce predation and to free the ossicones from any interspecific function. The pointed ossicones of modern Okapia (and probably of the extinct palaeotragines) are used in lateral display and the shafts for buffeting a rival opponent, but the points are brought into use only in interspecific combat with predators (Walther 1960, 1962). Dependence mainly on lateral display for intraspecific confrontations reduces the chance of damage to the opponent by the pointed ossicones and, when confronted with interspecific combat, maximizes the chance of seriously damaging the enemy. As the various lineages evolved, the larger giraffes and sivatheres reduced their susceptibility to predation and could depend on their large size, massiveness, and hooves to protect them. They could then specialize their intraspecific defenses. Because the okapis remained in a denser, forested habitat, their smaller size was an advantage that outweighed the advantages of increased size, and there-

Churcher

Giraffidae

531

F i g u r e 25.10 Fossil and living giraffids: (A) Zarafa zelteni from Gebel Zelten, Libya; early Miocene. (B) Samotherium sp., afterS. boissieri from Samos, Greece; Pliocene. (C) Giraffa camelopardalis rothschildi, Baringo or Rothschild's giraffe, from Uasin Gishu Plateau, Kenya, showing the three-horned condition; Recent. (D) Prolibytherium magnieri, from Gebel Zelten, Libya; early Miocene. (E) Sivatherium maurusium, from North, East, and South Africa; Pleistocene. (F) Palaeotragus primaevus, from Fort Ternan, Kenya; middle Miocene. (G) Okapia johnstoni, okapi from the Semliki Forest, Congo Basin, Zaire; Recent. ( Η ) Giraffa jumae, extinct giraffe, from Rawe, Kenya, and Olduvai Bed II, Tanzania; Pliocene and Pleistocene. (I) Giraffa camelopardalis camelopardalis, Nubian giraffe, from the Sudan; Recent. Heads not drawn to same scale; (Ε), (G), and (J) drawn to scale. fore t h e y r e t a i n e d t h e p o i n t e d o s s i c o n e s s u i t a b l e for t h i s h a b i t a t . S o m e of t h e p a l a e o t r a g i n e s (e.g., Giraf-

fokeryx punjabiensis,

Palaeotragus

quadricornis)

"experimented" w i t h four ossicones, a s did s o m e of t h e s i v a t h e r e s , b u t w i t h o u t i n c r e a s i n g t h e i r size to a n y g r e a t e x t e n t . P r e s u m a b l y t h e s e f o r m s still u s e d t h e i r o s s i c o n e s for d i s p l a y a n d for b u f f e t i n g , s i n c e t h e s h a f t s w e r e still m o r e v e r t i c a l t h a n l a t e r a l a n d t h e t i p s w e r e still pointed. I am most appreciative for all the assistance that I have received from W. R. Hamilton of the British Museum of Natural History, London, and from J. M. Harris of the National Museum of Kenya, Nairobi, both of whom allowed me to study their reports on African Giraffidae when they were still unpublished.

J. M. Harris also spent many hours discussing the taxonomy of the East African species of Giraffa, the morphology of Giraffa and Sivatherium, and the South African giraffid records. I am indebted to Η. B. S. Cooke of the Department of Geology, Dalhousie University, Halifax, Nova Scotia, for assisting me with the names of the geological stages and for editorial comments, many of which he offered before he became an editor for this volume. I thank R. Singer of the Department of Anatomy, University of Chicago, Illinois, and P. Robinson of the University of Colorado Museum, Boulder, Colorado, for reading drafts of the paper and commenting on various aspects of the geology and palaeontology of Africa and the Giraffidae. P. Robinson allowed me to examine specimens from Tunisia that have assisted me in my comprehension of the evolution of the African Giraffidae. I wish to thank W. J. White and T. Woolf, my assistants during the period when the paper was first prepared, for their help with

532

Churcher

geographic records and the bibliography, and M. L. Richardson, my assistant during the final preparation and revision, for her help and for allowing me to examine and cite materials in her care and for her critical reading of the paper. The chapter was originally completed in June 1974, and finally revised in November 1976.

References Abadie, J., J. Barbeau, and Y. Coppens. 1959. Une faune de vertebres villafranchiens au Tchad. C. R. hebd. Seances Acad. Sei., Ser. D, 248:3328-3330 Aguirre, E., and P. Leakey. 1974. Nakali, nueva fauna de Hipparion del Rift Valley de Kenya. Estud. Geol. 30: 219-227. Arambourg, C. 1947. Contribution ä l'etude geologique et paleontologique du bassin du Lac Rudolphe et de la basse vallee de l'Omo. Deuxieme partie, Paleontologie. Miss. Scient. Omo, 1932-1933, I, Geol.-Anthrop.: 232562. 1948a. Les vertebres fossiles des formations des plateaux Constantinois (Note preliminaire). Bull. Soc. Hist. Nat. Afr. Nord 38:45-48. 1948b. Un sivatherine nord africain: Libytherium maurusium Pomel. C. R. Soc. Geol. France 1948:177179. 1948c. Les mammiferes pleistocene d'Afrique. Bull. Soc. Geol. France, Ser. 5, 17:301-310. 1949. Les gisements de vertebres villafranchiens de l'Afrique du Nord. Bull. Soc. Geol. France, Ser. 5, 19:195-203. 1951. Observations sur les couches äHipparion de la vallee de l'Oued el Hammam (Algerie) sur l'epoque d'apparition de la faune de vertebres dite "pontienne." C. R. hebd. Seances Acad. Sei., Ser. D, 232:2464-2466. 1952. La paleontologie des vertebres en Afrique du Nord Frangaise. Congr. Geol. Internat., 19: Monog. Region., s.i., Algerie; hors serie, 1-63. 1954. La faune ä Hipparion de l'Oued el Hammam (Algerie). Congr. Geol. Internat., 19 Sess., Alger 1952. Assoc. Serv. Geol. Afr., Pt. 2, fasc. 21:295-302. 1958. La faune de vertebres Miocenes de l'Oued el Hammam (Oran, Algerie). C. R. Soc. Geol. France 6: 116-119. 1959. Vertebres continentaux du Miocene superieur de l'Afrique du Nord. Publ. Serv. Carte. Geol. Algerie (Alger), N.S., Paleont. Mem., No. 4:1-161. 1960. Precisions nouvelles sur Libytherium maurusium Pomel, Giraffide du Villafranchien d'Afrique. Bull. Soc. Geol. France, Ser. 7, 2:888-894. 1961. Prolibytherium magnieri, un Vellericorne nouveau du Burdigalien de Libye (Note preliminaire). C. R. Somm. Geol. France 1961(3):61-63. Arambourg, C., and M. Arnauld. 1949. Note sur les fouilles paleontologiques executees en 1947-48 et 1949 dans les gisements villafranchiens de la Garet Ichkeul. Bull. Soc. Sei. Nat. Tunisie 2:149-157. Arambourg, C., and R. Coque. 1958. Le gisement villa-

Giraffidae

franchien de 1'Αϊη Brimba (Sud-Tunisien) et sa faune. Bull. Soc. Geol. France 8:607-614. Bishop, W. W., G. R. Chapman, A. Hill, and J. A. Miller. 1971. Succession of Cainozoic vertebrate assemblages from the northern Kenya Rift Valley. Nature 233: 389-394. Bishop, W. W., and Μ. H. L. Pickford. 1975. Geology, fauna and palaeoenvironment of the Ngorora Formation, Kenya Rift Valley. Nature 254:185-192. Bohlin, B. 1926. Die Familie Giraffidae. Palaeont. Sinica, Ser. C, 4:1-179. Bone, E. L., and R. Singer. 1965. Hipparion from Langebaanweg, Cape Province, and a revision of the genus in Africa. Ann. S. Afr. Mus. 48:273-397. Burollet, P. F. 1956. Contribution ä l'etude stratigraphique de la Tunisie Centrale. Ann. Mines Geol. Tunisie 18:1-350. Churcher, C. S. 1970. Two new upper Miocene giraffids from Fort Ternan, Kenya, East Africa: Palaeotragus primaevus n. sp. and Samotherium africanum n. sp. In L. S. B. Leakey and R. J. G. Savage, eds., Fossil Vertebrates of Africa. London: Academic Press, Vol. 2, pp. 1-109. 1974. Sivatherium maurusium (Pomel) from the Swartkrans Australopithecine site, Transvaal (Mammalia, Giraffidae). Ann. Transv. Mus. 29:65-70. Colbert, Ε. H. 1933. A skull and mandible of Giraffokeryx punjabiensis Pilgrim. Am. Mus. Novit. No. 632:1-14. 1935. Siwalik mammals in the American Museum of Natural History. Trans. Am. Phil. Soc., N.S., 26: 1-401. 1936. Palaeotragus in the Tung Gur Formation of Mongolia. Am. Mus. Novit. No. 894:1-17. Cooke, Η. B. S. 1949. Fossil mammals of the Vaal River deposits. Mem. Geol. Surv. S. Afr. 35:1-117. 1955. Some fossil mammals in the South African Museum collections. Ann. S. Afr. Mus. 42:161-168. 1963. Pleistocene mammal faunas of Africa, with particular reference to southern Africa. In F. C. Howell and F. Bourliere, eds., African ecology and human evolution. Chicago: Aldine, pp. 65-116. 1974. The fossil mammals of Cornelia, O.F.S., South Africa. In K. W. Butzer, J. D. Clark, and Η. B. S. Cooke, The geology, archaeology and fossil mammals of the Cornelia Beds, O.F.S. Mem. Nas. Mus., Bloemfontein 9:63-88. Cooke, Η. B. S., and S. C. Coryndon. 1970. Pleistocene mammals from the Kaiso Formation and other related deposits in Uganda. In L. B. S. Leakey and R. J. G. Savage, eds., Fossil Vertebrates of Africa. London: Academic Press, vol. 2, pp. 107-224. Cooke, Η. B. S., and L. H. Wells. 1947. Fossil mammals from the Makapan Valley, Potgietersrust. III. Giraffidae. S. Afr. J. Sei. 43:232-235. Coppens, Y. 1960. Le quaternaire fossilifere de Koro-Toro (Tchad). Resultats d'une premiere mission. C. R. hebd. Seances Acad. Sei., Ser. D, 251:2385-2386. 1962. Deux gisements de vertebres villafranchiens

Churcher

du Tchad. Actes IV Pan-Afr. Congr. Prehist., Leopoldville, 1959:229-315. 1967. Les faunes de vertebres quaternaires du Tchad. In W. W. Bishop and J. D. Clark, eds. Background to evolution in Africa. Chicago: University of Chicago Press, pp. 89-97. 1971. Les vertebres villafranchiens de Tunisie: gisements nouveaux, signification. C. R. hebd. Seances Acad. Sei., Ser. D, 273:51-54. Coque, R. 1957. Decouverte d'un gisement de mammiferes villafranchiens dans le Sud-Tunisien. C. R. hebd. Seances Acad. Sei., Ser. D, 245:1069-1071. Coryndon, S. C. 1966. Preliminary report on some fossils from the Chiwondo Beds of the Karonga District, Malawi. Am. Anthrop. 68:59-66. Cross, M. W., and V. J. Maglio. 1975. A bibliography of the fossil mammals of Africa, 1950-1972. Princeton: Princeton Univ., 291 pp. Crusafont, M. 1952. Los jiräfidos fosiles de Espana. Mem. Comm. Inst. Geol. Barcelona 8:1-239. Dagg, A. I., and J. B. Foster. 1976. The giraffe: its biology, behavior, and ecology. New York: Van Nostrand Reinhold, 210 pp. Dietrich, W. O. 1941. Die säugetierpaläontologischen Ergebnisse der Kohl-Larsen'schen Expedition, 1937-1939 in nördlichen Deutsch-Ostafrika. Zentbl. Min. Geol. Paläont. 1941B:217-223. 1942. Ältestquartäre Säugetiere aus der südlichen Serengeti, Deutsch-Ostafrika. Palaeontographica, Abt. A, 94:1-133. Dreyer, T. F., and A. Lyle. 1931. New fossil mammals and man from South Africa. Bloemfontein: Dept. Zool., Grey Univ. Coll., 60 pp. Flamand, G. Β. M. 1902 (1901). Sur l'utilisation, comme instruments neolithiques, de coquilles fossiles ä taille intentionale (Littoral du Nord-Africain). C. R. Assoc. Frang. Avanc. Sei. 30:729-734. Geist, V. 1965. The evolution of horn-like organs. Behaviour 27:175-214. 1971. The relation of social evolution and dispersal in ungulates during the Pleistocene, with emphasis on the Old World deer and the genus Bison. Quartern. Res. 1:285-315. Godina, A. Y. 1962. Family Giraffidae Gray, 1821. Giraffes. In Osnovy paleontologii (Fundamentals of paleontology), Orlov, Yu Α., chief ed.; Vol. 13, Gromova, V. I., Ed. Moskva: Akademiya Nauk SSSR, pp. 379-383. Trans. Israel Program Sei. Translat., Jerusalem, 1968, pp. 523-530. Hamilton, W. R. 1973. The lower Miocene ruminants of Gebel Zelten, Libya. Bull. Brit. Mus. Nat. Hist. (Geol.) 21:73-150. Harris, J. M. 1974. Orientation and variability in the ossicones of African Sivatheriinae (Mammalia; Giraffidae). Ann. S. Afr. Mus. 65(6):189-198. 1976a. Pliocene Giraffoidea (Mammalia, Artiodactyla) from the Cape Province. Ann. S. Afr. Mus. 69(12):325-353.

Giraffidae

533

1976b. Pleistocene Giraffidae (Mammalia, Artiodactyla) from East Rudolf, Kenya. In R. J. G. Savage, ed., Fossil Vertebrates ofAfrica. London: Academic Press, vol. 4, pp. 283-332. Haughton, S. H. 1922. A note on some fossils from the Vaal River gravels. Trans. Geol. Soc. S. Afr. 24:11-16. Heintz, E. 1975. Origine, migration et paleobiogeographie des Palaeotraginae (Giraffidae, Artiodactyla) antevallesiens de l'ancien monde. Coll. Internat., C.N.R.S., No. 218:723-730. Hendey, Q. B. 1968. New Quaternary fossil sites near Swartklip, Cape Province. Ann. S. Afr. Mus. 52(2):4373. 1969. Quaternary vertebrate fossil sites in the southwestern Cape Province. S. Afr. Archaeol. Bull. 24(3 & 4):96-105. 1970a. A review of the geology and palaeontology of the Plio/Pleistocene deposits at Langebaanweg, Cape Province. Ann. S. Afr. Mus. 56(2):75-117. 1970b. The age of the fossiliferous deposits at Langebaanweg, Cape Province. Ann. S. Afr. Mus. 56(3): 119-131. 1972. Further observations on the age of the mammalian fauna from Langebaanweg, Cape Province. In E. M. van Zinderen Bakker, Sr., ed., Palaeoecology of Africa, the surrounding islands and Antarctica. Cape Town: Balkema, vol. 6, pp. 172-176. 1974a. The late Cenozoic Carnivora of the southwestern Cape Province. Ann. S. Afr. Mus. 63:1-369. 1974b. Faunal dating of the late Cenozoic of Southern Africa, with special reference to the Carnivora. Quartern. Res. 4(2):49-161. 1976. The Pliocene fossil occurrences in "E" Quarry, Langebaanweg, South Africa. Ann. S. Afr. Mus. 69(9):215-217. Hopwood, Α. Τ. 1928. Mammalia. In W. P. Pycraft et al., eds., Rhodesian man and other associated remains. London: Trustees of the British Museum, pp. 70-73. 1934. New fossil mammals from Olduvai, Tanganyika Territory. Ann. Mag. Nat. Hist. Lond., Ser. 10, 14:546-550. 1936. New and little-known fossil mammals from the Pleistocene of Kenya Colony and Tanganyika Territory-I. Ann. Mag. Nat. Hist. Lond., Ser. 10, 17:636-641. Hopwood, A. T., and J. P. Hollyfield. 1954. An annotated bibliography of the fossil mammals of Africa (17421950). Brit. Mus. Nat. Hist., Fossil Mammals of Africa, No. 8:1-194. Howe, B., and H. L. Movius. 1947. A stone age cave in Tangier. Pap. Peabody Mus. Amer. Archaeol. Ethnol. Harvard 28(l):l-23. Innis, A. 1958. The behaviour of the giraffe, Giraffa camelopardalis, in the eastern Transvaal. Proc. Zool. Soc. London 131(2):245-278. Jaeger, J. G., J. Michaux, and B. David. 1973. Biochronologie du Miocene moyen et superieur continental du Maghreb. C. R. hebd. Seances Acad. Sei., Ser. D, 277:2477-2480.

534

Churcher

Joleaud, L. 1926. Etudes de geographie zoologique sur la Berberie (Les ruminants cervicornes). Glasn. Hrv. Pirodasl. Dr., Zagreb, 37/38:263-322. 1934. Vertebres subfossiles de L'Azaoua (Colonie du Niger). C. R. hebd. Seances Acad. Sei., Ser. D, 198:599-601. 1936. Gisements de vertebres quaternaires du Sahara. Bull. Soc. Hist. Nat. Afr. Nord 26:23-29. 1937. Remarques sur les giraffides fossiles d'Afrique. Mammalia 1:85-96. Kent, P. E. 1942a. A note on Pleistocene deposits near Lake Manyara, Tanganyika. Geol. Mag. 79:72-77. 1942b. The Pleistocene beds of Kanam and Kanjera, Kavirondo, Kenya. Geol. Mag. 79:117-132. Kohl-Larsen, L. 1939. Vorlaüfiger Bericht über meine Afrika-Expedition 1937-1939. Forsch. Fortschr. 15: 339-340. Lankester, E. R. 1901. On Okapia, a new genus of Giraffidae, from central Africa. Trans. Zool. Soc. London 16(1):279-314. 1910. Monograph of the okapi. Atlas. London: Trustees of the British Museum, pp. i-viii. Lavocat, R. 1961. Le gisement de vertebres de Beni Mellal (Maroc). fitude systematique de la faune de mammiferes et conclusions generates. Notes Mem. Serv. Geol. Maroc 155:29-94 & 109-145. Leakey, L. S. B. 1959. A preliminary reassessment of the fossil fauna from Broken Hill, Northern Rhodesia, pp. 225-231. In J. D. Clark, ed., Further excavations at Broken Hill, Northern Rhodesia. J. Roy. Anthrop. Inst. Gr. Brit. 89:201-232. 1965. Olduvai Gorge 1951-1961, Vol. l.A preliminary report on the geology and fauna. Cambridge: Cambridge University Press, 109 pp. 1970. Additional information on the status of Giraffa jumae from East Africa. In L. B. S. Leakey and R. J. G. Savage, eds., Fossil Vertebrates of Africa. London: Academic Press, vol. 2, pp. 325-330. Madden, C. T. 1972. Miocene mammals, stratigraphy and environment of Muruorot Hill, Kenya. Paleobios 14:112.

Maglio, V. J. 1971. Vertebrate faunas from the Kubi Algi, Koobi Fora and Ileret Areas, East Rudolf, Kenya. Nature 231:248-249. 1972. Vertebrate faunas and chronology of the hominid-bearing sediments east of Lake Rudolf, Kenya. Nature 239:379-385. 1974. Late Tertiary fossil vertebrate successions in the northern Gregory Rift, East Africa. Ann. Geol. Surv. Egypt 4:269-286. Matthew, W. D. 1929. Critical observations upon Siwalik mammals. Bull. Am. Mus. Nat. Hist. 56:437-560. Mawby, J. E. 1970. Fossil vertebrates from northern Malawi: preliminary report. Quaternaria 13:319-323. Meladze, G. K. 1964. [About the paleobiological study of Sivatheriinae.] Soob. Akad. Nauk Gruz. SSR 33(3): 567-600. Oakley, K. P. 1954. Study tour of early hominid sites in

Giraffidae

southern Africa, 1953. S. Afr. Archaeol. Bull. 9(35):7587. Pallary, P. 1900. Note sur le girafe et le chameau du quaternaire algerien. Bull. Soc. Geol. France, Ser. 3, 28: 908-909. Pickford, M. 1975. Late Miocene sediments and fossils from the northern Kenya Rift Valley. Nature 256:279284. Pilgrim, G. E. 1910. Notices of new mammalian genera and species from the Tertiaries of India. Ree. Geol. Surv. India 40(1):63-71. 1911. The fossil Giraffidae of India. Palaeont. Indica, N.S. 4(l):l-29. Pomel, A. 1892. Sur 1 eLibytherium maurusium, grand ruminant du terrain pliocene de l'Algerie. C. R. hebd. Seances Acad. Sei., Ser. D, 115:100-102. 1893. Cameliens et Cervides. Carte Geol. Algerie, Paleont., Monogr., pp. 1-52. Reck, H., and L. Kohl-Larsen. 1936. Erster Uberblick über die jungdiluvialen Tier- und Menschenfunde Dr. Kohl-Larsens im Nordöstlichen Teil des NjarasaGrabens (Östafrika) und die Geologischen Verhältnisse des Fundgebietes. Geol. Rundschau 27:401-441. Robinson, P., and C. C. Black. 1969. Note preliminaire sur les vertebres fossiles du Vindobonien (formation Beglia), du Bled Dourah, Gouvernorat de Gafsa, Tunisie. Notes Serv. Geol. Tunisie 31:67-70. 1974. Vertebrate faunas from the Neogene of Tunisia. Ann. Geol. Surv. Egypt 4:319-332. Roman, F. 1931. Description de la faune pontique de Djerid (El Hamma et Nefta). Ann. Univ. Lyon, N.S., Sei. Med, l(48):30-42. Roman, F , and M. Solignac. 1934. Decouverte d'un gisement de mammiferes pontiens ä Douaria (Tunisie septentrionale). C. R. hebd. Siances Acad. Sei., Ser D, 199:1649-1650. Romer, A. S. 1928. Pleistocene mammals of Algeria. Fauna of the paleolithic station of Mechta-el-Arbi. Bull. Logan Mus. 1:80-163. 1966. Vertebrate paleontology. Edit. III. Chicago: University of Chicago Press, 468 pp. Singer, R. 1957. Recent discoveries in southern Africa. Mitteil. Anthrop. Gesellsch. Wien 88/89:133-136. Singer, R , and E. L. Bone. 1960. Modern giraffes and the fossil giraffids of Africa. Ann. S. Afr. Mus. 45(4):375603. Singer, R , and J. R. Crawford. 1958. The significance of the archaeological discoveries at Hopefield. J. Roy. Anthrop. Inst. Gr. Brit. 88(1):11-19. Stromer, Ε. 1907. Fossile Wirbeltier-Reste aus dem Uadi Färegh und Uadi Natrün in Ägypten. Abh. Senkenb. Naturf. Ges. 29(2):97-132. Suess, M. 1932. Sur la presence de gastropodes et de vertebres dans le gres de Bou Hanifa, feuille de Mascara (department d'Oran). C. R. hebd. Seances Acad. Sei., Ser. D, 194:1970-1972. Thenius, E. 1967. Die Giraffen und ihre Vorfahren. Kosmos 63(4):160-164.

Churcher

Trevor, J. C. 1953. A new fossil man from South Africa. Nature 172:652. van Hoepen, E. C. N. 1932. Voorlopige beskryving van Vrystaatse soogdiere. Paleont. Navors. Nas. Mus., Bloemfontein 2:63-65. Vaufrey, R. 1947. Olorgesaillie. Un site acheuleen d'une exceptionelle richesse. Anthropologie 51:367. Walker, E. P. 1969. Mammals of the world, vol. II. Baltimore: Johns Hopkins Press, pp. 647-1500.

Giraffidae

535

Walther, F. 1960. "Antilopenhafte" Verhalfensweisen im Paarungszeremoniell des Okapi (Okapia johnstoni Sclater, 1901). Zeits. Tierpsychol. 17:188-210. 1962. Uber ein Spiel bei Okapia johnstoni. Zeits. Säugetierk. 27:245-251. Wells, L. H. 1959. The nomenclature of the South African fossil equids. S. Afr. J. Sei. 55(3)64-66.

536

26 Tragulidae and Camelidae Alan W. Gentry

Families Tragulidae and Gelocidae The main living subgroup of the Ruminantia, the infraorder Pecora, comprises the families Giraffidae (giraffe and okapi), Cervidae (deer), Antilocapridae (north American pronghorn), and Bovidae (cattle, sheep, goats, and antelopes). Also surviving today is the less advanced family Tragulidae (Asian chevrotains Tragulus and African water chevrotain Hyemoschus) belonging to the infraorder Tragulina. Apart from some poorly known Eocene-Oligocene European fossils referred to the family Amphimerycidae, fossil tragulines are referable either to the surviving and more primitive family Tragulidae or to a more advanced extinct family Gelocidae. Both families occur in the African fossil record, where nearly all occurrences are in East Africa. The Tragulidae may be defined as ruminants lacking frontal appendages, with a bunodont or primitive selenodont dentition, large upper canines in males, with fusion of the two central metapodials to form a cannon bone in the hind leg and sometimes in the front leg as well, complete side metapodials, the ectocuneiform joined with the naviculo-cuboid, and often a bony carapace above the pelvic girdle in males. The mainly Asiatic Gelocidae differ from tragulids in having more advanced selenodonty, strong cingula on the cheek teeth, Pi separate from the other premolars by a short diastema, no fusion of ectocuneiform with naviculo-cuboid, and the side metatapodials with entirely atrophied shafts. Gelocus whitworthi Hamilton (1973, p. 140), known by mandibular pieces from Songhor and Rusinga, is the first member of this family to be recognized in Africa. Hopwood (1929, p. 5) had referred a much worn dentition from the Miocene of Namibia to "tragulid indet. cf Bachitherium," but the family status of this piece cannot be regarded as established.

Fossil Tragulids of Africa All African fossil tragulids are referred to Dorcatherium, a Miocene form only doubtfully distinguishable from Hyemoschus by the probable presence of a preorbital fossa (now lost), the premaxilla contacting the nasal, more compressed bladelike lower premolars, and the occasional persistence of Pi. It differs from Tragulus only by the probable preorbital fossa, the occasional P1; and the lack of a cannon bone in the forelimb. There seems every likelihood that Tragulus and Hyemoschus will eventually be sunk in Dorcatherium. No described African Dorcatherium is as large as the European Vindobonian species D, peneckei Hofmann (including D.

Gentry

Tragulidae and Camelidae

537

rogeri?) with M r M 3 up to 50 mm long, or the Siwaliks Chinji species D. majus Lydekker. On the other hand, AfricanDorcatherium includes species smaller than those fossilized in the Miocene of other continents and as small as the living Tragulus of tropical Asia. Dorcatherium chappuisi Arambourg (1933) was described on a mandible from Losodok (= Moruorot), and Whitworth (1958) referred to it much material from Rusinga (mainly Hiwegi Formation as understood by van Couvering and Miller [1969] but also from the Kiahera Formation), Mfwanganu, and possibly Maboko. Gentry (1970, p. 301) thought it was present at Fort Ternan, but perhaps this material is from a more advanced species. D. chappuisi has M r M 3 with an occlusal length of about 43 mm, about the same size as in D. naui Kaup, the type species of the genus from Eppelsheim, Germany (which includes D. crassum from Sansan, France). However, D. chappuisi differs from the latter by its more bunodont molars, less prominent mesostyle on the upper molars, and a shallower mandibular ramus, all of which are more primitive characters. It also differs by having no basal pillars on the lower molars and no cingulum on the upper or lower molars. D. chappuisi differs by the same characters, insofar as they are known, from the similarly sized Chinji to Dhok PathanZ). minus Lydekker. It is interesting that this African form is so very distinctive; its apparently advanced condition in not having basal pillars or cingula suggests that it could be generically separated from the main Tragulus-DorcatheriumHyemoschus stock. It is also interesting that it retained an independent ectocuneiform in its tarsus, as noted by Whitworth (1958, p. 41), unlike other tragulids. In many of its characters it is like the Chinji and Nagri Dorcabune Pilgrim, which is not represented in Africa and which has been taken as a tragulid (Colbert 1935, pp. 301-306). However, the alternative possibility that Dorcabune is an anthracothere has not yet been eliminated.

Hopwood (1929, p. 5) recorded a large tragulid from Namibia. D. libiensis Hamilton (1973) from Gebel Zelten, Libya, has teeth almost as large as in D. chappuisi but without the latter's distinctive characters. D. pigotti Whitworth (1958) is a smaller species than D. chappuisi with M!-M3 having an occlusal length of about 30 mm. It is recorded from the Hiwegi and Kiahera Formations of Rusinga, Mfwanganu, Moruorot (Madden 1972), Karungu, and Ombo, all in Kenya, and from Bukwa, Uganda. It is slightly smaller than the European fossils named D. puyhauberti Arambourg and Piveteau 1929 from Salonica. D. parvum Whitworth (1958) is a still smaller species with Μχ-Μ3 having an occlusal length of about 20 mm. Some individuals are known to have lacked Pi. It is known from the Hiwegi and Kiahera Formations of Rusinga, Karungu, Maboko, and Moruorot in Kenya and from Bukwa in Uganda. D. songhorensis Whitworth (1958) is the Songhor species in which the Mi-M3 is about 24 mm long and thus intermediate in size between D. pigotti and D. parvum. A second species of Dorcatherium at Fort Ternan, smaller than D. chappuisi, was recorded by Gentry (1970, p. 302). Bishop and Pickford (1975) recorded two species of tragulids at Ngorora, Kenya. Nobody has yet investigated the possible relationship of these small African tragulids to the living tropical Asian Tragulus. At the present time the only surviving African tragulid, Hyemoschus aquaticus, has become restricted to parts of West Africa. The tragulids disappeared from the Palaearctic region in the Miocene, and their place was taken, according to Flerov (1971), by the cold-adapted musk deer, Moschus, an isolated form with bovidlike teeth.

The extent to which Pi occurs in Miocene tragulids is not known. It is supposed to be absent in most of the Sansan specimens but occurs in some and also in the Eppelsheim holotype of Dorcatherium naui. Probably this tooth was tending to become smaller and to fall out earlier in life, and this process may have gone on at different evolutionary rates in different stocks. Without large samples for study it is not possible to determine the significance of this feature. The holotype of D. chappuisi has a larger Px than the holotype of D. naui. A large tragulid other thanD. chappuisi has been found at Bukwa, Uganda (Walker 1969, p. 592).

Family Camelidae Camels and llamas are placed in a distinct suborder, Tylopoda, of the Artiodactyla. There are two living camels, the two-humped Bactrian camel, Camelus bactrianus, and the one-humped dromedary, which is not known in the wild state. The dromedary is found in North Africa and parts of southwestern Asia and is distinguished from the Bactrian camel mainly by its orbits being directed more downward (at least in many skulls), a smaller or completely absent ethmoidal fissure, less relative displacement of anterior and posterior parts of the medial wall of the lower molars (Grattard, Howell, and Coppens 1976,

538

Gentry

p. 269), and more slender limb bones with relatively longer distal elements. Until a decade ago camels were believed to have been introduced into sub-Saharan Africa relatively recently. Fossil evidence for their earlier presence is only just coming to light.

Fossil Camels of Africa An extinct wild camel is known in Africa as early as the Pliocene. Howell, Fichter, and Wolff (1969) recorded a damaged lower molar from Member Β and a distal metatarsal from Member F in the Shungura Formation, lower Omo Valley, Ethiopia. Subsequently Grattard, Howell, and Coppens (1976, p. 268, table 1) established its presence in Members D and G as well. Gentry and Gentry (1969) recorded two camel molars from Marsabit Road, Kenya, and from site BK, upper Bed II, Olduvai Gorge, Tanzania, the latter being the southernmost record in Africa. A camel is also known in North Africa, from Pliocene beds equivalent in age to the lower Villafranchian (Arambourg 1962, p. 104; Coppens 1971). North African material is usually referred to Camelus thomasi Pomel (1893), a species described and well illustrated on rather scrappy material principally from the lower Pleistocene of Ternifine, Algeria. It is a large camel, held to have some similarities to the Bactrian camel. Gautier (1966, table 1) showed that the proportions of a Ternifine metatarsal agreed with the Bactrian camel but acknowledged that the specimen was not complete. C. thomasi is known until the end of the Pleistocene in North Africa (Arambourg 1962), and Gautier (1966) has identified as this species a camel skeleton from deposits in the northern Sudan dated to 22,000 years B.P. This skeleton is large for a camel and the proportions of its calcaneum agree with that of the Bactrian camel. Gautier also referred to a fragmentary large camel metapodial found by Coppens at Bochianga, near Koro Toro, in Chad. It seems safe to admit that the fossil camel of Africa was large, but other evidence for resemblance to the Bactrian camel does not appear to be conclusive. It amounts to the proportions of a calcaneum and the characters of the lower molars discovered by Grattard, Howell, and Coppens (1976). However, the Bactrian camel is the only surviving wild species, so that the resemblances may testify less to a C. thomasi-bactrianus relationship than to the distinctiveness of the characters of domestication in the dromedary. Further, the restricted distributional area of C. bactrianus as a relict species weakens its value

Tragulidae

and

Camelidae

for zoogeographical analysis. Finally, samples of camels in any one museum collection are small and sometimes the identifications and provenances are suspect. We have to be sure that even combined samples from several museums will be adequate for the differentiation of wild Bactrian camels, domestic Bactrian camels, and dromedary. We will probably have to turn to fossil evidence to answer the vital question of whether there was a separate wild species from which the dromedary is descended. When better African fossils are available, it will be profitable to compare them with the Pliocene C. sivalensis Falconer and Cautley from the Pinjor Formation of the Siwaliks. Complete metapodials of this species are large but appear to have proportions closer to the dromedary than to the Bactrian camel (Colbert 1935, figs. 133 and 134). Fossil camels from elsewhere in the Old World (see Howell, Fichter, and Wolff 1969 for references) have less reduced premolar rows than do living camels, and it is likely that any Pliocene or Pleistocene African camel will be similar. References Arambourg, C. 1933. Mammiferes miocenes du Turkana (Afrique Orientale). Ann. Paleont. 22:121-148. 1962. Les faunes mammalogiques du Pleistocene circummediterraneen. Quaternaria 6:97-109. Arambourg, C., and J. Piveteau. 1929. Les vertebres du Pontien de Salonique. Ann. Paliont. 18:59-138. Bishop, W. W., and Μ. H. L. Pickford. 1975. Geology, fauna and palaeoenvironments of the Ngorora Formation, Kenya Rift Valley. Nature 254:185-192. Colbert, Ε. H. 1935. Siwalik mammals in the American Museum of Natural History. Trans. Am. Phil. Soc. n.s. 26:1-401. Coppens, Y. 1971. Les vertebres Villafranchiens de Tunisie: gisements nouveaux, signification. C. R. Hebd. Seanc. Acad. Sei., Ser. D 273:51-54.

Flerov, C. C. 1971. Evolution of certain mammals in the Cenozoic. In Κ. T. Turekian (ed.), The late Cenozoic glacial ages. New Haven: Yale University Press, pp. 479491. Gautier, A. 1966. Camelus thomasi from the northern Sudan and its bearing on the relationship C. thomasiC. bactrianus.

J. Paleont.

40:1368-1372.

Gentry, A. W. 1970. The Bovidae (Mammalia) of the Fort Ternan fossil fauna. In L. S. B. Leakey and R. J. G. Savage, eds., Fossil Vertebrates of Africa. London: Academic

Press, vol. 2, pp. 243-323. Gentry, A. W., and A. Gentry. 1969. Fossil camels in Kenya and Tanzania. Nature 222:898. Grattard, J. L., F. C. Howell, and Y. Coppens. 1976. Remains of Camelus from the Shungura Formation, lower Omo Valley. In Y. Coppens, F. C. Howell, G. LI. Isaac, and R. E. F. Leakey, eds., Earliest

man and

environ-

Gentry

merits in the Lake Rudolf Basin. Chicago: Univ. Chicago Press, pp. 268-274. Hamilton, W. R. 1973. The lower Miocene ruminants of Gebel Zelten, Libya. Bull. Brit. Mus. Nat. Hist. (Geol.) 21 (3):73-150. Hopwood, A. T. 1929. New and little-known mammals from the Miocene of Africa. Am. Mus. Novit., no. 344: 1-9. Howell, F. C., L. S. Fichter, and R. Wolff. 1969. Fossil camels in the Omo Beds, southern Ethiopia. Nature 223:150-152. Madden, C. T. 1972. Miocene mammals, stratigraphy and

Tragulidae and Camelidae

539

environment of Muruarot Hill, Kenya. Paleobios 14: 1-12.

Pomel, A. 1893. Cameliens et cervides. Carte Geol. Alger. Paleont. Monogr. pp. 1-52. Van Couvering, J. Α., and J. A. Miller. 1969. Miocene stratigraphy and age determinations, Rusinga Island, Kenya. Nature 221:628-632. Walker, A. C. 1969. Lower Miocene fossils from Mount Elgon, Uganda. Nature 223:591-596. Whitworth, T. 1958. Miocene Ruminants of East Africa. Brit. Mus. Nat. Hist., Fossil Mammals of Africa, no. 15, pp. 1-50.

Many bovids are found in open country and even in arid regions, and it is this adaptive facility that undoubtedly explains the great success of the family. Linked with life in open areas is the tendency to live in herds. The only continents without native bovids are South America, Australia, and Antarctica. Bovids have horns consisting of a hollow keratinized sheath fitting over a bony core that may or may not have internal sinuses; neither sheath nor core is branched or seasonally shed. Horns may be present in both sexes or only in males. The horns of different species are distinctive and permit easy identification; in fact with fossil cores it is often more difficult to group than to separate the specimens. There are no upper incisors or canines, the upper and lower first premolars are missing, and the cheek teeth are selenodont (with crescentic cusps) and frequently hypsodont (high-crowned). Originally in artiodactyls the upper and lower incisors were used together for biting off foliage, the premolars for cutting it into smaller pieces, and the molars for crushing and grinding the small pieces. In bovids the lower incisors and tongue seize and tear off the food in different ways, and the cutting function of the premolars has declined. The premolars come to occupy a shorter length of mandible to allow more space for the relatively larger molars. However the premolars are not so reduced as in camels, nor do they become so molarized as in perissodactyls. The third and fourth metapodials have fused to form cannon bones in the front and back legs, and the lateral and medial metapodials have become more reduced than in deer. Cursorial adaptations of the limb bones, a feature of all artiodactyls, are most pronounced in some bovids.

Bovid Classification A modified version of Simpson's (1945) classification will be followed, with some improvements, mainly from Ansell (1971), and some of my innovations. Family Bovidae Subfamily Bovinae Tribe Tragelaphini Tribe Boselaphini Tribe Bovini Subfamily Cephalophinae Tribe Cephalophini Subfamily Hippotraginae Tribe Reduncini Tribe Hippotragini

bushbuck and allies Indian nilgai and four-horned antelope cattle and buffaloes duikers reedbuck group roan and allies

Gentry

Subfamily Alcelaphinae Tribe Alcelaphini Subfamily Antilopinae Tribe Neotragini Tribe Antilopini Subfamily Caprinae Tribe "Rupicaprini"

Tribe Ovibovini Tribe Caprini

wildebeest and group, impala

hartebeest

dik-dik group [incl. Saigini] gazelles, Springbok and others goral, serow group. Rupicapra itself might be better placed in the Caprini. muskox, takin, extinct allies sheep and goats

In the account of each tribe the fossils will be considered in reverse time order. A high proportion of the known bovid species are alive today and they give a lot of information about way of life that would not be available from fossils alone. It is therefore appropriate to use the living forms as a starting point for the fossil history of each tribe.

Fossil Localities The Pleistocene will here be divided into lower, middle, and upper parts as suggested in Butzer and Isaac (1975) and taken as having started around 1.8 m.y. (Berggren and van Couvering 1974). As a consequence the expression "lower Pleistocene" now denotes a period of time from 1.8 to 0.7 m.y., whereas in many recent studies of fossil mammals it has denoted the Villafranchian and equivalent periods from about 3.5 to 1.8 m.y. African sites of middle and upper Pleistocene age are mostly represented in Morocco, Algeria, and Tunisia in the north, and in South Africa. Important South African sites with bovids are Florisbad and Cornelia in the Orange Free State and Elandsfontein, Melkbos, and Swartklip on the coast of Cape Province. The later assemblages from the Sterkfontein, Swartkrans, and Kromdraai cave sites in the Aboukir Abu Hugar Afar Αϊη Boucherit Αϊη Brimba Αϊη Hanech Αϊη Jourdel, near Constantine Bel Hacel Beni Mellal Bled Douarah Bloembos Bolt's Farm Border Cave Broken Hill (now Kabwe)

Bovidae

541

Transvaal are also of middle and upper Pleistocene age (Vrba 1975). Faunal spans have been set up for South African sites (see chapter 1), to which the mammal ages of Hendey (1974) are equivalent. The latest of these, the Florisbad faunal span, or Florisian mammal age, is likely to run from 125,000 years B.P. or more until the Pleistocene-Holocene boundary, and is therefore coincidental with the upper Pleistocene. The preceding Cornelia faunal span dates from 700,000 years B.P. or earlier and is therefore nearly entirely of middle Pleistocene age, but it may have begun late in the lower Pleistocene. Some major East African sites are of Pliocene to Pleistocene age going back to about 3.0 m.y. They embrace the Shungura Formation in Ethiopia, East Turkana in Kenya, and Olduvai Gorge Beds I-IV in Tanzania. Some deposits in this group have slight temporal overlap at their upper levels with those of the South African Cornelian faunal span. South Africa also has sites that by faunal correlation are thought to be of Pliocene or lower Pleistocene age, for example the earlier assemblages from Sterkfontein, Swartkrans, and Kromdraai, as well as the other important site at Makapansgat Limeworks. There is considerable North African material of Pliocene to lower Pleistocene age, but the definitive account of the bovids has yet to be published as a posthumous continuation of Arambourg (1970). Among the sites older than 3.0 m.y. are Langebaanweg on the coast of Cape Province, Laetolil in Tanzania, Afar in Ethiopia, and most of the localities of the Baringo sequence in Kenya. A well-preserved early bovid fauna is known from Fort Ternan, Kenya, and has been dated to 14 m.y. (Bishop, Miller, and Fitch 1969, p. 685), but it comprises only four species. Nonbovid ruminants were still present in some numbers at this period, and only a few bovids are known from sites older than Fort Ternan. The following localities are mentioned in this chapter:

Algeria Sudan (see Hadar Formation) Algeria Tunisia Algeria Algeria Algeria Morocco Tunisia South Africa: CP South Africa: TR South Africa: Natal Zambia

(?middle) Pleistocene upper Pleistocene Pliocene (LV) Pliocene (LV) Pliocene (UV) Pliocene (LV) Pliocene (UV) middle Miocene middle Miocene (?upper) Pleistocene lower Pleistocene upper Pleistocene middle or upper Pleistocene

542

Gentry

Bovidae

Buffalo Cave Chelmer Cornelia Djerid East Turkana (= East Rudolf) Elandsfontein (= Hopefield)

South Africa: TR Rhodesia South Africa: OFS Tunisia Kenya South Africa: CP

Florisbad Fort Ternan Garaet Ichkeul Gebel Zelten Hadar Formation, Afar Isimila Kaiso Formation Kan am Kanjera Karmosit Beds, Baringo Kibish Formation, Omo Kom Ombo Koobi Fora Formation, East Turkana Kranskraal Kromdraai faunal site (KA) Laetolil Langebaanweg Losodok (= Moruorot) Lukeino Formation, Baringo Maboko Mahemspan Makapansgat Limeworks Mansoura, near Constantine Marceau Marsabit Road Melkbos Mockesdam Modder River Mpesida Beds, Baringo Mursi Formation, Omo Namib Desert Nelson Bay Cave Ngorora Formation, Baringo Olduvai Gorge, Beds I-IV Oued Bou Sellam, near Setif Oued el Atteuch Oued el Hammam (= Bou Hanifia) Peninj Power's site, Vaal River younger gravels Rusinga Island Sahabi Shungura Formation, Omo Sidi Bou Kouffa Singa Songhor Sterkfontein type site (STS) Sterkfontein "upper quarry" Steynspruit Swartklip Swartkrans main assemblage (SKa) Taung

South Africa: OFS Kenya Tunisia Libya Ethiopia Tanzania Uganda Kenya Kenya Kenya Ethiopia Egypt Kenya South Africa: OFS South Africa: TR Tanzania South Africa: CP Kenya Kenya Kenya South Africa: OFS South Africa: TR Algeria Algeria Kenya South Africa: CP South Africa: OFS South Africa: OFS Kenya Ethiopia Namibia South Africa: CP Kenya Tanzania Algeria Algeria Algeria Tanzania

Pliocene or Pleistocene upper Pleistocene middle Pleistocene middle Miocene Pliocene to Pleistocene middle Pleistocene, also a little upper Pleistocene material upper Pleistocene middle Miocene Pliocene (LV) early Miocene Pliocene lower Pleistocene Pliocene, with two faunal levels Pliocene to Pleistocene lower Pleistocene Pliocene upper Pleistocene to Recent upper Pleistocene Pliocene to lower Pleistocene upper Pleistocene lower (?middle) Pleistocene Pliocene Pliocene early Miocene late Miocene (?middle) Miocene upper Pleistocene to Recent Pliocene or lower Pleistocene Pliocene (UV) late Miocene Pliocene or lower Pleistocene upper Pleistocene upper Pleistocene middle or upper Pleistocene late Miocene Pliocene early Miocene upper Pleistocene to Recent middle to late Miocene lower Pleistocene middle or upper Pleistocene Pliocene middle Miocene lower Pleistocene

South Africa: CP Kenya Libya Ethiopia Tunisia Sudan Kenya South Africa: TR South Africa: TR South Africa: OFS South Africa: CP South Africa: TR South Africa: CP

middle Pleistocene early Miocene late Miocene Pliocene to lower Pleistocene Pliocene (UV) upper Pleistocene early Miocene Pliocene or lower Pleistocene upper Pleistocene or Recent upper Pleistocene upper Pleistocene lower Pleistocene Pliocene or lower Pleistocene

Gentry

Temara Ternifine Vlakkraal Wadi Natrun Wonderwerk Cave

Morocco Algeria South Africa: OFS Egypt South Africa: CP

The abbreviations used in the list are CP = Cape Province, LV = lower "Villafranchian," OFS = Orange Free State, UV = upper "Villafranchian," TR = Transvaal, and KA, SKa, STS = site designations used in Vrba (1975).

Systematics Subfamily Bovinae Tribe Tragelaphini The Tragelaphini are medium to large-sized browsing antelopes almost confined to Africa and tending to live where there is cover or even in forests. Living ones have spiraled horn cores in which the torsion is anticlockwise on the right side and in which there are two or three keels (anterior, posterolateral, and posteromedial). Other major characters are absence of internal sinuses in the frontals or horn pedicels, braincase not very angled on the face axis, teeth rather brachydont, basal pillars on molars small or absent, central cavities without a complicated outline, upper molars without prominent lateral ribs between the styles, lower molars with narrowly pointed lateral lobes and without goat folds, long premolar rows and large anterior premolars, P4's often with paraconid-metaconid fusion that closes the anterior part of the medial wall, and mandibles with shallow horizontal rami (see figure 27.1). There are two living genera, Tragelaphus (including Strepsiceros and Boocercus) and Taurotragus. The living species are Tragelaphus scriptus, the widespread bushbuck of small to medium size with anteroposteriorly compressed horn cores in which the posterolateral keel is the most prominent; T. spekei, the sitatunga that is the most aquatic of antelopes, larger than the bushbuck but with similar horns; T. angasi, the nyala with a restricted range in southeastern Africa and very similar cranially to the sitatunga; T. imberbis, the lesser kudu of parts of East Africa and perhaps Arabia (Harrison 1972, p. 629), with horn cores unlike the sitatunga by being longer, more spiraled, less divergent, and with a less strong posterolateral keel; T. strepsiceros, the greater kudu with strongly spiraled horn cores with the anterior keel more pronounced than the others and no anteroposterior compression; T. buxtoni, the

543

Bovidae

upper Pleistocene lower Pleistocene upper Pleistocene late Miocene upper Pleistocene

mountain nyala of Ethiopia with some features of its horn cores like the greater kudu; T. eurycerus, the bongo in which both sexes have horns; Taurotragus oryx, the eland, the largest known antelope, with horns in both sexes, a strong anterior keel, without anteroposterior compression, quite tightly twisted but not openly spiraled. Of these species, probable fossil ancestors are known for all except Tragelaphus buxtoni and T. eurycerus. In the middle and upper Pleistocene of South Africa there is an eland that can be safely referred to the living species. In the abundant material from Elandsfontein the insertion of the horn cores may not be at quite such a low angle as in the extant form of the species. An eland, apparently the living species, is known from the later Pleistocene of North Africa but disappeared at the start of the Holocene (Arambourg 1962). Taurotragus arkelli L. S. B. Leakey (1965) is known as a single cranium from Olduvai Bed IV, Tanzania. It is less advanced than the living eland and has horn cores slightly more uprightly inserted and a braincase less drastically shortened. It must almost certainly predate the Elandsfontein T. oryx. Arambourg (1962, p. 106) refers to an eland at Ternifine, Algeria, but details have yet to be published. Remains of eland from other sites are only fragmentary. Another tragelaphine in the later Pleistocene of the Cape Province is an enigmatic kudu known only from horn cores and teeth (Hendey 1968, p. 108). The horn cores agree quite well with the greater kudu but are much more tightly spiraled, and the teeth are of a size between greater and lesser kudu. It may be a subspecies of Tragelaphus strepsiceros or a separate species. Tragelaphus strepsiceros is the most frequently found tragelaphine at Olduvai but is sufficiently different from living examples for subspecific names to be applicable. T. s. grandis L. S. B. Leakey (1965, p. 38) is the large form from middle and upper Bed II, represented by the holotype cranium, horn cores, and dental remains. The horn cores are more divergent at their bases and have less mediolateral compression than in living kudus. There is a posterolateral keel for a short distance at the base of some horn cores. Linear measurements of the holotype indicate a size about 10 to 20% greater than at the present day. Appropriately sized teeth for this form

544

Gentry

Bovidae

Tragelaphini

Cephalophini

Bovini

Hippotraqini

Reduncini

Alcelaphini

Antilopini

Μ'

Μ.

Figure 27.1 Occlusal views of teeth of the right side in African tribes of Bovidae. The bovine P4 is shown with the close approach of paraconid and metaconid seen in African bovines, but not in those of other regions. The molars of Aepyceros are very like those of Antilopini, but there is fusion of paraconid and metaconid on P4. Neotragine teeth differ from genus to genus and have not been illustrated. Apart from overall size, degree of hypsodonty, and relative lengths of premolar and molar rows, the following characters help in identifying bovid teeth: (a) development of ribs on lateral side of upper molars and medial side of lowers, (ό) a complicated or simple outline of central cavities, (c) medial lobes of upper molars constricted or not, (d) presence or absence of basal pillars, (e) a transverse "goat fold" at the front of the lower molars, i f ) projecting hypoconid on P4, (g) size and shape of metaconid on P4, and (h) paraconid-metaconid fusion on P4. occur also at Peninj, Tanzania, and Makapansgat Limeworks, South Africa, but subspecific assignment for teeth is perhaps unwise. Bed I and the lower part of Bed II contain T. s. maryanus L. S. B. Leakey (1965, p. 40), about the same size or slightly smaller than present day

kudus. It has strong mediolateral compression of its horn cores and a braincase roof more bent down on the face axis than in later kudus. It is curious that out of seven known P 4 specimens of this subspecies paraconid-metaconid fusion occurs in five (including two in late wear) and is just starting on a further

Gentry

two in middle wear; the same fusion was seen in only 13 of 24 examples of the living species. The only complete lower cheek toothrow of T. s. maryanus has a premolar row shorter than in most living greater kudus. T. strepsiceros is well represented in the Koobi Fora Formation of East Turkana, both below and above the KBS Tuff (Harris 1976, p. 295). Only a single horn core fragment from Member G represents T. strepsiceros in the Shungura Formation, Ethiopia. The common kudu in Members Ε to G of this formation is a smaller form. Compared with later kudus it has horn cores spiraled more closely to their longitudinal axis but not tightly spiraled. The horn cores are less mediolaterally compressed and inserted at a lower angle. The braincase is relatively narrower, presumably an allometric effect of smaller overall size. This kudu is unlike the lesser kudu in its strong anterior keel and in its tendencies for its horn cores to become more mediolaterally compressed from Member Ε up to Member G and to reduce the posterolateral keel. Nonetheless it could be ancestral to T. imberbis if it showed a degree of evolutionary reversal in response to competition from the closely related T. strepsiceros. The holotype horn core of T. gaudryi (Thomas 1884, p. 15) came from A'in Jourdel, and a frontlet came from Mansoura (Gervais 1867-69, pi. 19, fig. 4), both sites being of Villafranchian-equivalent age. T. gaudryi can be used provisionally to include the Omo kudu, but this may need changing if the Algerian kudu should turn out to be an early stage of the T. strepsiceros lineage. So far kudus have not been reported from later deposits in North Africa. Some horn core pieces from the Mursi Formation, Ethiopia, appear to belong to a kudu. Horn cores from various sites have been referred to Tragelaphus cf spekei or T. cf angasi, but the remains are too scrappy to give much information about their evolution. At one time there was an unfigured skull from Olduvai described as T. spekei stromeri Schwarz (1932, 1937), but this was destroyed during the Second World War. The poorly preserved Laetolil cranium and horn cores of Τ. cf buxtoni Dietrich (1942, p. 118) is of this stock and is not related to the mountain nyala, which has horn cores like those of greater kudu in their open spiraling, upright insertions, and slight anteroposterior compression, but has its main keel posterolateral instead of anterior. It does not belong to the early fauna of the Laetolil Beds (Leakey et al. 1976). Pieces of horn cores of T. cf spekei or Τ. cf angasi are known from the early fauna of the Kaiso Formation

Bovidae

545

near North Nyabrogo in Uganda (Cooke and Coryndon 1970, p. 200), Kanam East, Kanam East Hot Springs, and Kanjera in Kenya, and Makapansgat Lime works. The Makapansgat Lime works specimen shows less anteroposterior compression than in living sitatunga and nyala. There are also some appropriately sized teeth at Makapansgat Lime works, already referred to T. cf angasi by Wells and Cooke (1956, p. 10). The only definite evidence for the ancestry of T. scriptus is a pair of horn cores from Member C of the Shungura Formation, Omo, and some tooth remains from Makapansgat Limeworks. The horn cores differ from those of the living species by being less anteroposteriorly compressed and more uprightly inserted. However, they agree with the living species in the strong posterolateral keel and the slightly less strong anterior keel. The dentitions at Makapansgat Limeworks were referred by Wells and Cooke (1956, p. 12, fig. 5) to Cephalophus pricei but are clearly tragelaphine by their less massive premolars, paraconid-metaconid fusion on P4, P2 insufficiently large relative to P3 and P4, and less rounded outbowings on the medial walls of the lower molars. The holotype for the name C. pricei was one of the mandibles, but the paratype was a horn core of a fossil species of Raphicerus. It remains to be seen if and when horn cores conspecific with the mandibles are discovered at Makapansgat Limeworks whether the name pricei can be applied to the Shungura species. Tragelaphus nakuae Arambourg (1941, 1947, p. 418) comes from Members Β to Η of the Shungura Formation and also from the Koobi Fora Formation as high as "the lower part of the M. andrewsi zone" (Harris 1976, p. 295), which would be just above the KBS Tuff. It is not a kudu and differs from the contemporaneous T. gaudryi by its larger size, a transverse ridge across the cranial roof above the occipital, and horn core characters. The material in Members C to G differs from that in Member Β by increased size and horn cores shorter and more curved with less twisting of the keels, a weaker anterior keel arising from a more anterolateral insertion, greater anteroposterior compression, and wider basal divergence. T. nakuae is very interesting in possessing skull characters in common with certain boselaphines and unlike other tragelaphines. The braincase is set high and not at all inclined to the axis of the face, there are strong temporal ridges on the cranial roof, the dorsal parts of the orbital rim project quite strongly, the supraorbital pits are not narrowed and drawn out anteroposteriorly, and one specimen shows a large preorbital fossa. There are

546

Gentry

strong similarities to Selenoportax vexillarius Pilgrim (1937) from the Nagri Formation of the Siwaliks, 1 and the question arises of whether T. nakuae could be better classified as a boselaphine. It seems best to leave it as a tragelaphine until more is known about tragelaphines and boselaphines in Africa prior to the Pliocene. S. vexillarius could as easily be related to Bovini as to Tragelaphini, and its similarities to T. nakuae lie more with the Members C to G representatives than with that from Member B. Moreover, the teeth of Τ. nakuae are adequately known and are satisfactorily tragelaphine in morphology. What appears to be an ancestral species to Tragelaphus nakuae has recently been discovered earlier in the Pliocene of the Afar area of Ethiopia (see figure 27.2). A number of tragelaphine horn cores at Langebaanweg, South Africa, have less anteroposterior compression and more upright insertions than in living T. angasi or T. spekei. Tragelaphine teeth are known from the Mpesida Beds and apparently from Ngorora, Kenya (KNMMP 071, 072; KNM-BN 1235), making them the oldest record of the tribe. One or more boselaphine 1 The holotype, a skull roof with horn cores, cannot be later than the Nagri Formation according to Pilgrim. I incorrectly wrote that it was from the Dhok Pathan Formation (Gentry 1970a, pp. 260, 315), but this formation contains only teeth attributed to Selenoportax.

Bovidae

lineages must have developed tragelaphine tooth characters quite early, and this will lead to difficulties in tribal classification. Tribe Boselaphini The Boselaphini are a mainly fossil group with only two living species, the large Boselaphus tragocamelus, the Indian nilgai, and Tetracerus quadricornis, the small Indian four-horned antelope. However, the tribe has occurred in the past in Africa. The main characters of the tribe are horn cores with keels but no transverse ridges, braincases little angled on the face axis, strong temporal ridges on the cranial roof, brachydont cheek teeth with enamel that is often rugose, lower molars generally without goat folds, and premolar rows long. The most interesting boselaphine in Africa is Mesembriportax acrae Gentry (1974) from Langebaanweg. It is large with short divergent horn cores, very mediolaterally compressed, with a prominent anterior keel in their lower part but with a circular reduced cross-section terminally. The horn sheaths were probably bifurcated, unlike in any living bovid. There is an advanced system of internal sinuses in the horn pedicels and frontals, which raise the frontals so much that the inclination of the dorsal part of the orbital rims becomes almost vertical. The skull is low and wide and the temporal ridges are strong. The premolar rows are long with large upper and

Figure 27.2 Tentative phylogeny for Tragelaphini. Living species are entered above the dashed line and age is shown on the left in millions of years. No fossils are known related to Tragelaphus buxtoni and T. eurycerus.

Gentry

lower P2s and no paraconid-metaconid fusion on P4, and the first incisors are not greatly enlarged. The species is not very close to other boselaphines but may be descended from Protragocerus labidotus. An M3 from Wadi Natrun, Egypt, identified by Andrews (1902, p. 437, pi. 21, figs. 7 and 8) as Hippotragus ?cordieri appears to be a large boselaphine or tragelaphine. It chances to resemble living Oryx except in being lower crowned, and the anterior goat fold is not usual in boselaphines or tragelaphines. Parabos cordieri is a French Pliocene bovine or boselaphine, but there is no reason to link the Wadi Natrun tooth with this species in particular. Protragocerus labidotus Gentry (1970a, p. 247) from the Fort Ternan Miocene is a small species and has mediolaterally compressed horn cores with the anterior keel extended downward as a ridge onto the pedicel, no internal hollowing of the frontals, upper canine alveoli less reduced than in later bovids, and enlarged first incisors in the mandible. The limb bones are not advanced in cursorial specialization. It is congeneric with a European and Siwaliks species and close to the origins of Tragoportax, Selenoportax, and Pachyportax. A slightly later ProSyncerus

547

Bovidae

tragocerus comes from the Ngorora Formation, Kenya (Bishop et al. 1971), and an earlier member of the genus has been recorded from Libya (see "The Earliest Bovids of Africa" below). A frontlet, apparently of the well-known later Miocene Eurasian boselaphine Miotragocerus, has been found at Sahabi, Libya, and is now in Rome. This genus was formerly called Tragocerus until Kretzoi discovered that the name was preoccupied by a beetle (see Gentry 1971, p. 234). Miotragocerus is not otherwise known from Africa (see figure 27.3). Roman's (1931, p. 34, pi. 4, figs. 1-4) record from the Djerid area of Tunisia is based on teeth that could easily belong to the caprine Pachytragus solignaci. The same applies to a horn core from Djerid that Boule (1910, p. 50) thought was similar to Miotragocerus or Hemitragus. The Miotragocerus from Marceau, Algeria (Arambourg 1959, pi. 17, figs. 5-7), is based on some cheek teeth that are not exclusively identifiable as Miotragocerus. Tribe Bovini The Bovini comprise large-sized descendants of boselaphines known from the later Tertiary onward.

caffer Pelorovis^antiquus

Syncerus sp (Olduvai)

Bos primigenius

P. oldowayensis Simatherium kohllarseni

Ugandax gautieri

Mesembriportax acrae

Leptobos syrticus

10 Protragocerus sp. CNgorora)

\ 15

/ TRAGELAPHIN1

Protragocerus labidotus

/ Walangania africanus

E o t r Q

^

s

\

S

P

(Zelten)

Protragocerus sp. (Zelten)

h 20

Figure 27.3 Tentative phylogeny for African Boselaphini and Bovini. The single living species is entered above the dashed line and age is shown on the left in millions of years.

548

Gentry

The living ones have low and wide skulls (allometrically linked with their large size), horn cores in both sexes emerging transversely from the skulls, frontals and horn cores with internal sinuses, a short braincase and triangular basioccipital, molars with basal pillars and complicated central cavities, upper molars with prominent outbowed ribs between the styles, and lower molars without large goat folds. There are three living genera, Bos for the cattle and bison of Eurasia with one species in North America, Bubalus for the Asiatic water buffaloes (one mainland species and two allopatric island forms), and Syncerus for the African buffalo, S. caffer. The latter is shortfaced, has short horn cores inserted just behind the orbits, and the paraconid of P4 has usually fused with the metaconid to close the anterior part of the medial wall. It is a complexly variable species (see Grubb 1972). Even in the more recent deposits there are no signs of buffaloes with the uparched frontals, large basal horn bosses, and downturned horns found in living S. caffer caffer south of Ethiopia. A nearly complete S. caffer skeleton from the Kibish Formation, Omo (Leakey 1969, p. 1,132), resembles the less advanced West African and Ethiopian savanna buffaloes yet it is as large as a large male S. c. caffer. Another large S. caffer is represented by a cranium, horn core pieces, and tooth remains at Melkbos, South Africa (Hendey 1968, p. 104). It has internally hollowed horn cores and surface rugosity of the frontals, but again the horn cores pass outward at their bases and do not turn immediately downward. A large bovine lower jaw from the Zululand coast, Bubalus andersoni Scott (1907), is likely to be yet another large S. caffer. Syncerus has not been definitely recorded north of the Sahara Desert. In Olduvai middle and upper Bed II are a number of crania and one or two other finds of a buffalo, presumably ancestral to Syncerus caffer, in which the horn cores are internally hollowed only near their bases, have a triangular cross-section formed by upper, lower, and front surfaces, P4 with paraconid and metaconid growing closer to one another but not fused, and the basioccipital perhaps not so narrowed anteriorly. The species has horn cores emerging transversely and without large basal bosses; in this it resembles the less advanced forms of S. caffer. Skull parts from Members Β to G of the Shungura Formation have short horn cores and show a substantially more primitive morphology than the living species by their less shortened braincases and strong temporal ridges. Ugandax gautieri Cooke and Coryndon (1970, p. 206) from deposits of unknown age in the Kaiso For-

Bovidae

mation was published as a hippotragine but appears to be an early member of the Syncerus lineage (see figure 27.3). It has a braincase even less shortened and less low and wide than in the Shungura specimens and differs from the living buffalo by its smaller, less dorsoventrally flattened horn cores that pass more upward and backward than outward. It is like Proamphibos Pilgrim from the Tatrot Formation of the Siwaliks, ancestor of Hemibos and Bubalus, but differs in its shorter horn cores with less marked keels, less triangular basioccipital, and upper molars with smaller basal pillars and less pronounced styles and ribs on the lateral walls. An interesting bovine, apparently a more advanced Ugandax than U. gautieri, has been discovered at Afar. In the middle and upper Pleistocene of northern, eastern, and southern Africa are remains of an extinct lineage of larger, long-horned buffaloes. With the possible exception of the abundant Elandsfontein remains, they may be regarded as belonging to one species, Pelorovis antiquus (Duvernoy 1851), which includes Βubalus baini Seeley and 5 . nilssoni Lönnberg. Besides the longer horn cores, they also differ from S. caffer by their horn cores being less dorsoventrally compressed and without basal bosses and by longer metapodials. The holotype came from Oued Bou Sellam near Setif, Algeria, and is of middle or upper Pleistocene age. Other North African specimens have been illustrated by Pomel (1893). A horn core, WK-East 2305, from Bed IV and possibly some teeth from Beds III and IV provide evidence of P. antiquus at Olduvai. Homoioceras singae Bate (1949,1951) from Singa and Abu Hugar, Sudan, was based on a skull with horn bases thought to be of one of these long-horned buffaloes. The new generic name was intended to apply to all of them and its founding emphasized that they were not related to the south Asian water buffaloes of the genus Bubalus. It is unfortunate that the holotype of H. singae, the type species, probably belongs to a large member of the Syncerus stock (an idea suggested to me by J. W. Simons) as shown by the great dorsoventral compression of the horn bases and the closeness of their dorsal edges on the top of the skull. However, the older generic name Pelorovis Reck (1928) is available, and Bate is correct about the dissimilarity to Bubalus. Pelorovis is more like Syncerus in its shorter face, irregular or absent keels on the horn cores, nasals without lateral flanges anteriorly, premaxillae without a long nasal contact, vomer not fused to the back of the palate, lower and wider occipital surface, and tendency to fuse paraconid and metaconid on P4. Pelorovis oldowayensis Reck 1928 (including Bu-

Gentry

Bovidae

549

larchus arok Hopwood according to Gentry 1967) is known from Olduvai middle and upper Bed II, Kanjera, and high in the Koobi Fora Formation, East Turkana (Harris 1976, p. 295). The holotype is said to come from Bed IV at Olduvai. It has horn cores inserted close together and very posteriorly near to or above the occipital surface. With the toothrow in a horizontal plane, the horn cores curve backward on leaving the skull; this makes their support difficult, especially in an animal that had attained the size of P. oldowayensis. Probably this species went on to give rise to P. antiquus with its generally downsweeping horn cores. The change in their course was linked with the insertions moving forward again to a more primitive position just behind the orbits. The occlusal pattern of the cheek teeth of P. oldowayensis is simpler than in P. antiquus, except for the Elandsfontein examples. There are few other differences, and we can regard P. antiquus and P. oldowayensis as congeneric. The priority of the generic name Pelorovis circumvents the problem of the type species of Homoioceras being probably a Syncerus, and Pelorovis need not be abandoned even if P. antiquus were not the descendant of P. oldowayensis. Teeth belonging to P. oldowayensis in Olduvai Bed II are well differentiated from those of Syncerus by their greater size and less pronounced occlusal complexity. This distinction between the two lineages is less apparent in teeth from the Shungura Formation, where a few horn core fragments show the presence P. oldowayensis or its ancestor.

Smith (1972) and Churcher (1972) confirm B. primigenius at this site. The species is known in North Africa back to early in the Amirian (= middle Pleistocene), but it is not found at Ternifine (Arambourg 1962, p. 106). In earlier Villafranchian-equivalent horizons is a large bovine related to Leptobos Rütimeyer (Arambourg 1962, p. 104); the Bos sp. teeth from Sidi Bou Kouffa, Tunisia, mentioned by Coppens (1971) could be the same species. Petrocchi (1956) described Leptobos syrticus from Sahabi, Libya. Several characters of the three complete or partial crania of this species suggest that it is rather primitive compared with other Leptobos: the closeness of the horn core insertions to the orbits in the anteroposterior plane, the absence of any lengthening of the horn pedicels, the retention of an anterior keel and a posterolateral keel in parts of the horn core's course, and the closeness of the supraorbital pits to the longitudinal midline of the skull. These are all resemblances to Parabos boodon (Gervais) from Perpignan, France, particularly to the cranium in Paris figured by Deperet (1890, pi. 7, fig. 4), but L. syrticus is a separate lineage from Parabos. L. syrticus is unlikely to be related to Syncerus because of the smoothness of the frontal bone surfaces between the horn core bases, insufficient inclination of the braincase roof, the excessively developed temporal ridges, and perhaps the weakness of the posteromedial (= posterodorsal) keel. Pilgrim (1937, p. 817) has pointed out the lingering of faint keels in a specimen of L. falconeri.

Simatherium kohllarseni Dietrich (1941, p. 222, 1942, p. 119) is based on a badly preserved Laetolil cranium that could be ancestral to Pelorovis. There are three differences from P. oldowayensis: the horn cores are inserted closer to the orbits, are wider apart, and diverge less toward their tips. The first and third are to be expected in an ancestor, and the second is not incompatible with that form being ancestral. Similar horn core fragments are known from Langebaanweg and were mistakenly attributed by Gentry (in Hendey 1970, p. 114) to a kudu. In fact, the keel is much too prominent for a kudu but is not unlike the condition on the Shungura Formation horn core fragments or even the irregular keels or ridges that occur occasionally on the Elandsfontein Pelorovis. Bos primigenius is known from the late Pleistocene of North Africa, where it survived into Neolithic and perhaps Roman times. Pomel (1894a) illustrated some remains. Bubalus vignardi Gaillard (1934, pi. 5, fig. 1) from Kom Ombo, upper Egypt, also looks like a Bos primigenius, and Churcher and

Subfamily Cephalophinae Tribe Cephalophini Cephalophines or duikers are mostly forest-living small to moderate-sized antelopes, rather heavily built and with short horn cores. All but one of the species are placed in the genus Cephalophus. The exception is Sylvicapra grimmia, which has more upright horn cores and longer legs and lives in areas with less dense cover than those inhabited by Cephalophus. Speciation has been extensive in cephalophines, but they are extremely uncommon as fossils. They are seen in some late archaeological contexts, e.g., at Nelson Bay, Cape Province (Klein 1972). Cephalophus parvus Broom (1934, p. 477, fig. 7) was described on a maxilla with P3 and P 4 from Taung, South Africa. Wells (1967, p. 101) believed it to be conspecific with the living blue duiker, C. "caerulus" (= C. caeruleus, a junior synonym of C. monticola). From Makapansgat Limeworks there is a piece of a right mandible, BPI M-22, with part of

550

Gentry

P3 to M2 described by Wells and Cooke (1956, p. 15) as cf Cephalophus caeruleus. Cooke (1949, p. 38) referred to a mandible and a broken lower M3 of Sylvicapra grimmia from the Vaal River younger gravels at Power's site, but Vrba (1973, p. 288) was satisfied that the mandible belongs to Antidorcas bondi. Schwarz (1937, p. 25) referred a mandible and vertebrae from Olduvai to Philantomba (= Cephalophus) monticola, but without illustrating them. The material was destroyed during the Second World War. A fragmentary maxilla from the Ngorora Formation, KNM-BN 96, may represent a duiker. The two surviving molars are brachydont with rugose enamel and have a pronounced rib between parastyle and mesostyle. The fossil is unlikely to belong to a neotragine or gazelle. The tooth of Cephalophus sp. Arambourg (1959, p. 127, pi. 17, figs. 8, 8a, and 8b) from Marceau is smaller than the contemporaneous gazelle species but otherwise indeterminate. Subfamily Hippotraginae Tribe Reduncini Reduncines are moderate to large-sized antelopes commonly found near water and now confined to Africa. Living ones have horn cores only in males, without keels or spiraling but with transverse ridges, and mostly inserted at a low angle in side view. They also have little internal hollowing within the frontals, temporal ridges on the braincase roof often approaching closely to one another, a large maxillary tuberosity prominent in ventral view, palatal ridges close together on the maxillae in front of the toothrow, cheek teeth rather small in relation to skull and mandible size, and quite hypsodont, basal pillars on upper and lower molars, medial lobes of upper molars and lateral lobes of lowers constricted, upper molars with small but wellprotruding ribs between styles, lower molars with goat folds, upper and lower P2s small, lower premolars with the appearance of being anteroposteriorly compressed, P4 with a strongly projecting hypoconid, and rarely with paraconid-metaconid fusion. There are two living genera, Redunca and Kobus. Redunca redunca the bohor reedbuck has short horns over most of its range, inserted at a very low angle; R. arundinum has longer horns, some other distinguishing skull characters, and a more southerly distribution; the small R. fulvorufula has short and little divergent horn cores. Species of Kobus are more gregarious. K. kob, the kob and puku, has mediolaterally compressed horn cores inserted rather uprightly in side view and close together; K. leche, the Central African lechwe, has mediolaterally com-

Bovidae

pressed horn cores inserted widely apart and at a low angle and with an initial backward curvature; K. ellipsipyrmnus,

the Waterhuck, is the largest liv-

ing reduncine and has horn cores with little mediolateral compression inserted widely apart and without any basal backward curvature; K. megaceros, the Nile lechwe, has horn cores with little mediolateral compression inserted farther behind the orbits than in other reduncines, wide nasals, and very strong longitudinal ridges on the basioccipital. Fossils are known that could be ancestral or related to two reedbucks and to all living species of Kobus except the Nile lechwe, and there are two or more lineages without living descendants. Redunca arundinum is common as a fossil at Elandsfontein and other sites (Hendey 1968, p. 110; Hendey and Hendey 1968, p. 51). The Elandsfontein form has horn cores more anteroposteriorly compressed, less divergent, and possibly shorter than in living representatives of the species. The last two characters, and possibly the first, give it some resemblance to R. redunca, suggesting a recent common ancestry. Redunca redunca is present in the late Pleistocene and early Holocene of North Africa, far to the north of its present limits (Arambourg 1938, p. 44). Antilope (Oegoceros) selenocera Pomel (1895, pi. 6, figs. 1-3), A. (Dorcas) triquetricornis Pomel (1895, p. 28, pi. 11, figs. 1 and 2) and the teeth called A. (Nagor) maupasii Pomel (1895, p. 38, pi. 10, figs. 1-11) are all synonyms of R. redunca. Redunca darti Wells and Cooke (1956, p. 17) is an earlier species from Makapansgat Limeworks. It differs from living R. redunca and R. arundinum, to which it may be ancestral, by its horn cores being more uprightly inserted and the posteromedial basal flattened surface lying more medial than posterior. It is larger than i?. fulvorufula and has larger, more divergent, and less upright horn cores. Two pieces, BPI M-690 and M-2798, suggest that it may have had a small preorbital fossa unlike any living Redunca. A horn core from Kanam East Hot Springs, BM M-15928, is similar to R. darti. A few tooth remains likely to be of Redunca are known from Olduvai. Horn cores from Kanam and a horn core base from Kanjera, BM M-26930, M-26931, M-26932, M-26934, resemble/?, redunca rather than the more primitive R. darti. Coppens (1971) refers to Redunca from Villafranchian-equivalent sites in Tunisia, of which further details are to be given by Arambourg (to be published posthumously). Other fossil reduncines form a more heterogeneous group and are related or assignable to Kobus. Potentially very interesting is a late Pleistocene

Gentry

horn core base from Abu Hugar, Sudan (Bate 1951, fig. 3), which Wells (1963) discussed and identified as "cf Kobus sp." The fossil looks reduncine, but I cannot identify it with any known extant species. It is somewhat like a very large kob but shows pronounced mediolateral compression. A second reduncine in the upper Pleistocene of South Africa is Kobus venterae Broom (1913, p. 13), of which the holotype cranium and some horn cores come from Florisbad and other horn cores from Mahemspan, Mockesdam, and Vlakkraal. It appears not to be distinguishable from the living K. leche, and it presumably indicates the former presence of seasonally inundated floodplains. Reduncines are common in the Shungura Formation but rather rare at Olduvai Gorge. Kobus sigmoidalis Arambourg (1941) has long, mediolaterally compressed, divergent horn cores with an initial curve backward at the base. Its teeth are not as advanced as in present-day reduncines in that upper molars have central cavities that are often not very complicated and ribs between the styles that are not very localized or accentuated. The lower molars frequently have less constricted lateral lobes than in the living waterbuck. Horn cores and teeth of K. sigmoidalis are scarce in Members C and D, abundant in Members Ε to G of the Shungura Formation, and occur in Olduvai Beds I and lower II. A few horn cores of the same lineage but with mediolaterally thicker bases and scarcely any backward curvature are known from Bed III and perhaps Bed I Olduvai and from the upper levels of the Shungura Formation. Thus K. sigmoidalis was probably evolving into K. ellipsiprymnus. Remains from the Koobi Fora Formation are comparable with those in Member G and above in the Shungura Formation (Harris 1976, pp. 296-297). K. leche may also descend fromK. sigmoidalis. In middle and later Bed II at Olduvai are found sparse remains ofK. kob. It was smaller than if. sigmoidalis and had horn cores inserted at a low angle with little backward basal curvature. Metapodials were a little shorter than at the present day. The kob from Bed III was larger and had larger, more mediolaterally compressed horn cores curving backward at the base from a slightly more upright insertion. Compared with the living kob, it is larger and the horn cores less upright at their bases and their insertions less close together. The evolutionary change in compression and insertion plane of the horn cores is the reverse of what occurred in the waterbuck lineage. Menelikia lyrocera Arambourg (1941, 1947, p. 392) is a medium-sized extinct reduncine from

Bovidae

551

Members Ε to J of the Shungura Formation, the Koobi Fora Formation (Harris 1976), and Marsabit Road, Kenya. Its most surprising character is the extensive internal hollowing of the frontals, almost as much as in alcelaphines. Its horn cores rise close together, then diverge outward, and finally curve upward at the tips. The braincase is a little angled on the facial axis, and a small preorbital fossa is present (absent in all living reduncines). The anterior tuberosities of the basioccipital are large and close together. Arambourg founded a new subfamily for this form, but it is satisfactory as a reduncine by its horn cores with transverse ridges, temporal ridges approaching closely on the cranial roof, prominent maxillary tuberosity in ventral view, the extreme closeness of the palatal ridges on the maxilla in front of the toothrow, large anterior tuberosities of the basioccipital, basal pillars on the molars, constricted medial lobes of the upper molars, lower molars with goat folds, lower premolars with the appearance of being anteroposteriorly compressed, P4 with a strongly projecting hypoconid and often a deep and narrow lateral valley in front of it, and no paraconid-metaconid fusion on the P4. An earlier species of Menelikia with longer horns less distinct from those of K. sigmoidalis occurs in Member C of the Shungura Formation. A short-horned variety of M. lyrocera is known from Members Η and J. The Kaiso Village locality of the Kaiso Formation has yielded horn cores of cf Menelikia lyrocera (Cooke and Coryndon 1970, p. 214). Also, a horn core collected by Bishop from Nyawiega lower in the same formation (referred by Cooke and Coryndon to Pultiphagonides cf africanus) appears to belong to Kobus sigmoidalis. Kobus ancystrocera (Arambourg 1947, p. 416) occurs in Members Β to J of the Shungura Formation, Omo, and above the KBS Tuff at East Turkana (Harris 1976, p. 296). It has long horn cores inserted at a very low angle and curving strongly upward and even slightly forward at the tips. They are very divergent and have some mediolateral compression with a flattened lateral surface. Arambourg placed the species in Redunca, but it seems rather remote from the neat group formed by the three living species and ft. darti. Its larger size and long horn cores with their low insertion angle, great divergence, curvature at the tips, and flattened lateral surface are all unlike R. darti. Thaleroceros radiciformis Reck (1935, 1937) is known only from the holotype frontlet with right horn core from Olduvai Bed IV. It is the only Olduvai antelope to have survived the Second World War in Munich and is perhaps the oddest antelope to

552

Gentry

have come from Olduvai. It is moderate to largesized with massive horn cores passing backward then upward from the base and parallel to one another. There is a sudden diminution of crosssectional area near the tips. The massive united horn core pedicel is without internal sinuses and has paired anterior protuberances just below the horn core bases. It is possible that this Bed IV frontlet represents a descendant of K. ancystrocera, specialized in its massive horn cores and united pedicel. A pair of unidentified horn cores, BM M-15925, from Kanam "Museum Cliff" are structurally intermediate. "T. radiciformis (? )" of Leakey (1965, p. 65) is an alcelaphine and is unrelated to the present species. Another reduncine, common in Member Β of the Shungura Formation, shows a short braincase hardly at all angled on the face axis, long horn cores with some anteroposterior compression, their insertions close together, and small supraorbital pits. This form has not been found elsewhere, but a related or ancestral species occurs at Afar. It is abundant in the middle part of the Hadar Formation. A short-horned Kobus, perhaps related to K. sigmoidalis, is known from Langebaanweg. A similar form, but with slightly longer horn cores, comes from Sahabi, and a cast of a right horn core from Wadi Natrun, BM M-8200 labeled "?Gazella," is also very similar. Reduncines with koblike characters are known from the Dhok Pathan, Tatrot, and Pinjor Formations of the Siwaliks. They are difficult to group into specieslike taxa, a problem that Pilgrim (1939, pp. 99-129) solved by giving almost every specimen a different name. The Pinjor forms seem well fitted to be ancestral to both kobs and K. sigmoidalis, but without knowing their stratigraphic levels of origin or having radiometric dates, one cannot be sure if they are ancestral or merely primitive Indian survivals already isolated from African forms. They have strong temporal ridges on the braincase roofs, which suggests the likelihood of a boselaphine origin, and early tooth remains are also not unlike Boselaphini. It is interesting that the male skulls BM M-487 and M-2402, but not the females M-36673 and M-39569, retain small preorbital fossae. The Tatrot and Dhok Pathan remains are of smaller animals and have been described under the names Dorcadoxa porrecticornis (Lydekker), ?Gazella superba Pilgrim, cf Indoredunca theobaldi Pilgrim and cf Hydaspicobus auritus Pilgrim. Horn cores of similar reduncines have been found at Baard's Quarry, Langebaanweg, and a similar frontlet, KNM-LU Oil, in the Lukeino Formation of the Lake Baringo sequence, Kenya. The horn cores are

Bovidae

somewhat mediolaterally compressed, inserted above the back of the orbits, fairly divergent, with a slight and regular backward curvature. They can be distinguished from gazelles only by their larger size, greater divergence, deep postcornual fossa, and larger supraorbital pit. These data suggest the near contemporaneous occurrence of similar primitive reduncine forms in Asia and Africa during the earliest Pliocene (see figure 27.4). Tribe Hippotragini The Hippotragini are a tribe of medium to large, rather stocky antelopes. The few species cover a wide range of habitats. Both sexes have long, not very divergent horn cores without keels or transverse ridges, hollowed pedicels to the horn cores, teeth quite hypsodont without much premolar reduction, molars with basal pillars, lower molars with a tendency to have anterior goat folds, and P4s without fusion of paraconid and metaconid. There are three living genera, Hippotragus, Oryx, and Addax. The species of Hippotragus comprise H. leucophaeus, the small South African blaauwbok exterminated in about 1799; H. equinus, the roan antelope with uprightly inserted and backwardly curved horn cores; H. niger, the sable antelope with longer, more mediolaterally compressed horn cores and higher frontals between the horn bases than in the roan; Oryx gazella, the gemsbok and beisa, with a lower and wider skull and straighter, more anteroposteriorly compressed horn cores inserted very obliquely; O. dammah, the scimitar oryx with more curved horn cores; O. leucoryx, the Arabian oryx; and Addax nasomaculatus, the addax of the Sahara with spiraled horns. Fossils are known related to living Oryx and Hippotragus, and there is also an extinct lineage of Hippotragus. In the later Pleistocene of South Africa three species of Hippotragus are known. A small one having teeth with an advanced occlusal pattern and long premolar rows occurs at Swartklip (Hendey and Hendey 1968, p. 54), Elandsfontein, and more recent sites, and Cooke (1947) referred similar mandibles from Bloembos to H. problematicus. All these remains are most likely to belong to the extinct blaauwbok. Mohr's (1967, p. 64) belief that if. problematicus had different teeth from H. leucophaeus arose from a comparison of the fossil with a skull in Glasgow, identified as H. leucophaeus. However, Mohr's illustrations suggest that the Glasgow skull could well be a sable, in which case the difference of problematicus from leucophaeus remains to be demonstrated. Hippotragoides broomi Cooke (1947) from the

Gentry

Bovidae

553

megaceros or Redunca fulvorufula.

Sterkfontein "upper quarry" (not the australopithecine site) is likely to be an example of the living roan antelope. In a most interesting paper, Klein (1974) has demonstrated both that H. equinus occurs in the late Pleistocene and Holocene of the southern Cape Province and that at a number of sites it existed contemporaneously with H. leucophaeus. A third species of Hippotragus is the extinct H. gigas L. S. B. Leakey (1965), which occurs abundantly at Elandsfontein and as a few teeth at Florisbad but appears not to have survived into the Holocene. It differs from sable or roan by less mediolaterally compressed horn cores, a larger oryxlike dentition with relatively shorter premolar rows, P2 as small as in Oryx, a simpler occlusal pattern of the molars, strong temporal ridges on the braincase roof, a low and wide occipital surface, and a short basioccipital with anterior tuberosities localized more as in Oryx than as in other Hippotragus. Further back in the Pleistocene there is no record of H. leucophaeus, but H. gigas is found in Beds I through III at Olduvai, its type locality. This material has some differences from the Elandsfontein remains, so that the latter should certainly have subspecific status. At Olduvai the temporal ridges are less strong, the occipital less low and wide, and the horn cores larger. The dimensions of the horn cores probably exceed all but those of the largest

sable. The holotype (Leakey 1965, p. 49, pis. 56 and 58) is from Bed II, probably above the Lemuta Member. The adult females and the young of both sexes probably had more inclined horn core insertions than the males, and females probably had less backward curvature. It is difficult to distinguish hippotragine teeth from those of bovines in Olduvai Beds I and II. They seem to be smaller and the premolar rows shorter; the lower molars may have less narrowed lateral lobes, less pronounced outbowings on the medial walls, and goat folds extending down to the neck of the teeth. When Bovini have goat folds, these are smaller and disappear before the tooth is completely worn down. The hippotragine lower P4 has a large bulbous metaconid (less apparent in living Hippotragus) but little tendency for paraconid and metaconid to fuse. In hippotragine upper molars the mediolateral width is smaller nearer the base and the central cavities are less curved. Horn cores of H. gigas are also known from East Turkana (Harris 1976), Kanjera, and Peninj and of H. gigas or an ancestral species from Member G of the Shungura Formation, Omo. It is interesting that some large Hippotragus upper molars from Kanjera, BM M-25711, M-25695, and M-25702 are more advanced thani/. gigas teeth from other sites. Perhaps H. equinus and H. niger descended from some population ofH. gigas, while other populations elsewhere

554

Gentry

evolved into the later subspecies of H. gigas as known from Elandsfontein. H. niger is more specialized than the other two species in its narrow braincase, compressed horn cores of the males, narrower dorsal orbital rims, and frontals raised between the horn core bases. Presumably it evolved more recently than H. equinus. Oryx sp. indet. is known from Olduvai Bed I by one or possibly two horn cores and from above the KBS Tuff at East Turkana by two partial crania (Harris 1976, p. 297). There is a frontlet from Member G of the Shungura Formation, but the upper molar that Arambourg (1947, p. 410) referred to Oryx is bovine. An oryx horn core was recorded from Mansoura near Constantine, Algeria (Joleaud 1918, p. 90, fig. 1), in deposits ofVillafranchian-equivalent age. Coppens has referred to hippotragines of the same age at Bel Hacel and Ain Hanech, Algeria, and at Garaet Ichkeul, Tunisia, but further details have yet to be published in the continuation of Arambourg (1970). Balout (1942) attributed a palate from Algiers to Addax nasomaculatus, but the identity cannot be taken as established. The horn core of Praedamalis deturi Dietrich (1950, pi. 2, fig. 23) from Laetolil is likely to be hippotragine. It is inserted rather uprightly and has almost no curvature, hence it is difficult to know whether to relate it to Oryx or Hippotragus. Teeth of Aeotragus garussi Dietrich (1950, pi. 3 figs. 37-40, 42) may be conspecific, although the horn core assigned to this name is indeterminate. A slender, backwardly curved horn core from the earlier North African site of Sahabi is possibly on the Hippotragus lineage. Less certainly hippotragine are the Laetolil horn cores Gazella kohllarseni and Aepycerotinae gen. et sp. indet. (Dietrich 1950, pi. 1, fig. 7; pi. 4, fig. 45). A Wadi Natrun tooth assigned by Andrews (1902) to Hippotragus ?cordieri has already been mentioned. Studer (1898) thought a Wadi Natrun horn core was of the oryx group, and Andrews tentatively associated it with the above tooth, but it is from an unidentified alcelaphine. 'Hippotragus sp." from Langebaanweg (Gentry in Hendey 1970, p. 115) now appears to be an alcelaphine. The holotype cranium of Sivatragus bohlini Pilgrim (1939) from the Pinjor Formation of the Siwaliks would be suitable as an ancestor of H. gigas. It is smaller but has a short braincase and a short, small basioccipital with localized anterior tuberosities. Its stronger temporal ridges and a braincase less inclined than in H. gigas are compatible with descent from the Boselaphini. S. brevicornis Pilgrim (1939) is a form of uncertain tribal affinity. Since S. bohlini

Bovidae

is the type species and appears to be congeneric with H. gigas, it should be possible to sink Sivatragus in Hippotragus (see figure 27.5). A fossil Oryx is also known from the Pinjor Formation, represented by the two conspecific skulls, Antilope sivalensis Lydekker (1878) and the immature Sivoryx cautleyi Pilgrim (1939). Oryx sivalensis, as it should probably be known, has a more angled braincase roof without temporal ridges and horn core insertions more inclined than in Hippotragus bohlini, but a large shallow preorbital fossa would fit boselaphine ancestry. Subfamily Alcelaphinae Tribe Alcelaphini The Alcelaphini are medium to large antelopes of open country and are now confined to Africa. Their main osteological features are a long skull, horns in both sexes, horn cores often of irregular course and often with transverse ridges, generally no keels, extensive internal sinuses of frontal and horn pedicels, short braincase strongly angled on the face axis, zygomatic bars usually deepened anteriorly under the orbits, a central longitudinal groove on the basioccipital, cheek teeth strongly hypsodont, without basal pillars, complicated central cavities, rounded medial lobes of upper molars, and lateral lobes of lowers, widely outbowed ribs of upper molars, lower molars without goat folds, short premolar rows with P2s reduced or absent, P4 with fused paraconid and metaconid, mandible with deep horizontal rami, limb bones with markedly cursorial adaptations. There are four living genera, Alcelaphus, Damaliscus, Connochaetes, and Beatragus, and I would include Aepyceros, which is normally put in the Antilopini. Alcelaphus buselaphus is the widespread hartebeest with a united horn pedicel and immense local variation in horn core morphology; A. lichtensteini is the Central African Lichtenstein's hartebeest with horn cores set widely apart; Damaliscus lunatus includes the topi, tiang, and tsessebe with horn cores inserted more forward, nearer to the orbits, and without the abrupt alterations in course seen in hartebeests; D. dorcas is the smaller South African bontebok and blesbok with skull morphology very similar to D. lunatus; Connochaetes gnou, the black wildebeest, has a lower and wider skull than any of the preceding species and horn cores with broad bases passing forward to end in sharply recurved tips; C. taurinus, the larger blue wildebeest, also has a wide skull but a longer face, and horn cores without such large bases and not passing forward; Beatragus hunteri, the herola or Tana

Gentry

I

Bovidae

\

Η. bohlini

555

0. s/vo/ensis

/

Figure 27.5 Tentative phylogeny for Hippotragini. Living or recently extinct species are entered above the dashed line and the age is shown on the left in millions of years. Non-African hippotragines are shown in italics. No fossils are known related to Addax nasomaculatus.

River hartebeest, with long horns diverging outward then having long parallel or subparallel distal parts. Aepyceros melampus, the impala, is a smaller antelope with horns somewhat like those of the preceding species but only found in males. Alcelaphine fossils assignable to all the living and several extinct lineages are abundant. Megalotragus priscus (Broom 1909) is the terminal species of an extinct lineage of veiy large alcelaphines. It is known from late Pleistocene sites of South Africa but not from northern Africa, and in my opinion it includes Pelorocerus helmei (Lyle), P. elegans Van Hoepen, Megalotragus eucornutus Van Hoepen, Lunatoceras mirum (Van Hoepen), and probably Pelorocerus broomi Cooke (1949). It had a relatively narrow skull for its great size, being larger than any living alcelaphine. It had molar teeth without a complicated occlusal pattern, very short premolar rows with reduction of anterior premolars, and long legs. The long curved horn cores are quite like those of the bovine Pelorovis oldowayensis, but do not reach the great size of the 01duvai Bed II representatives of that species and have a stronger upward component to their curvature. No complete skull has been found. The degree of basal divergence in the horn cores is weak in specimens from Cornelia, Elandsfontein, Steynspruit, and Mahemspan but strong in those from Florisbad, Mockesdam, and Kranskraal and in the holotype

cranial piece from the Modder River (all localities except Elandsfontein being in the Orange Free State). It remains to be seen whether this distribution will stand up to an interpretation of the two varieties as temporal subspecies or whether those horn cores with greater divergence, being also the longer ones, are from older individuals. In the earlier Shungura-Olduvai time span Megalotragus is represented by M. kattwinkeli (Schwarz 1932, p. 4), which includes as synonyms Alcelaphus howardi Leakey (1965) and Xenocephalus robustus Leakey (1965). It has short horn cores inserted not quite as far back as in M. priscus. M. kattwinkeli is known from Olduvai middle Bed II to Bed IV, and the same species or its ancestor is known from the Koobi Fora Formation (Harris 1976), earlier Olduvai levels, and the later levels of the Shungura Formation. Wildebeest survived in North Africa until the late Pleistocene (Ficheur and Brives 1900, reporting toothrows figured by Pomel 1894b, pi. 2, figs. 1-4), and perhaps into the Neolithic (e.g., Debruge 1906, an identification based on a metacarpal and teeth). Arambourg (1938, p. 42) is doubtful about Neolithic records. Wildebeests that can best be classified as belonging to extinct subspecies of Connochaetes gnou are known from later South African sites. C. gnou antiquus Broom (1913, p. 14) from Florisbad has horn cores passing less markedly forward from

556

Gentry

the base and tips perhaps less recurved. C. g. laticornutus (Van Hoepen 1932, p. 65) is an earlier stage from Cornelia and Elandsfontein, known from crania and horn cores but not from facial parts. The crania resemble C. taurinus at about the time level of Olduvai Beds II through IV, but the base of the horn core has begun to be expanded in dorsal view and a rugose surface texture has begun to spread across the frontals. These features are scarcely visible in the holotype cranium from Cornelia (Van Hoepen 1932, fig. 3), which I attribute to its being a female. The wildebeest from Olduvai middle Bed II to Bed IV is Connochaetes taurinus olduvaiensis, differing from the living subspecies by its horn cores being inserted at a slightly less posterior level and passing less downward as they emerge from the skull. A cranium from the upper Pleistocene of Temara in Morocco, assigned by Arambourg (1938, p. 38) to the Ternifine (lower Pleistocene) form C. taurinus prognu Pomel, is very similar to that from Olduvai. However, Pomel (1894b) did not illustrate any of the Ternifine (=Palikao) horn cores, so that more information is required before we can adequately assess the validity of these two subspecific names. Rhynotragus semiticus Reck (1935, p. 218, fig. 1) is an odd skull fragment from Olduvai that Schwarz (1937, pp. 60, 85) took as a subspecies of C. taurinus and to which he assigned much Olduvai material. The absence of a sharply outlined temporal fossa between horn core base and orbit and the sloping braincase roof of Reek's fossil make it unlikely to have been a wildebeest, but its true affinities remain uncertain. In the Koobi Fora Formation above the KBS Tuff (Harris 1976, p. 298) and the earlier sediments of Olduvai, up to and including the lowest levels of middle Bed II in the HWK area, the wildebeest present has horn cores inserted less far back on the skull and their distal parts not so markedly turned upward and inward. It is presumably a precursor of C. taurinus. A still earlier wildebeest is Oreonagor tournoueri (Thomas 1884, pi. 7, fig. 1) from Αϊη Jourdel, Algeria, in which the horn cores are less extremely divergent. A problematical specimen is the Olduvai Bed II holotype skull of Connochaetes africanus (Hopwood 1934, p. 549), originally the type species of the genus Pultiphagonides. The face has closer resemblances to C. gnou than to C. taurinus, and the rather small horn cores are inserted extremely widely apart—almost as if space were being left for descendants to develop the basal bosses of living C. gnou. Thus the C. gnou lineage formerly may have occurred north of South Africa, as did Antidorcas, and the identification of fossil wildebeest,

Bovidae

especially less complete specimens, may be more uncertain than has been implied here. Alcelaphus buselaphus is known only from very late sites in southern and northern Africa; like Syncerus caffer caffer, it can have appeared only in the very recent past. I discovered Alcelaphus lichtensteini horn cores among material from Broken Hill, Zambia, hitherto placed with Connochaetes; the horn cores differ from those of the extant members of the species by being less curved upward and passing more forward as they rise. Rabaticeras arambourgi Ennouchi (1953) is known from Morocco, Elandsfontein, and Olduvai Bed III. It has horn cores arising close together and above the orbits, inserted uprightly and curving forward with a torsion that is clockwise from the base upward on the right side, and with a short braincase much angled on the facial axis. R. porrocornutus Vrba (1971) from Swartkrans, South Africa, is similar. I believe that both the living species of hartebeest descended from Rabaticeras, in which case the united horn pedicel of A. buselaphus must have appeared very quickly in geological time. It is noteworthy that the characters whereby the Broken Hill fossils of A. lichtensteini differ from extant specimens make the fossils less removed from Rabaticeras. Also, a frontlet from Aboukir, Algeria, ofBoselaphusprobubalis Pomel (1894b, pi. 4, figs. 14 and 15) is certainly more primitive than the living hartebeest and may be a subspecies of R. arambourgi, although Arambourg (1938, p. 37) had made Pomel's name a synonym of A. buselaphus. An earlier Rabaticeras-like cranium from the Lemuta Member at the top of lower Bed II, Olduvai, has horn cores inserted more widely apart and at a lower angle and has a less inclined braincase. A possible interpretation of this fossil is that it represents the ancestry of Alcelaphus lichtensteini, already separated from that of A. buselaphus. If this were true, Alcelaphus would become a diphyletic genus and nomenclature would need to be changed. Horn cores ofDamaliscus dorcas, or a probable ancestor of that species, are known from Florisbad and Vlakkraal; they are larger and less mediolaterally compressed, with the lyration nearer to the base. A new and unnamed species of Damaliscus is represented in Olduvai Beds II through IV, and possibly I. It is about the size of D. dorcas but differs by some details of skull morphology and, surprisingly, by shorter premolar rows without P2. Except for its premolar characters, it seems suitable as an ancestor of both living Damaliscus species. Damaliscus niro (Hopwood 1936, p. 640) belongs

Gentry

to an extinct lineage known from Olduvai lower Bed II to Bed IV, Peninj, and later South African sites. It has horn cores curved backward, very compressed with flattened medial and lateral surfaces, and generally with well-spaced strong transverse ridges on the front surface. Some horn core variations are encountered at Olduvai in which there is not an even backward curvature but a sharper alteration in course near their midpoints. Some males reached a large size, but average size seems to have become smaller over the time span from Cornelia to Florisbad. At Florisbad and Vlakkraal its horn cores are smaller than those of the other Damaliscus, the medial surface less rounded, and the whole horn core more mediolaterally compressed, the widest part of the transverse section more anteriorly placed, and without the slight lyration. A horn core from Wonderwerk Cave, Cape Province ("cf Copra walie," Wells 1943, p. 268; also see Cooke 1941), still had part of the horn sheath attached. It is unfortunate that there is not a single toothrow definitely associated exclusively with D. niro horn cores. The species is not known in northern Africa. Damalops palaeindicus (Falconer) from the Pinjor

Connochaetes gnou C.taurinus Megalotragus priscus

M.

/

kattwinkeli

Bovidae

557

Formation of the Siwaliks seems most likely to be related to the Alcelaphus-Rabaticeras-Damaliscus group of alcelaphines. Parmularius Hopwood is an extinct stock of medium-sized alcelaphines with horn cores tending to have basal swellings, a tendency to small bosses on the parietal roofs, small preorbital fossae, short premolar rows, and deep mandibles. They are known principally from Olduvai. P. angusticornis (Schwarz 1937, p. 55), formerly placed in Damaliscus and including D. antiquus Leakey (1965), is a well-known species from middle and upper Bed II with heavy, straight horn cores and a much shortened braincase. It is also known from Kanjera and Isimila. It is probably descended from the slightly smaller type species P. altidens Hopwood (1934) of Olduvai Bed I and lower Bed II and the highest part of the Shungura Formation. This latter form has less massive and more backwardly turned horn cores. It is possibly a different lineage that produced P. rugosus Leakey (1965) of Beds III and IV and possibly earlier and that appears to have outlived P. angusticornis (see figure 27.6). A frontlet, SAM-16561, and some horn cores from

Beatragus hunteri

Damaliscus dorcas & lunatus

C.g. antiquus C.g.laticornutus

? Beatragus sp. (Elandsfontein) /

Alcelaphus lichtenstein A.buselaphus

Aepyceros melampus

Damaliscus niro Parmularius rugosus

C.t.olduvaiensis

Damaliscus sp. (Olduvai)

_ . . Rabaticeras arambourgi

P. angusticornis C.africanus

B. antiquus P. altidens

Damalops palaeindicus

Oreonagor tournoueri ?Parmularius sp. (LaetoliO

Aepyceros sp. (Shungura)

1959.233 (laetolil)

-4

Figure 27.6 Tentative phylogeny for Alcelaphini. Living species are entered above the dashed line and age is shown on the left in millions of years. The single non-African alcelaphine is shown in italics.

558

Gentry

Elandsfontein seem most likely to be a species of Beatragus with short and very divergent horn cores, however other interpretations are possible. Beatragus antiquus Leakey (1965) was evidently more widespread than the living B. hunteri, being known from Olduvai Beds I to II, and as an immature cranium from just below Tuff Η of the Shungura Formation. It differed from B. hunteri by its larger size, wider skull, and more uprightly inserted horn cores and may have been its ancestor. Aepyceros, apparently A. melampus, is known by a few small horn core pieces from Olduvai Gorge and Peninj. An impala is present in the Koobi Fora Formation (Harris 1976) and abundant in the Shungura Formation, both localities being north of the present-day range of Aepyceros. The material in the earlier part of these sequences is sufficiently distinct to justify specific separation from A. melampus. In Shungura Member Β it is smaller, has a less lengthened face, supraorbital pits placed less widely apart, the vestiges of a preorbital fossa, back of the nasals more narrowly pointed, and a longer premolar row. Horn cores of an even more primitive species occur in the Mursi and Hadar Formations, Ethiopia, and in the Karmosit Beds, Kenya. It is also possible that the supposed gazelle at Fort Ternan could be connected with Aepyceros instead (cf Gentry 1970a, p. 292). Alcelaphines other than Aepyceros are known from African sites predating the Shungura Formation. A Laetolil cranium, 1959.277, could be primitive in its rather long, little-angled braincase, and other characters. It has a low parietal boss and could be an ancestor of Parmularius altidens and Damaliscus niro; for the present it is best classified as ?Parmularius sp. The Laetolil horn core called "Reduncini gen. et sp. indet." by Dietrich (1950, pi. 2, fig. 21) is probably conspecific with the latter taxon. The alcelaphine species that Dietrich named Parestigorgon gadjingeri will remain of uncertain identity until better finds become available. An alcelaphine found at Afar may be conspecific with a partial cranium and horn cores, 1959.233, from Laetolil, and it may become possible to link this stock with Damalops mentioned above. Two species of primitive alcelaphines occur at Langebaanweg. Their frontals between the horn core bases are higher than the orbital rims, and the horn pedicels are internally hollowed as in later alcelaphines. One species has a braincase strongly angled on the facial axis, the distal parts of the horn cores tending to pass outward rather than backward and the long axis of the horn core cross-section set at a considerable angle to the long axis of the skull.

Bovidae

The other species has a less angled braincase, horn cores with a backward curvature, and normally orientated pedicels. The alcelaphine teeth at Langebaanweg are markedly primitive in the upper molars with a more pronounced mesostyle and less rounded medial lobes than in later forms, basal pillars on Ml and occasionally on the upper molars, less constriction across the central cavities, and less rounded medial and lateral lobes of the lower molars. They are more primitive than alcelaphine teeth at Laetolil. A cast of an M3 from Wadi Natrun, BM M-12361, which Andrews (1902, p. 439) thought was from a large gazellelike form, could perhaps represent an alcelaphine at the Langebaanweg level of development. It is less hypsodont than Pleistocene or living alcelaphines. A badly preserved horn core base, BM M-8199 (cast), labeled Hippotragus ?cordieri, also appears to be alcelaphine by the large, smooth-sided internal cavity in its horn pedicel. Subfamily Antilopinae Tribe Neotragini The Neotragini includes 14 living small-sized antelopes, possibly not very closely related to one another. They have simple, short, and small horn cores that are straight or curved slightly forward. They have large preorbital fossae. There are generally no basal pillars on the molars and no outwardly bowed ribs between the styles on the lateral walls of the upper molars; the enamel outer walls of the molars tend to be straight with pointed rather than rounded corners. The living genera are Neotragus (including Nesotragus) with three species, Madoqua (including Rhynchotragus) with five living species, Oreotragus, Dorcatragus, and Ourebia with one species each, and Raphicerus with three species. Fossils related to most of these genera are known but are not common. The Vaal rhebbok, Pelea capreolus, now recognized as not a reduncine (Roberts 1937, p. 86; Wells 1967, p. 100), might be related to Caprinae (Gentry 1970b, p. 65) or it might be a large neotragine. Fossil Pelea occurs in the Transvaal caves (Vrba 1975) as far back as the Kromdraai faunal site (KA) and the Swartkrans main assemblage (SKa), i.e., probably back to the lower Pleistocene. A Raphicerus at Melkbos and Swartklip is not distinguishable from living grysbok or steinbok, but the Elandsfontein Raphicerus was larger and had more inclined horn cores with a tendency toward posterolateral keels. The paratype horn core of "Cephalophus" pricei from Makapansgat Limeworks appears to be a Raphicerus with a keeled horn core

Gentry

like some of those from Elandsfontein. Some horn cores of Raphicerus at Langebaanweg are larger than at Elandsfontein and have a still more irregular cross-section. A few mandibular pieces at Laetolil are about the right size for Raphicerus. At Olduvai the only neotragine remains are two horn cores from FLKN in Bed I, perhaps of Raphicerus. Oreotragus major Wells (1951, p. 167) is based on a skull from a red breccia of unknown age near Makapansgat. It is probably a late form, judging by its closeness to the living klipspringer. The Makapansgat Limeworks frontlet assigned to Oreotragus major by Wells and Cooke (1956, p. 35) has very short horn cores with some degree of anteroposterior compression and is not very like the holotype; it would be better referred to Oreotragus sp. The associated dentitions at Makapansgat Limeworks have premolar rows that are probably slightly longer relative to the molar rows than in living O. oreotragus. Palaeotragiscus longiceps Broom (1934, p. 477, fig. 6) is a maxilla from Taung, thought to be an odd neotragine. A right upper molar, BM M-25708 from Kanjera, is probably of Ourebia. Madoqua avifluminis (Dietrich (1950, p. 34) is known by many horn cores and dentitions from Laetolil. Dietrich referred them to Praemadoqua gen. nov., but the horn cores and proportionate lengths of the premolar and molar rows agree with living Madoqua. There is a small back lobe on the M3 agreeing more with subgenus Rhynchotragus than with Madoqua. A complete metatarsal in the Nairobi collection is shorter and thicker than in living dik-diks. A maxilla fragment from the Ngorora Formation, KNM-BN 92, is small enough to fit Neotragini and appears not to belong to a duiker. Tribe Antilopini The Antilopini are small- to medium-sized antelopes that generally live in arid areas. The main features of the living species are braincase not strongly bent down on the face axis, complicated midfrontal and parietofrontal sutures, teeth hypsodont and without basal pillars, central cavities with a simple outline, and P4 normally without paraconid-metaconid fusion. The horn cores of the constituent species differ so much from one another that it is impossible to generalize about them. The living genera are Antilope, Gazella, Antidorcas, Litocranius, and Ammodorcas, but only Gazella, the gazelles, has more than one living species. Antilope cervicapra is the spiral-horned Indian blackbuck and Antidorcas marsupialis, the South African springbok. An African fossil belonging to Antilope is known. Other

Bovidae

559

African fossils are of Antidorcas and Gazella, the former having sinuses in the frontals that raise their level between the horn bases, male horn cores usually bent backward quite strongly, highercrowned teeth, shorter premolar rows, P2 generally absent, and deeper mandibular rami. Antidorcas australis Hendey and Hendey (1968, p. 56) from Swartklip, Elandsfontein, and other Cape Province coastal sites, was originally described as a subspecies of A. marsupialis but is now thought to be a separate species (Hendey 1974, p. 52). It differs from A. marsupialis by its smaller and more mediolaterally compressed horn cores without any sharp bending backward and outward and by its smaller dentitions. It could be another form peculiar to the southern Cape, like Hippotragus leucophaeus and the unnamed kudu. Some horn cores from Elandsfontein and Bolt's Farm are indistinguishable from the Olduvai species A. recki (see below). They are less common at Elandsfontein than is A. australis, and Hendey (1974, p. 52) believes that the A. recki may have been more a species of the inland plateau and ancestral to the living springbok. Almost certainly to be included in Antidorcas is Gazella bondi Cooke and Wells (1951), originally described from Chelmer, Rhodesia, but also at Vlakkraal and Florisbad. It has small teeth, but they are so extremely hypsodont that the lower edge of the mandibular ramus in subadults is seriously distorted. Vrba (1973) has described the first known horn cores and crania from Swartkrans and concludes that A. bondi is not ancestral to A. marsupialis. Klein (pers. comm.) has identified it from Late Stone Age levels of Border Cave, South Africa, dated to about 36,000 B.P. (Beaumont 1973, p. 45). The holotype of Gazella wellsi Cooke (1949, p. 38, fig. 11), a left adult mandible from Power's Site on the Vaal River, is possibly conspecific with A. bondi or A. recki and is discussed by Vrba (1973, p. 311). The common antilopine at Olduvai is Antidorcas recki (Schwarz 1932, 1937, p. 53), originally placed in a new genus, Phenacotragus. It was smaller than A. marsupialis and had horn cores more mediolaterally compressed and often more sharply bent backward in their distal parts, premolar rows less reduced, and shorter radii, tibiae, and metapodials. The Olduvai horn cores show considerable variability, part of which may be due to a temporal succession of different varieties. Gentry (1966, p. 56) had referred most of the Bed I examples to Gazella wellsi, but it now seems wisest not to complicate the East African story with the introduction of this name. It is possible that extensive horn core vari-

560

Gentry

ability in time and space has characterized other Antidorcas species. A. recki is also at East Turkana (Harris 1976), Peninj, and Kanjera. It, or its immediate ancestor, is known from the Shungura Formation, Omo, by a few horn cores with much compression and a slight but localized bending back of their distal parts. It seems that an Antidorcas occurred in Villafranchian-equivalent deposits of Αϊη Brimba, Tunisia (Arambourg and Coque 1958); Coppens (1971) indicates that it was also present at Oued el Atteuch and Αϊη Boucherit in Algeria, and Bayle (1854) had earlier referred to what may be this species from Mansoura. Antidorcas sp. Arambourg (1947, p. 390) from the Omo is the mandible of a small alcelaphine. The description of Gazella gazella praecursor Schwarz (1937) from Olduvai suggests that it could have been A. recki, but the material was lost in Germany during the Second World War. Leakey (in Clark 1959, p. 230) identified Litocranius on some limb bones from Broken Hill, Zambia, but Gazella or Antidorcas seem a more likely identification. Fossil gazelles are well known in North Africa. G. atlantica Bourguignat (1870, p. 84), abundant in the middle and late Pleistocene, seems to have been about the size of a springbok and had rather short, thick horn cores with strong backward curvature and a flattened lateral surface. It is difficult to know which living gazelle to link it with, but it ought to be compared with the early Holocene G. decora and G. arista Bate (1940, pp. 419, 429) of Palestine. G. tingitana Arambourg (1957) is another late Pleistocene North African species, having very long, backwardly curved, and slender horn cores. It might be ancestral to the living G. leptoceros, which has long and slender but straight horn cores. Other late Pleistocene fossils belong to gazelles that survived in the region until recent centuries (G. dorcas, G. rufina, G. cuvieri). Many names proposed by Pomel (1895) for North African gazelles have been amended and synonymized by Arambourg (1957). G. thomasi (Pomel 1895, p. 18), the corrected name for the junior homonym G. atlantica Thomas 1884, comes from Villafranchian-equivalent deposits of North Africa and is a horn core with strong mediolateral compression. The horn cores of G. praethomsoni Arambourg (1947, p. 387) from the Shungura Formation are very similar but with more backward curvature; the mandibular piece assigned to G. praethomsoni by Arambourg (1947, p. 388, pi. 27, fig. 1) is antilopine, but probably Antidorcas. G. setifensis (Pomel 1895, p. 15, pi. 10, figs. 14,15) is another North African gazelle of similar age to G.

Bovidae

thomasi, somewhat larger and with less mediolateral compression and more curvature of its horn cores. The fossil gazelles of eastern and southern Africa are somewhat easier to arrange in tentative lineages. An extinct Gazella is known from several horn cores and mandibles at Elandsfontein far to the south of the present limits for the genus in Tanzania. The same species is found in Olduvai Bed II, Peninj, and Kanam West, and it is less common than Antidorcas at Olduvai, as at Elandsfontein. It has only slightly compressed horn cores, inserted at a low angle and little curved backward. The level of the greatest mediolateral diameter lies more anteriorly in the cross-section than in living G. dorcas; this and the low inclination might align it with the living African G. thomsoni and G. rufifrons. A horn core from Olduvai Bed I and another from the base of middle Bed II are more mediolaterally compressed. The mandibles assigned to this species have a shallower ramus and longer premolar row than Antidorcas. G.janenschi Dietrich (1950, p. 25, pi. 2, fig. 22) is known by horn cores and dentitions from Laetolil and could be ancestral to the Olduvai and Elandsfontein gazelle (see figure 27.7). It differs by its slightly smaller horn cores, divergence lessening a little toward the tips, no flattening of the lateral surface, and stronger backward curvature. Teeth of Gazella hennigi Dietrich (1950, pi. 5, fig. 47) are probably conspecific, but the G. hennigi horn cores are perhaps of an Antidorcas. A second gazelle lineage is represented by G. vanhoepeni (Wells and Cooke 1956, p. 43) from Makapansgat Limeworks, originally placed in Phenacotragus but later removed to Gazella (Wells 1969). It is a large gazelle with strongly compressed horn cores, and Wells was disposed to link it with the three large living African gazelles, G. granti, G. soemmerringi, and G. dama. I think he was right, and I think that the Makapansgat Limeworks G. gracilior Wells and Cooke (1956, p. 37) is likely to be the female of G. vanhoepeni. The species would have had rather less sexual dimorphism of its horn cores than living gazelles. A Langebaanweg gazelle has very compressed horn cores that are curved backward. These features cause it to resemble G. vanhoepeni, to which it may be ancestral. It is possible that the North African G. thomasi and the Omo G. praethomsoni, already mentioned above, are related to the G. vanhoepeni lineage, but the available material of them is insufficient for decision. A large Laetolil antilopine is represented by the teeth of Gazella kohllarseni Die-

Gentry

Antidorcas marsupialis

A.bondi

Gazella soemmerringi, dama.granti

G.thomsoni, rufifrons

G.vanhoepeni

Antidorcas sp. (Laetolil)

Antilope cersicapra

Gazella leptoceros G.atlantica

A.australis

A.recki

561

Bovidae

G.sp.COIduvai, Elandsfontein)

G.tingitana Antilope ?subtorta

G.janenschi Gazella sp. (Langebaanweg)

Pros trepsiceros rotundicornis -10

G.praegaudryi

Gazella sp. (Fort Ternan)

-15 Gazella sp.(Zelten)

Figure 27.7 Tentative phylogeny for African Antilopini. Living species are entered above the dashed line and age is shown on the left in millions of years. Related non-African antilopines are shown in italics. The North African Gazella dorcas, G. cuvieri, and G. rufina are known from the late Pleistocene of North Africa but are of unknown ancestry; they have not been shown. G. praethomsoni and some other enigmatic, but named, horn cores have not been entered.

trich (1950, pi. 2, fig. 16, pi. 5, fig. 49), although the horn core assigned to this name is probably alcelaphine. A gazelle mandible, KNM-MP 129, comes from the Mpesida Beds, Kenya, and G. praegaudryi Arambourg (1959, p. 123, pi. 17, figs. 9-11) was described from Oued el Hammam. Early gazelles have relatively small and little-compressed horn cores and rather brachydont teeth with long premolar rows and lower M3 without a very noticeably enlarged back lobe. The gazelle at Fort Ternan (Gentry 1970a, pp. 292, 299) has several distinctive features and has already been mentioned. Gazella helmoedi Van Hoepen (1932, p. 65, fig. 2) from Cornelia is a long horn core, very narrow for its length and slightly curved in two planes. It is not a gazelle and I can only suggest that it may be an oryx horn core, distorted in life or after fossilization. Wells (1967, p. 102) has suggested that it might belong to an alcelaphine genus (? Parmularius). Two Antilope horn cores from Member C of the Shungura Formation, Omo, can be rather doubtfully likened to A. subtorta Pilgrim (1937), a Siwaliks species from the Pinjor Formation. A. subtorta is an-

cestral to the Indian blackbuck, A. cervicapra (Linnaeus), from which it differs by less twisted horn cores with a trace of a posterolateral keel. Subfamily Caprinae Tribe Ovibovini The Ovibovini are a very diverse group of mostly moderate to large-sized bovids. They have short and often divergent horn cores, which are strongly specialized in some genera. Other characters are a high infraorbital foramen, large occipital condyles and basioccipital, hypsodont cheek teeth without basal pillars, with a short premolar row and paraconidmetaconid fusion on P4. The only two living species are Ovibos moschatus and Budorcas taxicolor, the muskox of the North American Arctic and the takin of Tibet. There are many Eurasian fossils (see Gentry 1971, p. 289) and some in North America. There is growing evidence of the presence of Ovibovini in the African fossil record, although with one exception they are only poorly known. The exception is Makapania broomi Wells and Cooke (1956) from Makapansgat Limeworks. This is a

562

Gentry

moderate to large-sized antelope with horn cores emerging almost transversely from the skull just behind the orbits. Gentry (1970b) believed that it was an ovibovine related to the European Villafranchian Megalovis latifrons Schaub, mainly on details of basioccipital anatomy. Its teeth are fairly hypsodont, without basal pillars, with poor to moderate styles and ribs on the lateral walls of the upper molars, and having a fused paraconid and metaconid on P4. Vrba (1975) records Makapania from the Sterkfontein type site. Bos makapaani Broom (1937, p. 510, figured) from Buffalo Cave in the Makapan Valley appears to be a second ovibovine species, but of uncertain generic identity. The position of sutures on the top of the holotype frontlet suggest that the convex edge of the short curved horn cores is anterior or anterodorsal, whereas Broom evidently thought that the convex edge was posterior. A horn core from Bed I Olduvai, BM M-14531, could be the same species. A fine ovibovine skull, not of Makapania broomi, has been found in the Hadar Formation, Ethiopia. A damaged Langebaanweg cranium, SAM-L 13105, resembles the cranium of the living takin. Damalavus boroccoi Arambourg (1959, p. 120) from Oued el Hammam was described as an alcelaphine but is quite possibly ancestral to Palaeoryx Gaudry, a genus best known from the Samos and Pi-

Bovidae

kermi Hipparion faunas (see figure 27.8). I consider it to be an early ovibovine (Gentry 1970a, p. 132, 1971, p. 284). Tribe Caprini Caprini are moderate-sized bovids, tending to have rather high and narrow skulls, both horn cores and frontals internally hollowed in living species, wide anterior tuberosities of the basioccipital, and a small angle on the mandible. They have hypsodont cheek teeth without basal pillars, rather flat lateral walls between the styles of the upper molars, lower molars often with goat folds, short premolar rows, P4 with paraconid and metaconid fused, and little enlargement of the central incisors. Living species, which are mostly Eurasian, are usually found in hilly habitats or in rocky open areas. Capra, or Ammotragus, lervia is known from North African sites in the late Pleistocene, normally in association with Mousterian stone implements. Churcher (1972) found it in the late Pleistocene of southern Egypt. Romer's (1928, p. 119) reference to it in the Villafranchian appears to be a misinterpretation of a possible Antidorcas reported by Bayle (1854). Fossils related to either of the living African ibexes have not been discovered. Numidocapra crassicornis Arambourg (1949),

Ovibos Budorcas

Capra

Makapania broomi

Megalovis

Numidocapra crassicornis

ALCELAPHINI Pa/oeoryx

Pachy tragus

spp.

-10 Damalavus boroccoi

Pachytragus solignaci Benicerus theobaldi

?Pseudotragus potwaricus Oioceros tanyceras

15

Figure 27.8 Tentative phylogeny for African Ovibovini and Caprini. Living genera are entered above the dashed line and age is shown on the left in millions of years. Non-African forms are shown in italics.

Gentry

based on a frontlet from Αϊη Hanech, Algeria, is large with long, thick horn cores curving upward and forward from their close insertions and with very little divergence. The top of the orbital rim is well below the level of the frontals between the two pedicels, and the short braincase is strongly inclined. Its affinities are unknown, but it is not totally unlike the smaller Procamptoceras brivatense Schaub from the Villafranchian of Seneze, France, which could be related to the chamois,Rupicapra rupicapra, which I think is likely to belong to the Caprini. Therefore I tentatively include Numidocapra in the Caprini. Coppens (1971) refers to a Capra at Αϊη Brimba, for which Arambourg will give further details. The lower of two faunas in the Beglia Formation at Bled Douarah, Tunisia, contains predominantly savanna forms (Robinson and Black 1969), including an abundant caprine Pachytragus solignaci Robinson (1972). Its horn cores are like those of the Samos P. crassicornis Schlosser but have stronger mediolateral compression and more marked anterior keels. The teeth are more brachydont with more rugose enamel, basal pillars on the lower molars, the medial lobes of the upper molars not joined to one another or to the lateral side of the teeth until late in wear, and a smaller hypoconid on P 4 . Probable references to this species by Boule (1910) and Roman (1931) have already been mentioned. P. solignaci is a problematical species. That it has been correctly assigned to the Caprini is shown by the downward slope of the braincase in profile, the upright insertion of the horn cores, the fairly deep ramus of the mandible, and the absence of boselaphinelike specializations of the horn cores. However, the combination of primitive teeth and precociously compressed horn cores is striking, and the animal must be a different lineage from the Samos Pachytragus, which is likely to be at least 3 m.y. younger. One can see analogies in the differences of Hemitragus (tahrs) from Capra (goats) at the present day. Heintz (1973) has named a horn core from Beni Mellal, Morocco, as Benicerus theobaldi. He did not give an opinion about its tribal or subfamily affiliation but pointed out that its torsion was in the same direction as in Oioceros and that it had similarities to the Chinese Prosinotragus. Oioceros tanyceras Gentry (1970a) is well represented at Fort Ternan and is very like O. grangeri Pilgrim (1934) and O. noverca Pilgrim from Tung Gur, Mongolia. It has horn cores curving upward and outward with a slight clockwise torsion on the right side, a short premolar row, and limb bones with cursorial adaptations. It is not very like the

Bovidae

563

Oioceros species from the later Hipparion faunas of Pikermi, Maragha, and Samos. ?Pseudotragus potwaricus (Pilgrim 1939, p. 86) from the Nagri Formation of the Siwaliks also occurs at Fort Ternan (Gentry 1970a, p. 284). It has backwardly curved horn cores but may be fairly closely related phyletically to Oioceros tanyceras. A slightly larger descendant species is represented by a skull from the Ngorora Formation, KNM-BN 100. Both ?P. potwaricus and Ο. tanyceras may be ancestral to Alcelaphini, and their transfer to that tribe may become necessary to preserve monophylety.

The Earliest Bovids in Africa Cooke (1968, p. 248) had expected that a basic tragulid-pecoran stock would be first known in Africa in the later Oligocene, and it is now certain that bovids occur in the African Miocene before Fort Ternan. Hamilton (1973, pp. 126-128) has described from Gebel Zelten, Libya, three horn cores of Protragocerus, two of Eotragus, and a mandible and mandibular fragment of Gazella. The gazelle has a notably deep mandibular ramus, deeper than in the gazelles from Oued el Hammam or Fort Ternan, so it may represent a lineage that later gave rise to Antidorcas. Of course more information is needed to be sure of this. Protragocerus is not known in Europe before the Vindobonian, and Eotragus occurs from the Burdigalian to the Vindobonian. Savage (in Selley 1969, p. 458) and Savage and Hamilton (1973, p. 525) correlated Gebel Zelten with the early Burdigalian of Europe or even the Aquitanian but did not suppose that it was of an earlier age than the East African Miocene sites that predate Fort Ternan. Well known as a possible early bovid is Walangania africanus (Whitworth 1958, p. 19), a species in which Hamilton (1973, p. 146) has amalgamated "Palaeomeryx" africanus and W. gracilis Whitworth (1958, p. 30). The holotype is a mandible from Songhor. Many other tooth remains are known from Songhor, Rusinga, and other sites, and there are also an incomplete dentition and limb bones of a single juvenile individual from Rusinga. The femur and tibia are about the size of an adult of the Fort Ternan species Protragocerus labidotus so would have a size appropriate for the young of a slightly larger species. The morphology of the cheek teeth does not suggest much difference at a generic level from Eotragus Pilgrim (see Thenius 1952). In fact the morphology of P 3 and P4 shows more of a boselaphine than a caprine pattern (see Gentry 1970a, p. 265, fig. 2). On the other hand the transversely narrow rear

564

Gentry

part of the top articular surface of the metatarsal does not suggest an affinity with the Boselaphini (see Gentry 1970a, pp. 248, 280, fig. 12). Possibly Walangania was a boselaphine ancestor of some caprine lineage. Alternatively it may not be a bovid but an early pecoran close to Dremotherium or Amphitragulus as known from the Oligocene to Miocene of Europe. Other possible remains of early African bovids include the mandibular fragment of cf Strogulognathus (=Eotragus) sansaniensis from the Namib Desert of Namibia (Stromer 1926, p. 115), if this is not identical with Propalaeoryx austroafricanus (see below). There are two partial bovid crania from Maboko, BM M-15543 and M-15544, which must belong to different species. One of them could be a late occurrence of Walangania. Bovid teeth from Maboko have already been referred to by Whitworth (1958, p. 25) and Gentry (1970a, p. 303). Stromer (1926, p. 117) described Propalaeoryx austroafricanus from the Namib Desert as a bovid, but Whitworth (1958, p. 26) assigned it and an East African species, P. nyanzae, only to the Pecora, and Hamilton (1973, p. 142) has now placed Propalaeoryx in the giraffoid family Palaeomerycidae (see chapter 25). The shallow ramus of the first species does not easily suggest bovid affinities. An indeterminate bovid from Losodok (Arambourg 1933, p. 142, 1947, pi. 22, figs. 5, 6) is more likely to be giraffoid (Gentry 1970a, p. 302). In Europe the earliest bovid is Eotragus artenensis Ginsburg and Heintz (1968) from the Burdigalian of Artenay, France. In Mongolia supposed bovids have been described from the lower Miocene of the western Gobi (Gobiocerus mongolicus Sokolov 1952) and from the Hsanda Gol Formation of Oligocene age at Tatal Gol (Palaeohypsodontus asiaticus Trofimov 1958). The identity at family level of the Asiatic species, which are based only on teeth, cannot be regarded as certain, for middle Tertiary nonsuiform artiodactyls on several continents were showing trends such as widening of the premolars, increasing selenodonty of the molars, development of hornlike appendages on the top of the skull, fusion of carpal bones, and appearance of cannon bones. It is not yet clear what characters can be taken as diagnostic of what groups of what time levels, to what extent the Bovidae are monophyletic, or whether they evolved in one place. Can early bovidlike teeth, especially in Asia, be presumed not to belong to ancestors or relatives of the musk-deer? The final words on the origin of the Bovidae can be taken from Hamilton (1973, p. 134):

Bovidae

The Pecora probably originated from the Traguloidea during the Upper Eocene or Lower Oligocene and of the two traguloid families the Gelocidae are the most likely to have given rise to the Pecora. In the gelocids true selenodonty is developed from more bunodont forms; thus Lophiomeryx has very bunoid lower molars showing few signs of true selenodonty while Bachitherium and Prodremotherium have molars which are very similar to those of Dremotherium. A detailed study of this group is needed and it is here that the divergence of the Bovoidea and other higher ruminants probably occurred.

Discussion Classification and Zoogeography Schlosser (in Zittel 1925, p. 208), Pilgrim (1939, p. 10), and Simpson (1945, p. 270) have discussed the extent to which the bovids may be divided into two groups called Böodontia and Aegodontia. This grouping, although not used in the formal classification of this chapter, does reflect important features of bovid evolution. The limits of the two groups have not been agreed upon, but böodonts could be conceived as comprising the Bovinae, Cephalophinae, and Hippotraginae, and aegodonts the Alcelaphinae, Antilopinae, and Caprinae. Böodonts tend to have lower and wider skulls, braincases little angled on their faces, horn cores more frequently keeled, internal sinuses of the frontals less frequent, less hypsodont teeth, rugose enamel, basal pillars persisting and declining later in evolution if at all, a slower rate of fusion between the lobes of the molar teeth in ontogeny, longer premolar rows, shallower mandibles, and less cursorial limb bones. (Gentry 1970a, pp. 277-282, figs. 12-15 has listed cursorial characters of bovid limb bones.) Aegodonts have the converse characters. The differentiating characters of böodonts and aegodonts reach their clearest manifestation in the late Miocene to early Pliocene. Böodonts have not exploited such ecologically extreme environments as aegodonts, and it seems likely that aegodonts evolved from early bovids with predominantly böodont characters. Walangania could be an incipient aegodont. Figures 27.9 and 27.10 show representative horn cores of various bovid groups. As far as Africa is concerned, the basic split between browsing böodont Boselaphini and more cursorial aegodont Caprini was clear by 14 m.y. ago at Fort Ternan. At this stage there was still much similarity among the bovid faunas everywhere in the Old World. Only the species, not the genera, differed from continent to continent. Regional variation began to be more marked by the time of the famous

Gentry

Bovidae

565

Figure 27.9 Some African fossil bovids. (A) Megalotragus priscus (Alcelaphini), (Β) Pelorovis antiquus (Bovini), (C) Hippotragus

gigas

(Hippotragini), (.D) Tragelaphus

strepsiceros

Eurasian Hipparion faunas of Pikermi, Samos (8 to 9 m.y., van Couvering and Miller 1971), Maragha, and northern China. In Europe and China Miotragocerus had become the only boselaphine, but there had appeared various ovibovines such as Palaeoryx and Criotherium, the earliest Caprini, and spiralhorned Antilopini. The equivalents in Africa of these Hipparion faunas are not at all well known, but later at Langebaanweg and Laetolil tragelaphines, reduncines, and alcelaphines had appeared, the first two probably descended from boselaphines and the third from caprines. However boselaphines and ovibovines did not die out in Africa (cf Gentry 1970a, p. 313). In connection with this faunal divergence between Africa and Palaearctic Eurasia, it is interesting that Siwaliks bovids have more resemblances to African than to other Eurasian forms. This is evidenced by the similarity of Ugandax to Proamphibos (but the difference of either from the Bos ancestry is unknown), at least one reduncine lineage from the Dhok Pathan to the Pinjor Formation, the presence of two hippotragines in the Pinjor Formation, the Pinjor alcelaphine Damalops, and the presence of Antilope in the Shungura Formation. The presence of Oryx in Arabia, the former presence of an alcelaphine, probably hartebeest, in Palestine and Jordan (Garrod and Bate 1937, p. 215; Clutton-Brock 1970, p. 26), and the possible lesser kudu record for Arabia also suggest a faunal link from Africa across Arabia. Kurten (1957, p. 223) believed that the Indian fauna

grandis

(Tragelaphini). ( x 1/20.)

has only acquired its present resemblance to the Eurasian one since the Villafranchian. The decline of "African" bovids in that region may be connected with the burgeoning of deer. Bearing in mind the böodont-aegodont split and the affinity between Siwalik and African bovids, we may consider the tribes of antelopes from a zoogeographical standpoint.2 The Tragelaphini are now almost confined to Africa. All fossil records from other continents are incorrect or unsubstantiated. The most long-standing misidentifications are of Eurasian Tertiary spiral-horned Antilopini (see Gentry 1971). Pilgrim (1939, p. 131) claimed that the Chinji frontlet Sivoreas eremita was tragelaphine, but it now seems more likely (cf Gentry 1970a, p. 257) that it is boselaphine and perhaps even conspecific with Protragocerus gluten, also from the Chinji. Tragelaphines probably descended from early boselaphines, and basal pillars on the teeth became less prominent in tragelaphines than in other böodont tribes. The Boselaphini is the tribe that probably includes the oldest bovid fossils. They seem to have disappeared from most of the Palaearctic soon after the Samos time level but were abundant in the Dhok Pathan Formation (Miotragocerus, Tragoportax, Pachyportax, Selenoportax). At least one lineage lasted well into the Tertiary in Africa and had resem2 This discussion will not take into account the fact that some bovids have immigrated into North America.

566

Gentry

Bovidae

Figure 27.10 Some African fossil bovids. (A) Taurotragus arkelli (Tragelaphini), (Β) Tragelaphus nakuae (Tragelaphini), (C) Protragocerus labidotus (Boselaphini), (D) Menelikia lyrocera (Reduncini), (E) Thaleroceros radiciformis (Reduncini), (F) Antidorcas recki (Antilopini), (G) Rabaticeras arambourgi (Alcelaphini), (H) Parmularius altidens (Alcelaphini), (J) Ρachy tragus solignaci (Caprini). (x 1/8.) blances to Siwaliks forms. It is noteworthy that the only two present-day boselaphines are Indian (but see Clutton-Brock 1970, p. 25, for an upper Pleistocene record in Jordan) and that they therefore now constitute an element of faunal difference from Africa. The extensive internal sinuses of the frontals in the Langebaanweg boselaphine are unusual for a böodont. The Bovini are mostly large-sized descendants of Boselaphini that appeared in the later Tertiary. Two closely related lineages, Simatherium —* Pelorovis and Ugandax —* Syncerus, are confined to Africa, although Ugandax is quite like the Tatrot Proamphibos, which probably gave rise through the Pinjor Hemibos to Bubalus. Hemibos has been recorded from Palestine (Pilgrim 1941). The other bovine stock is the one that gave rise to Bos, a Palaearctic

and Indian genus not found in Africa south of the Sahara Desert. Its ancestry is unknown (Siwalik Pachyportax, European Pliocene Parabos?), and it is puzzling that the Villafranchian and Pinjor "ox," Leptobos, has hornless females, whereas in Bos the females are horned. With the Bovini, as with Tragelaphini, it is difficult to differentiate early members from the Boselaphini. Cephalophini are unknown outside Africa and, being mainly forest dwellers, are almost unknown in the fossil state. They are quite likely to have a boselaphine ancestry. Reduncines are found in Africa and the Siwaliks, and it is interesting that as a group they are dependent on habitats close to water, yet the intervening area of Arabia is today so arid. The earliest reduncines come from the Dhok Pathan Formation, Wadi

Gentry

Natrun, Sahabi, the Lukeino Formation, and Langebaanweg. Khomenko (1913) claimed reduncines from the Russian Tertiary, but I agree with Pilgrim's (1939, p. 95) reservations about their tribal assignations; at least one of them may be a Palaeoryx. Hippotragini are likewise confined to the Siwaliks and Africa. Gentry (1971) has argued that alleged Tertiary hippotragines from Samos and other Palaearctic Hipparion faunas are misidentified ovibovines and caprines. The earliest likely hippotragine comes from Sahabi. The Pinjor ones give hints of a boselaphine ancestry, but there are no earlier boselaphines that appear to be the actual ancestors. Alcelaphini, the first of the aegodont tribes in the list, are also limited to the Siwaliks, Africa, and the intervening areas. The only Siwalik representative, Damalops palaeindicus, could be related to the hartebeests, so this is the only stock for which an extraAfrican representation is likely. The earliest adequately preserved alcelaphines are the interesting undescribed species from Langebaanweg, but they are also known from Wadi Natrun, and earlier African and Siwalik caprines may include alcelaphine ancestors. The small Neotragini are only known from Africa. The Antilopini is a difficult tribe zoogeographically. Gazella is a very early bovid and has survived in arid areas of the Palaearctic, Indian, and Ethiopian realms. G. rufifrons and thomsoni on the one hand and G. dama, soemmerringi, and granti on the other constitute two African groups of gazelles. G. dorcas and related species constitute one or more Palaearctic groups. Procapra (which I would include in Gazella), Saiga, and Pantholops are Asiatic genera probably descended from Gazella, while Antidorcas, Ammodorcas, and Litocranius are African genera. Gentry (1970a, pp. 295-300) hypothesized that Eurasian fossil gazelles may be divisible into two groups, one ancestral to Procapra and the other to the G. dorcas group. In Africa, fossils are known that appear to be ancestral to Antidorcas and probably to both groups of African gazelles. Spiral-horned Antilopini are known from the Samos and other Hipparion faunas of the Palaearctic. By the Pleistocene they seem to have resolved themselves into two stocks, the more northerly Spirocerus (Asia) and Gazellospira (Europe) and the more southerly Antilope, which was now present in India (did it and the Leptobos/Bos bovines immigrate from elsewhere in Eurasia?) and northeastern Africa. Ovibovini are more abundant in the fossil record than at present. One archaic group with peculiar horn cores comprises the Miocene genera Criother-

Bovidae

567

ium, Palaeoreas, Urmiatherium (including Parurmiatherium and Plesiaddax), and Tsaidamotherium, none of which has been found in Africa. A second and longer-lasting group has been present in Africa. The only adequately known African species is Makapania broomi; together with Megalovis and Budorcas it may have formed a more southerly ranging stock than Ovibos. No ovibovines have been recorded from the Siwaliks. Caprini are basically a Palaearctic group that have scarcely penetrated Africa south of the Sahara. However, early African bovids from which Alcelaphini probably descended, have been "horizontally" classified with their contemporaneous Eurasian relatives as Caprini (Oioceros and ?Pseudotragus at Fort Ternan). North Africa contained Caprini at least at certain periods. The tribe called Rupicaprini is an essentially Eurasian stock and has not been recorded from Africa. In summary, and still ignoring Nearctic forms, we see that the Siwaliks and Africa contain rather more Tertiary boselaphines than does the Palaearctic. From such boselaphines there probably evolved the Tragelaphini, one southeastern Asian and two African lineages of Bovini, the Cephalophini, the Reduncini, and the Hippotragini. Cattle, bison, and their ancestors comprise the only major Palaearctic böodont group. Among the aegodonts the Antilopini can be broken down into Palaearctic and African-Siwaliks forms at generic level, except in Gazella, where the zoogeographical separation is at species or species-group level. The Alcelaphini and Neotragini are totally African-Siwalik, and the former tribe descended from early Caprini, which may be regarded as the aegodont parallel to early Boselaphini. Of the Ovibovini, one early group is Palaearctic, but another is found in Africa and the Palaearctic. The tribe called Rupicaprini is totally Palaearctic and southeast Asian, and the Caprini became almost totally Palaearctic after its beginning in the Miocene. The tribal changes in the Siwalik-Africa area from early Boselaphini to Tragelaphini and their other descendants, and from early Caprini to Alcelaphini, are an artificial result of the "horizontal" classification of the earliest bovids.

Faunal Evolution in Bovids Although nearly all tribes and genera of bovids are characteristic of particular faunal realms, it is difficult to generalize about regional bovid faunas in units smaller than realms. Admittedly the southern part of Cape Province appears to have had a number of endemic species in the recent past, as witnessed

568

Gentry

by the extinct kudu, the blaauwbok, the grysbok, and the extinct springbok, Antidorcas australis. This possible exception apart, it appears that regional groupings of species are not clear cut within the faunal realms, although species differ from habitat to habitat. Arising from such considerations, together with the inadequate knowledge of former species' ranges allowed by the fossil record, it is impossible to give precise information about where particular bovids evolved. It is worth repeating that the involvement of the Siwaliks in the history of African bovids does not necessitate the opinion that African bovids originated in India. It is better to say no more than that the two areas have shared many of their antelopes, at least at the generic level. Faunas in local areas must certainly have evolved as integrated units with a degree of interdependence between the species. Theoretically one may expect that a morphological change in one lineage, which enables or reflects a modification in its ecology, might well produce smaller compensating changes in a number of contemporaneous, sympatric, but phylogenetically unrelated lineages. Such a phenomenon would be difficult to identify in fossil faunas, and it would produce gradual evolution rather than dramatic changes. Any marked faunal changes that do occur in the fossil record are likely to reflect only local changes of ecology and habitats. North Africa is faunally interesting not for incipient endemism like South Africa but as the area where the Palaearctic and Ethiopian realms meet. The oldest known bovids in the region are those of Gebel Zelten, which are not zoogeographically distinctive, so far as they are known at present, and could correlate with other Old World sites down to Fort Ternan at 14 m.y. Later in North Africa the Benicerus of Beni Mellal, the Pachytragus of the Beglia Formation, and the Damalavus of Oued el Hammam are Caprinae with more or less convincing affinities with Eurasian forms. Later still at Sahabi and Wadi Natrun the affinities change. Sahabi has Miotragocerus, Leptobos, a reduncine, and a possible hippotragine, and Wadi Natrun has a reduncine and alcelaphine. One could align the Miotragocerus and Leptobos with European faunas and the rest with African faunas, or one could align all the bovids with the Siwaliks. It will be interesting to see if future research can substantiate a faunal change of some kind preceding the Sahabi and Wadi Natrun faunas or clarify the zoogeographical affinities of those faunas.

Character Evolution It looks as if changes in particular characters in bovid lineages have not always been from the more

Bovidae

primitive to the more advanced. Three examples are: (1) Pelorovis oldowayensis has a more posterior insertion of its horn cores than the later P. antiquus; (2) paraconid-metaconid fusion of kudu P 4 occurs in a higher percentage of Olduvai than extant specimens; and (3) the premolar row ofKobus sigmoidalis may have been shorter than in the living waterbuck. A partial explanation of the tooth characters may be that some living antelopes have less narrow adaptive ranges than their ancestors. A greater number of species in the past would have allowed narrower specializations of individual species. Also, living species may sometimes have evolved from atypical populations of their predecessors, or antelope lineages may have consisted of series of sequentially replacing species (and not chronospecies transitional to one another), as Martin (1972, p. 316) has postulated for lemurs. This latter possibility would allow slightly different adaptive characters in successive species (but, if correct, would necessitate splitting among some species names as used here).

Extinctions Tragelaphus nakuae, Menelikia, Makapania, and Numidocapra appear to have become extinct sometime in the early Pleistocene and Mesembriportax probably earlier. Apart from these, there may have been few extinctions until the end of the Pleistocene. Klein (1972, pers. comm.) has produced definite evidence of Pelorovis and Megalotragus in the Cape Province until about 12,000 to 11,000 B.P. and 15,000 to 14,000 B.P. respectively, and Antidorcas australis, A. bondi, and Damaliscus niro may have been other late surviving species. In East Africa there is no evidence for the extinction of Thaleroceros, Hippotragus gigas, or Parmularius within the sequence of Olduvai Beds I to IV. Beatragus hunteri could be a species naturally on the verge of extinction at the present time.

Correlations The contribution of bovids to faunal correlations has been implicit in many of the statements in this chapter. Only a summary of major points can be given. At Elandsfontein the eland is definitely more advanced than the Olduvai Bed IV Taurotragus arkelli, the Rabaticeras arambourgi corresponds closely with that in Olduvai Bed III, and the gazelle appears identical with that of middle and upper Bed II. Many Elandsfontein bovids are not greatly different from living South African species. A likely conclusion from all this is that much of the Elandsfontein time span corresponds to Olduvai Bed IV or later.

Gentry

Makapansgat Limeworks has a Tragelaphus with rather a primitive horn core cross-section, Redunca darti that is substantially less advanced than living Redunca, Gazella vanhoepeni that is perhaps descended from the Langebaanweg gazelle, and Makapania broomi whose closest relative is in the European Villafranchian. None of this is very conclusive, but an age about the same or a little earlier than 01duvai Bed I would probably best fit the bovid evidence. The Shungura Formation is known by potassiumargon dating to immediately predate and overlap Olduvai. Its bovids comprise mainly tragelaphines, reduncines, and impala and so cannot be compared easily with those of Olduvai, which are mainly alcelaphines (not impala) and antilopines. At Laetolil the Simatherium kohllarseni, ?Parmularius sp., Beatragus sp., and Gazella janenschi and the supposed hippotragine teeth are all more primitive than Olduvai forms and suggest that some horizons substantially predate Olduvai. No reduncines are known from Laetolil, a fact that is ecologically interesting. The chief interest of the Langebaanweg bovids is that the alcelaphine teeth are more primitive than any at Laetolil, which would suggest an earlier date. The Tragelaphus has a primitive horn core crosssection, and the gazelle predates that at Makapansgat Limeworks. In the Kaiso Formation the Menelikia lyrocera of the later fauna would best fit a time level around Member F of the Shungura Formation. The tragelaphine horn core of the early fauna seems insufficiently primitive to match the supposed age of the fauna and the Kobus sigmoidalis is notably early.

The Contribution of Bovids to the Definition of Land Mammal Ages The most recent faunal division presenting itself in the history of African bovids is that marked by the extinctions of a number of lineages near the end of the Pleistocene. So far reasonably precise dates have been secured for only two lineages, but should other extinctions be found to have occurred within a few millenia of them, then the boundary of two land mammal ages could reasonably be fixed here. In South Africa the immediately preceding part of the later Pleistocene, in which are present the extinct forms and slightly different subspecies of some living species, has been called the Florisbad faunal span or Florisian mammal age. It seems doubtful that the bovids of the still earlier Cornelia faunal span (Elandsfontein and Cornelia sites) are sufficiently distinct from Florisbad-Florisian ones to help in the definition of another land mammal age.

Bovidae

569

Passing still further back into the lower Pleistocene, one notices increasing differences among the bovids. Bed IV at Olduvai contains Taurotragus arkelli instead of the living T. oryx, and upper Bed II contains Pelorovis oldowayensis instead of the later P. antiquus. At a level somewhere between the top of Bed I and the later parts of Bed II the differences become sufficiently marked to justify having a boundary between two land mammal ages. There is a faunal change at this level at Olduvai, and it may be coeval with the ending of the Sterkfontian faunal span in South Africa. In the preceding period back as far as, say, Member C of the Shungura Formation the bovids are definitely different and often more archaic than living ones. One has Tragelaphus pricei instead of T. scriptus, Kobus sigmoidalis and Redunca darti instead of Κ. ellipsiprymnus and the living Redunca species, a Connochaetes ancestral to C. taurinus, Parmularius altidens instead of the later P. angusticornis or P. rugosus, as well as lineages that may scarcely have survived into the Pleistocene at all: Tragelaphus nakuae, Menelikia lyrocera, and Makapania broomi. As studies proceed it is becoming clear that another distinct stratum of bovids exists at such earlier sites as Laetolil, Afar, and certainly Langebaanweg. It may become convenient in the future to visualize this faunal level as extending upward to the top of Member Β of the Shungura Formation. Member Β contains a distinctive variety of Tragelaphus nakuae and a Kobus species, which both have antecedents at Afar. Thus the three boundary lines one can draw across the flow of African bovid evolution lie at the end of the Pleistocene, around the Olduvai lowermiddle Bed II junction corresponding to the end of the Sterkfontian in South Africa, and low in the Shungura Formation. It would be premature to extend this process any further back into the Tertiary for the present. I am grateful to G. B. Corbet, J. Clutton-Brock, W. R. Hamilton, and my wife for reading the first draft of this paper.

References Andrews, C. W. 1902. Note on a Pliocene vertebrate fauna from the Wadi-Natrun, Egypt. Geol. Mag. (4) 9:433-439. Ansell, W. F. H. 1971. Artiodactyla. In J. Meester and H. W. Setzer, eds., The mammals of Africa: an identification manual, part 15. Washington: Smithsonian Institution Press, pp. 1-93. Arambourg, C. 1933. Mammiferes miocenes du Turkana (Afrique Orientale). Ann. Paleont. 22:121-148.

570

Gentry

1938. Mammiferes fossiles du Maroc. Mem. Soc. Sei. Nat. Phys. Maroc 46:1-74. 1941. Antilopes nouvelles du Pleistocene ancien de l'Omo (Abyssinie). Bull. Mus. Natn. Hist. Nat. Paris (2) 13:339-347. 1947. Contribution ä l'etude geologique et paleontologique du bassin du lac Rudolphe et de la basse vallee de l'Omo. Mission Scient. Orno 1932-1933 I, Giol. Anthrop., pp. 232-562. 1949. Numidocapra crassicornis, nov. gen., nov. sp., un Ovicaprine nouveau du Villafranchien constantinois. C. R. Soc. Geol. Fr. no. 13:290-291. 1957. Observations sur les gazelles fossiles du Pleistocene superieur de l'Afrique du Nord. Bull. Soc. Hist. Nat. Afr. Nord 48:49-81. 1959. Vertebres continentaux du Miocene superieur de l'Afrique du Nord. Mem. Carte Geol. Algerie, n.s,,Paleont. 4:1-159. 1962. Les faunes mammalogiques du Pleistocene circummediterraneen. Quaternaria 6:97-109. 1970. Les vertebres du Pleistocene de l'Afrique du Nord. Archs. Mus. Natn. Hist. Nat. Paris (7) 10:1-126. Arambourg, C., and R. Coque. 1958. Le gisement villafranchien de l'Aün Brimba (sud Tunisien) et sa faune. Bull. Soc. Geol. Fr. (6) 8:607-614. Arambourg, C., and J. Piveteau. 1929. Note preliminaire sur un ruminant du Pliocene inferieur du Roussillon. C. R. Soc. Geol. Fr. no. 10:144-146. Balout, L. 1942. Note sur la presence des restes fossiles d'une Addax nasomaculata Blainv. parmi des ossements decouverts au Pare d'Hydra, commune de Birmandreis (Alger). Bull. Soc. Hist. Nat. Afr. Nord. 33:138-140. Bate, D. M. A. 1940. The fossil antelopes of Palestine in Natufian (Mesolithic) times, with descriptions of new species. Geol. Mag. 77:418-443. 1949. A new African fossil long-horned buffalo. Ann. Mag. Nat. Hist. Lond. (12) 2:396-398. 1951. The mammals from Singa and Abu Hugar. Brit. Mus. Nat. Hist., Fossil Mammals of Africa no. 2:1-28.

Bayle, E. 1854. Sur une collection d'ossements fossiles trouves pres de Constantine. Bull. Soc. Giol. Fr. (2) 11:343-345. Beaumont, P. B. 1973. Border Cave—a progress report. S. Afr. J. Sei. 69:41-46. Berggren, W. Α., and J. A. van Couvering. 1974. The late Neogene. Palaeogeogr. Palaeoclimat. Palaeoecol. 16: 1-216. Bishop, W. W., G. R. Chapman, A. Hill, and J. A. Miller. 1971. Succession of Cainozoic vertebrate assemblages from the northern Kenya Rift Valley. Nature 233:389394. Bishop, W. W., J. A. Miller, and F. J. Fitch. 1969. New potassium-argon age determinations relevant to the Miocene fossil mammal sequence in East Africa. Am. J. Sei. 267:669-699. Boule, M. 1910. Sur quelques vertebres fossiles du Sud de la Tunisie. C. R. Soc. Geol. Fr., pp. 50-51. Bourguignat, J. R. 1870. Histoire du Djebel Thaya et des

Bovidae

ossements fossiles recueillis dans la grande Caverne de la Mosquee. Paris, pp. 1-108. Broom, R. 1909. On a large extinct species ofBubalis. Ann. S. Afr. Mus. 7:279-280. 1913. Man contemporaneous with extinct animals in South Africa. Ann. S. Afr. Mus. 12:13-16. 1934. On the fossil remains associated with Australopithecus africanus. S. Afr. J. Sei. 31:471-480. 1937. Notices of a few more new fossil mammals from the caves of the Transvaal. Ann. Mag. Nat. Hist. Lond. (10) 20:509-514. Butzer, K. W., and G. LI. Isaac, eds. 1975. After the australopithecines: stratigraphy, ecology and culture change in the middle Pleistocene. The Hague: Mouton. Churcher, C. S. 1972. Late Pleistocene vertebrates from archaeological sites in the plain of Kom Ombo, upper Egypt. Contr. Life Sei. Div. Roy. Ont. Mus. 82:1-172. Churcher, C. S., and P. E. L. Smith. 1972. Kom Ombo: preliminary report on the fauna of late Paleolithic sites in upper Egypt. Science 177:259-261. Clark, J. D. 1959. Further excavations at Broken Hill, northern Rhodesia. J. Roy. Anthrop. Inst. Gr. Brit. 89:201-232. Clutton-Brock, J. 1970. The fossil fauna from an upper Pleistocene site in Jordan. J. Zool. Lond. 162:19-29. Cooke, Η. B. S. 1941. A preliminary account of the Wonderwerk Cave, Kuruman District. Section 2. The fossil remains. S. Afr. J. Sei. 37:303-312. 1947. Some fossil hippotragine antelopes from South Africa. S. Afr. J. Sei. 43:226-231. 1949. Fossil mammals of the Vaal River deposits. Mem. Geol. Surv. S. Afr. 35, part 3:1-109. 1968. Evolution of mammals on southern continents. II. The fossil mammal fauna of Africa. Quart. Rev. Biol. 43:234-264. Cooke, Η. B. S., and S. C. Coryndon. 1970. Pleistocene mammals from the Kaiso Formation and other related deposits in Uganda. In L. S. B. Leakey and R. J. G. Savage, eds., Fossil Vertebrates of Africa. London: Academic Press, vol. 2, pp. 107-224. Cooke, Η. B. S., and L. H. Wells. 1951. Fossil remains from Chelmer, near Bulawayo, Southern Rhodesia. S. Afr. J. Sei. 47:205-209. Coppens, Y. 1971. Les vertebres Villafranchiens de Tunisie: gisements nouveaux, signification. C. R. Hebd. Seanc. Acad. Sei., Ser D, 273:51-54. Debruge, A. 1906. La grotte du Fort Clauzel. C. R. Assoc. Frang. Avanc. Sei. (1905) 34:624-632. Deperet, C. 1890. Les animaux pliocenes du Roussillon. Mem. Soc. Giol. Fr., Paleont. no. 3:1-194. Dietrich, W. O. 1941. Die säugetierpaläontologischen Ergebnisse der Kohl-Larsen'schen Expedition 1937-1939 im nördlichen Deutsch-Ostafrika. Zentbl. Miner. Geol. Paläont. Ser. B, no. 8:217-223. 1942. Ältestquartäre Säugetiere aus der südlichen Serengeti, Deutsch Ost-Afrika. Palaeontogr. Abt. A 94:43-133. 1950. Fossile Antilopen und Rinder Aquatorialafrikas. Palaeontogr. Abt. A 99:1-62.

Gentry

Duvernoy, G. L. 1851. Note sur une espece de buffle fossile, Bubalus (Ami) antiquus, decouverte en Algerie. C. R. Hebd. Seanc. Acad. Sei. 33:595-597. Ennouchi, E. 1953. Un nouveau genre d'Ovicaprine dans un gisement Pleistocene de Rabat. C. R. Soc. Geol. Fr. no. 8:126-128. Ficheur, E., and A. Brives. 1900. Sur la decouverte d'une caverne ä ossements, ä la carriere des Bains-Romains, ä l'ouest d'Alger. C. R. Hebd. Seanc. Acad. Sei. 130: 1485-1487. Gaillard, C. 1934. Contribution ä l'etude de la faune prehistorique de l'figypte. Arch. Mus. Hist. Nat. Lyon 14, 3:1-126. Garrod, D. A. E., and D. M. A. Bate. 1937. The Stone Age of Mount Carmel I. Oxford: Clarendon Press. Gentry, A. W. 1966. Fossil Antilopini of East Africa. Bull. Brit. Mus. Nat. Hist. (Geol.) 12:45-106. 1967. Pelorovis oldowayensis Reck, an extinct Bovid from East Africa. Bull. Brit. Mus. Nat. Hist. (Geol.) 14:243-299. 1970a. The Bovidae (Mammalia) of the Fort Ternan fossil fauna. In L. S. B. Leakey and R. J. G. Savage, eds., Fossil Vertebrates of Africa. London: Academic Press, vol. 2, pp. 243-323. 1970b. Revised classification for Makapania broomi Wells and Cooke (Bovidae, Mammalia). Palaeont. Afr. 13:63-67. 1971. The earliest goats and other antelopes from the Samos Hipparion fauna. Bull. Brit. Mus. Nat. Hist. (Geol.) 20:229-296. 1974. A new genus and species of Pliocene boselaphine (Bovidae, Mammalia) from South Africa. Ann. S. Afr. Mus. 65:145-188. Gervais, P. 1867-1869. Zoologie et paleontologie generates. 1st series, Paris, 263 pp. Ginsburg, L., and E. Heintz. 1968. La plus ancienne antilope d'Europe, Eotragus artenensis du Burdigalien d'Artenay. Bull. Mus. Natn. Hist. Nat. Paris (2) 40: 837-842. Grubb, P. 1972. Variation and incipient speciation in the African buffalo. Zeits. Saügetierk. 37:121-144. Hamilton, W. R. 1973. The lower Miocene ruminants of Gebel Zelten, Libya. Bull. Br. Mus. Nat. Hist. (Geol.) 21:73-150. Harris, J. M. 1976. Bovidae from the East Rudolf succession. In Y. Coppens, F. C. Howell, G. LI. Isaac, and R. E. F. Leakey, edsEarliest man and environments in the Lake Rudolf Basin. Chicago: University of Chicago Press, pp. 293-301. Harrison, D. L. 1972. The mammals of Arabia, vol. 3. London: Benn. Heintz, E. 1973. Un nouveau bovide du Miocene de Beni Mellal, Maroc: Benicerus theobaldi n. g., n. sp. (Bovidae, Artiodactyla, Mammalia). Ann. Scient. Univ. Besancon (3) 18:245-248. Hendey, Q. B. 1968. The Melkbos site: an upper Pleistocene fossil occurrence in the southwestern Cape Province. Ann. S. Afr. Mus. 52:89-119. 1970. A review of the geology and palaeon-

Bovidae

571

tology of the Plio-Pleistocene deposits at Langebaanweg, Cape Province. Ann. S. Afr. Mus. 56:75-117. 1974. The Late Cenozoic Carnivora of the southwestern Cape Province. Ann. S. Afr. Mus. 63: 1-369. Hendey, Q. B., and H. Hendey. 1968. New Quaternary fossil sites near Swartklip, Cape Province. Ann. S. Afr. Mus. 52:43-73. Hopwood, Α. Τ. 1934. New fossil mammals from Olduvai, Tanganyika Territory. Ann. Mag. Nat. Hist. Lond. (10) 14:546-550. 1936. New and little-known fossil mammals from the Pleistocene of Kenya Colony and Tanganyika Territory. Ann. Mag. Nat. Hist. Lond. (10) 17:636-641. Joleaud, L. 1918. Etudes de geographie zoologique sur la Berberie. Ill: Les hippotragines. Bull. Soc. Geogr. Oran 38:89-118. Khomenko, J. P. 1913. La faune Meotique du village Taraklia du District de Bendery. Ezheg. Geol. Mineral. Rossii 15:107-143. Klein, R. G. 1972. The late Quaternary mammalian fauna of Nelson Bay Cave (Cape Province, South Africa). Quatern. Res. 2:135-142. 1974. On the taxonomic status, distribution, and ecology of the blue antelope, Hippotragus leucophaeus (Pallas, 1766). Ann. S. Afr. Mus. 65:99-143. Kurten, Β. 1957. Mammal migrations, Cenozoic stratigraphy, and the age of Peking man and the Australopithecines. J. Paleont. 31:215-227. Leakey, L. S. B. 1965. Olduvai Gorge 1951-61.1. Fauna and Background. Cambridge: Cambridge University Press. Leakey, M. D., R. L. Hay, G. H. Curtis, R. E. Drake, Μ. K. Jackes and T. D. White. 1976. Fossil hominids from the Laetolil Beds. Nature 262:460-466. Leakey, R. E. F. 1969. Early Homo sapiens remains from the Omo River region of southwest Ethiopia. Nature 222:1132-1133. Lydekker, R. 1878. Crania of ruminants from the Indian Tertiaries, and supplement. Pal. Indica (10) 1:88-181. Martin, R. D. 1972. Adaptive radiation and behaviour of the Malagasy lemurs. Phil. Trans. Roy. Soc. London, Series B, 264:295-352. Mohr, Ε. 1967. Mammalia depicta, der Blaubock Hippotragus leucophaeus (Pallas, 1766), eine Dokumentation. Hamburg: Verlag Paul Parey. Petrocchi, C. 1956. I Leptobos di Sahabi. Boll. Soc. Geol. Ital. 75(l):206-238. Pilgrim, G. E. 1934. Two new species of sheep-like antelope from the Miocene of Mongolia. Am. Mus. Novit, no. 716:1-29. 1937. Siwalik antelopes and oxen in the American Museum of Natural History. Bull. Am. Mus. Nat. Hist. 72:729-874. 1939. The fossil Bovidae of India. Pal. Indica (N.S.) 26, 1:1-356. 1941. A fossil skull of Hemibos from Palestine. Ann. Mag. Nat. Hist. Lond. (11) 7:347-360.

572

Gentry

Pomel, A. 1893. Bubalus antiquus. Carte Geol. Alger., Paleont. Monogr. pp. 1-94. 1894a. Les Boeufs Taureaux. Carte Geol. Alger., Paleont. Monogr. pp. 1-108. 1894b. Les Boselaphes Ray. Carte Geol. Alger., Paleont. Monogr. pp. 1-61. 1895. Les Antilopes Pallas. Carte Geol. Alger., Paleont. Monogr. pp. 1-56. Reck, H. 1928. Pelorovis oldowayensis n.g., n.sp. Wiss. Ergebn. Oldoway-Exped. 1913 (N.F.) 3:57-67. 1935. Neue Genera aus der Oldoway-Fauna. Zentbl. Miner. Geol. Paläont., Ser. B, 1935, pp. 215-218. 1937. Thaleroceros radiciformis n.g., n.sp. Wiss. Ergebn. Oldoway-Exped. 1913 (N.F.) 4:137-142. Roberts, A. 1937. The old surviving types of mammals found in the Union. S. Afr. J. Sei. 34:73-88. Robinson, P. 1972. Pachytragus solignaci, a new species of caprine bovid from the late Miocene Beglia Formation of Tunisia. Notes Serv. Geol. Tunisie 37:73-94. Robinson, P., and C. C. Black. 1969. Note preliminaire sur les vertebres fossiles du Vindobonien. Notes Serv. Geol. Tunisie 31:67-70. Roman, F. 1931. Description de la faune pontique du Djerid (El Hamma et Nefta). In M. Solignac, Le Pontien dans le sud Tunisien. Awn. Univ. Lyon (N.S.) I 48:1-42. Romer, A. S. 1928. Pleistocene mammals of Algeria. Fauna of the paleolithic station of Mechta-el-Arbi. Bull. Logan Mus. 1:80-163. Savage, R. J. G., and W. R. Hamilton. 1973. Introduction to the Miocene mammal faunas of Gebel Zelten, Libya. Bull. Brit. Mus. Nat. Hist. (Geol.) 22:513-527. Schwarz, Ε. 1932. Neue diluviale Antilopen aus Ostafrika. Zentbl. Miner. Geol. Paläont. ser. B, 1932, pp. 1-4. 1937. Die fossilen Antilopen von Oldoway. Wiss. Ergebn. Oldoway-Exped. 1913 (N.F.) 4:8-104. Scott, W. B. 1907. A collection of fossil mammals from the coast of Zululand. Rep. Geol. Surv. Natal Zululand 3:253-262. Selley, R. C. 1969. Near-shore marine and continental sediments of the Sirte basin, Libya. Quart. J. Geol. Soc. Lond. 124:419-460. Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bull. Am. Mus. Nat. Hist. 85:1-350. Sokolov, I. I. 1952. Remains of Bovidae, Mammalia from lower Miocene deposits of western Gobi. Trudy paleont. Inst. Akad. Nauk. SSSR (Moscow) 41:155-158. Stromer, Ε. 1926. Reste Land- und Süsswasser-Bewohnender Wirbeltiere aus den Diamantfeldern Deutsch-

Bovidae

Südwestafrikas. In E. Kaiser, Die Diamantenwüste Südwest-afrikas, Berlin, vol. 2, pp. 107-153. Studer, T. 1898. Ueber fossile Knochen vom Wadi-Natrun Unteregypten. Mitt. Naturf. Ges. Bern 1460:72-77. Thenius, E. 1952. Die Boviden des steirischen Tertiärs. Sber. Ost. Akad. Wiss. Abt. I, 161:409-439. Thomas, P. 1884. Recherches stratigraphiques et paleontologiques sur quelques formations d'eau douce de 1'Algerie. Mem. Soc. Geol. Fr., ser. 3, 3(2): 1-51. Trofimov, Β. A. 1958. New Bovidae from the Oligocene of Central Asia. Vert. Palasiat. 2:244-247. Van Couvering, J. Α., and J. A. Miller. 1971. Late Miocene marine and nonmarine time scale in Europe. Nature 230:559-563. van Hoepen, E. C. N. 1932. Voorlopige beskrywing van Vrystaatse soogdiere. Paleont. Navors. Nas. Mus. Bloemfontein 2(5):63-65. Vrba, E. S. 1971. A new fossil alcelaphine (Artiodactyla: Bovidae) from Swartkrans. Ann. Transv. Mus. 27:59-82. 1973. Two species of Antidorcas Sundevall at Swartkrans (Mammalia, Bovidae). Ann. Transv. Mus. 28:287-352. 1975. Some evidence of chronology and palaeoecology of Sterkfontein, Swartkrans and Kromdraai from the fossil Bovidae. Nature 254:301-304. Wells, L. H. 1943. A further report on the Wonderwerk Cave, Kuruman. Section 2, Fauna. S. Afr. J. Sei. 40: 263-270. 1951. A large fossil klipspringer from Potgietersrust. S. Afr. J. Sei. 47:167-168. 1963. Note on a bovid fossil from the Pleistocene of Abu Hugar, Sudan. Ann. Mag. Nat. Hist. Lond. (13) 6:303-304. 1967. Antelopes in the Pleistocene of southern Africa. In W. W. Bishop and J. D. Clark, eds., Background to evolution in Africa. Chicago: Univ. of Chicago Press, pp. 99-107. 1969. Generic position of "Phenacotragus" vanhoepeni. S. Afr. J. Sei. 65:162-163. Wells, L. H., and Η. B. S. Cooke. 1956. Fossil Bovidae from the Limeworks Quarry, Makapansgat, Potgietersrust. Palaeont. Afr. 4:1-55. Whitworth, T. 1958. Miocene ruminants of East Africa. Brit. Mus. Nat. Hist., Fossil Mammals of Africa, no. 15, pp. 1-50. Zittel, K. A. 1925. Textbook of Palaeontology, vol. 3, Mammalia. Translated and revised by Arthur Smith Woodward, London.

573

28 Sirenia Daryl P. Domning

The order Sirenia is the only extant group of mammals adapted to feed exclusively on aquatic plants. In view of the worldwide abundance of aquatic macrophytes and the few other large herbivores competing for this resource, it is noteworthy that Recent sirenians comprise only three genera and five species. One of these, Steller's sea cow (Hydrodamalis gigas) of the North Pacific, was exterminated by man in the eighteenth century. Uniquely among sirenians, it was adapted to cold climates and a diet of kelp and other algae (Domning 1977). All the living sirenians are tropical forms that feed preferentially on angiosperms, and this appears to have been the primitive condition for the order. The Indian Ocean and West Pacific tropics are today inhabited by a single species, Dugong dugon, distributed in nearshore marine waters from East Africa and the Red Sea to Japan, Micronesia, and Australia. The three species of manatees (Trichechus) occur on both sides of the tropical Atlantic: T. manatus in fresh and salt water from the southeastern United States through the Caribbean to beyond the eastern tip of Brazil; T. senegalensis in rivers and coastal waters of West Africa; and T. inunguis in the Amazon Basin of South America (Bertram and Bertram 1973). The fossil record of sirenians is extensive but uneven in both geographic and taxonomic distribution, and many of the described genera, subfamilies, and families are monotypic. The order seems to have reached its peak diversity in the Miocene with about a dozen known genera. Sirenians first appear in the Eocene, and the abundance of sirenian remains in the middle and upper Eocene of North Africa early led to the labeling of that area as the center of origin of the group. This was corroborated by recognition of anatomical similarities between sirenians and other characteristically African groups, such as moeritheres, proboscideans, and hyracoids. However, the earliest known sirenian fossils are from the lower Eocene of Hungary (Kretzoi 1953), and the presently known distribution of Eocene sirenians from the Caribbean to the East Indies (Reinhart 1976) points to the shores of the Tethyan Seaway, probably in the Old World, as the specification of choice for a center of origin. Significantly, the marine angiosperms also show signs of an originally Tethyan distribution, although they entered the water during the Cretaceous, well before the appearance of sirenians (den Hartog 1970). The last appearing (late Oligocene) group of "subungulates," the desmostylians, were formerly included within the Sirenia but are now accorded ordinal status in close alliance with moeritheres,

574

Domning

proboscideans, and sirenians. These extinct marine herbivores, of possible eastern Tethyan origin, were apparently restricted to the North Pacific except for a limited foray into the Caribbean (Reinhart 1976). All of the so-called subungulates presumably share a Paleocene condylarth ancestry. I provisionally follow Sickenberg (1934) and Simpson (1945) in recognizing four families of sirenians: Prorastomidae, Protosirenidae, Trichechidae, and Dugongidae. The former two are known only from middle Eocene deposits; the manatees have a scanty Miocene to Recent record; while the dugongids, comprising the majority of known forms, are fairly well documented from the middle Eocene to the present and are divided into several subfamilies. Prorastomus from Jamaica shows resemblances to condylarths in its ear region and atlas; however, the inflated rostrum, pachyostotic skull, and bilophodont molars are shared with later sirenians (Savage 1977). Sirenavus from the middle Eocene of Hungary has also been questionably referred to the Prorastomidae (Kretzoi 1941). Protosiren from Egypt is more advanced and in most ways a good structural ancestor for the other two families, although too late in time to be the actual ancestor. The trichechids are a small, conservative group apparently confined to South America from the Eocene until their dispersal to North America and Africa in the Quaternary. The dugongids, the most successful family, maintained a pantropical distribution throughout the Tertiary. Their main stem is generally considered to be the subfamily Halitheriinae (middle Eocene to Pliocene, pantropical), which gave rise to at least two specialized side branches, the Hydrodamalinae (middle Miocene to Recent, North Pacific) and the Dugonginae (Recent, Indopacific; fossil record lacking). I regard Rytiodus (early Miocene, Europe and Africa) as referable to the Halitheriinae rather than to a subfamily of its own, and I divide Reinhart's (1959) "Halianassinae" between the Halitheriinae and Hydrodamalinae. Miosiren (middle Miocene, Europe) is an aberrant, possibly molluscivorous form placed in its own subfamily (Miosireninae) within the Dugongidae, though separate descent from protosirenids may be an alternate possibility.

Structure The Sirenia, together with the Cetacea, are the only obligatorily aquatic mammals, sharing fusiform bodies, finlike forelimbs, horizontally expanded tail fins, and loss of hind limbs. The latter

Sirenia

were reduced and may have been nonfunctional even in Protosiren. Unlike many cetaceans, however, sirenians lack dorsal fins and show no aptitude or adaptations for echolocation. As slow-swimming herbivores, they have refined the art of hydrostasis to a unique degree in order to minimize energy expenditure in locomotion, evolving massive pachyostotic skeletons for ballast and horizontal diaphragms, elongate lungs, and monopodially-branching bronchial trees (Pick 1907) for maintaining fore-and-aft trim. Muscular contraction of the thorax evidently can adjust specific gravity at will and permit silent, vertical submergence. In conjunction with these apparent adaptations for statically keeping the body axis close to a horizontal position, the degree of downward flexion of the rostrum and anterior mandible is adjusted to the feeding habits of the species: Dugong, with a strongly downturned snout, is adapted to habitats (sea grass meadows) where the dominant growth habit of plants is short and bottom hugging, whereas sirenians with slight snout deflections are found in habitats with abundant floating or other near-surface vegetation (Domning 1977). The earliest sirenians, including Prorastomus (Savage 1977), Protosiren, and Eotheroides (Sicken3 15 3 berg 1934), had a dentition consisting of g'^'g'g · This soon became reduced, first by loss of P5 and retention of DP5 in the adult dentition, then by progressive loss of the anterior permanent and deciduous premolars and the canines and incisors. One pair of upper incisors, apparently the first, is retained as short tusks in many dugongids; their degree of development is at least sometimes (as in Dugong) sexually dimorphic; but the sexes never seem to differ to the extent of presence versus complete absence of tusks. In all sirenians the fronts of the upper and lower jaws bear tough pads that serve with the prehensile upper lip to pull food into the mouth, with or without the help of the flippers. While halitheriines always retain an adult cheek dentition of at least M2-3 (more commonly DP5-M3 in Neogene forms) and the teeth retain the primitive bunobilophodont-brachydont condition, the other dugongid subfamilies show more specialization: M3 of Miosiren is reduced to a simple, stout peg; adult Dugong retain only M2-3, lacking roots, enamel, and (after initial wear) cusps; and Hydrodamalis was completely toothless. Trichechids, in contrast, after reaching a halitheriine-like degree of reduction by the late Miocene, evolved a unique adaptation to increased tooth wear by continuing to produce supernumerary brachydont, bilophodont molars at the

Domning

rear of each toothrow. The entire row migrates forward, worn teeth falling out at the front and new ones replacing them at the rear, throughout the animal's life. This process has often been misleadingly compared to tooth replacement in elephants; but while forward tooth movement or "mesial drift" occurs to some degree in many mammals, the combination of this process with unlimited production of supernumerary teeth is unique to the advanced trichechids (Ribodon and Trichechus). Trichechus also differs from dugongids in having a rounded rather than whalelike tail fin (probably a primitive feature), in having only six cervical vertebrae, and in the structure of the shoulder joint and pelvis, features suggesting a long independent history. The most common sirenian fossils are fragments of the swollen, pachyostotic ribs (usually indeterminable), vertebrae, and skullcaps (fused parietals and supraoccipital). The latter have, faute de mieux, been much used in sirenian systematics, but I consider their diagnostic value very limited. Isolated teeth are also difficult to interpret. Indeed, so morphologically similar are most sirenians that precise identification may be impossible without relatively large portions of skulls or mandibles. Considerable numbers of high-quality specimens will be necessary to elucidate the relations between the many monotypic forms known and to sort out the apparently frequent instances of parallel evolution in sirenian history.

Fossil Sirenia in Africa The earliest geological record of African sirenians is that ofEotheroides aegyptiacum (Owen 1875) from the middle Eocene Nummulitic Beds of the Mokattam Hills near Cairo. Later discoveries were predominantly from the late Eocene marine beds of the Fayum. With the exception of three Miocene occurrences of "Halitherium sp."—from the Isthmus of Suez (Gervais 1872), Madagascar (Collignon and Cottreau 1927), and the Congo (Dartevelle 1935)— the known distribution of fossil sirenians in Africa remained restricted to the Egyptian Eocene until the reports of Pliocene material from Morocco (Ennouchi 1954), Eocene, Oligocene, and Miocene material from Libya (Savage and White 1965; Savage 1967, 1969, 1971), and Eocene and Oligocene remains from Somalia (Savage 1969). Fragmentary remains are now also known from the Oligocene and Miocene of Tunisia. Although the Egyptian material has been thoroughly described, the sirenian history of the rest of the continent has barely begun to be

Sirenia

575

investigated. A synopsis of the published African material follows. Order Sirenia Illiger 1811 Family Protosirenidae Sickenberg 1934 Genus Protosiren Abel 1904 This family appears to contain but a single species, P. fraasi Abel 1904, from the middle Eocene (basal Mokattam Formation) of Egypt. P. dolloi Abel 1904, from the upper Eocene of Italy, later referred to Mesosiren Abel 1906, has been most recently (Simpson 1945) considered synonymous with the dugongid Prototherium veronense. Sickenberg (1934) suggested the name ?Protosiren dubia for teeth from the middle Eocene of France, but their affinities are indeed dubious. Protosiren fraasi, after Prorastomus the most primitive known sirenian, is described by Sickenberg (1934) as having five (or possibly six) singlerooted premolars anterior to the three molars. Unlike later forms, it possesses an alisphenoid canal but shows a clear advance over Prorastomus in that the periotic is no longer fused to the rest of the skull (Savage 1977). A descending process of the frontal (the lamina orbitalis) forms part of the wall of the orbit, a feature seen also in trichechids and in Miosiren. The inside of the braincase roof, however, lacks a bony falx cerebri, and this probably derived condition together with Protosiren's near-contemporaneity with the first dugongids seems to rule it out of direct ancestry to the later families. Since, to my knowledge, Protosiren shares no derived characters with either modern family, I favor retaining the Protosirenidae as a primitive stem group from which both could have been derived. Family Dugongidae Gray 1821 Subfamily Halitheriinae Abel 1913 Genus Eotheroides Palmer 1899 SYNONYMY. Eotherium Owen 1875, nec Leidy 1853, Eosiren Andrews 1902, Archaeosiren Abel 1913, Masrisiren Kretzoi 1941. TYPE SPECIES. Eotherium aegyptiacum Owen 1875. Eotheroides contains five nominal species from the Eocene of Egypt: E. aegyptiacum (Owen 1875) (including "Manatus" coulombi Filhol 1878), E. abeli Sickenberg 1934, and "Eotherium" majus Zdansky 1938 from the middle Eocene Mokattam Formation (above the Protosiren horizon; Kellogg 1936, p. 235),

576

Domning

and E. libyca (Andrews 1902) and E. stromeri (Abel 1913) from the late Eocene Qasr el Sagha Formation of the Fayum. CEotherium markgrafi" Abel 1913, p. 337 is a nomen nudum.) The genus may also occur in southern Europe (Sickenberg 1934; Richard 1946; Fuchs 1973) along with its close late Eocene relative, Prototherium de Zigno 1887; the latter, however, has not been reported from Africa. All these names are in need of reexamination and some should probably be placed in synonymy. The specimen of E. majus in particular, consisting only of an isolated M 2 , is probably referable to one of the other species, as is the "Eotheroid.es sp. indet." skull from the Qasr el Sagha Formation described by Reinhart (1959). Eotheroides is the oldest and most primitive dugongid known and is probably generalized enough to have given rise to all the later forms. Its rostrum is moderately deflected and bears a pair of small tusks. Five premolar teeth are present, but DP5 does not appear to have still had a permanent replacement (Sickenberg 1934). The tooth cusp morphology is typical of Paleogene dugongids. The permanent premolars are single-rooted with one major cusp surrounded by a low multicuspate cingulum. Upper molariform teeth have a relatively straight protoloph composed of protocone, protoconule, and paracone; an unobstructed transverse valley; a straight or convex-forward metaloph formed by hypocone, hypoconule, and metacone; and small, variably cuspate anterior and posterior cingula. Lower molariform teeth similarly have two lophs slightly oblique to the tooth's axis, a distinct crista obliqua intruding on the transverse valley, a well-developed hypoconulid lophule (largest on M 3 ), and no anterior cingulum. The skull is principally marked by characters primitive for sirenians in general: expanded supraorbital processes; well-developed nasals and lacrimals; jugals with greatest ventral expansion posterior to orbit; prominent sigmoid ridge on posterior flank of squamosal; unexpanded dorsolateral rim of exoccipital; slender horizontal mandibular ramus; and several accessory foramina posterior to the large mental foramen. External hind limbs were completely lacking (Siegfried 1967). A new genus of primitive sirenian, possibly allied to Eotheroides, has been discovered in middle Eocene beds at Bu el Haderait, Libya, and is being described by G. Heal (Savage 1977). Genus Halitherium Kaup 1838 S Y N O N Y M Y . Pugmeodon Kaup 1834 (nomen oblitum), Manatherium Hartlaub 1886. T Y P E SPECIES. Hippopotamus dubius G. Cuvier 1824.

Sirenia

Halitherium is the common Oligocene sirenian of Europe (Lepsius 1882; Spillmann 1959) and also occurs in North America (Reinhart 1976). As the senior available generic name proposed for a fossil sirenian, Halitherium has also served in its time as a wastebasket name for a variety of Eocene to Miocene specimens, many fragmentary and indeterminable. At least seven nominal species are currently in the literature, and many more names have been erected and consigned to various degrees of oblivion; thorough revision is badly needed. Halitherium can be easily derived from dugongids of the Eotheroides type; apart from a greater snout deflection, larger tusks, loss of the most anterior premolars, and some reduction of the nasals and lacrimals, there is little to distinguish the two genera. H. christoli from the Austrian upper Oligocene (Spillmann 1959), with a somewhat deeper mandible, could be transitional to Miocene forms of the Metaxytherium grade. Only three specimens referred to Halitherium sp. have been reported from the African region. None of these, however, is sufficiently complete to justify generic assignment, and their post-Oligocene ages make assignment to Halitherium additionally questionable. Gervais (1872:341) records ribs, cf Halitherium, from the Carcharodon megalodon beds at Chalouf ( = El Shallüfa?), Isthmus of Suez, received by the Paris Museum. He also (p. 352) alludes to other ribs from Lower Egypt cited by de Blainville. In fact, de Blainville (1840:43, 51) was originally inclined to refer the fragmentary vertebrae and ribs in question to a pinniped, but later (1844:119-120) concluded that they were sirenian. These remains, from the "left bank of the Nile valley" and of uncertain age, appear to be the earliest recorded sirenian remains from Africa. The second Halitherium sp. deserves mention for its Zoogeographie interest. A fragmentary skull and skeleton were recorded from the island of Makamby on the northwest coast of Madagascar by Collignon and Cottreau (1927), who considered the deposits Miocene (Burdigalian to Helvetian). They compared the specimen with European Halitherium, pointing out that the latest known of the latter was H. bellunense from the basal Burdigalian of Italy. The Madagascar skullcap is elongated, with temporal crests meeting in the midline, suggesting the stage of evolution represented by Halitherium. This record at least establishes the presence in the Indopacific region of a Halitherium-like form from which Dugong might have been derived, but connecting links are lacking.

Domning

Dartevelle (1935) recorded Halitherium^) sp. from beds of probable Burdigalian age at Malembe, Congo, but as the material consisted only of ribs the only tenable identification is Sirenia indet. Genus Rytiodus Lartet 1866 TYPE SPECIES. Rytiodus capgrandi Lartet 1866. This poorly known dugongid was described on the basis of tusks and other fragments from the Aquitanian of France. A fragmentary skeleton was later reported by Delfortrie (1880), but no further material has come to light in Europe. The genus is chiefly notable for its flat, bladelike tusks, relatively large size (estimated at nearly 5 m), and (as restored by Delfortrie) straight, undeflected rostrum. That the latter interpretation is erroneous, however, was apparent from Delfortrie's own illustration (1880, pi. 5), which clearly shows a pronounced deflection of the anterior end of the maxilla, indicating a strongly downturned snout. This was confirmed by the discovery of an intact skull and other remains in the lower Miocene at Gebel Zelten, Libya, which are being described as a new species of Rytiodus by G. Heal (pers. comm.). The rostrum is indeed very sharply deflected, adding one more genus to the roster of dugongids apparently specialized for bottom feeding. Rytiodus was evidently an offshoot of Oligocene Halitherium, but its specializations do not impress me as warranting the subfamilial distinction usually accorded to it. Genus Metaxytherium de Christol 1840 SYNONYMY. Cheirotherium Bruno 1839, nec Kaup 1835, Fucotherium Kaup 1840, Pontotherium Kaup 1840, Felsinotherium Capellini 1871, Halysiren Kretzoi 1941. TYPE SPECIES. Hippopotamus medius Desmarest 1822.

As Halitherium has often served as a wastebasket term, usually connoting "Oligocene dugongid," Metaxytherium and Felsinotherium have served the same function for the Miocene and Pliocene respectively. The virtual impossibility of separating the latter two nominal taxa has long been acknowledged, and even Capellini himself regarded them as synonymous; but only recently has Felsinotherium been dropped from usage (Fondi and Pacini 1974; Reinhart 1976). The synonymy of Metaxytherium and Halianassa Studer 1887, to which many nominal species of Metaxytherium have at times been assigned, is a more complex problem, summarized by Kellogg (1966); the holotype of Halianassa may in fact pertain to yet another form, Thalattosiren (see

Sirenia

577

Thenius 1952). Clearly the Neogene halitheriines are as badly in need of revision as the Paleogene ones; at least a dozen nominal species of Metaxytherium are still recognized. As noted above, Eotheroides, Halitherium, and Metaxytherium seem to form a straightforward sequence of structural stages, although the details on the specific level remain to be worked out. The European Metaxytherium continued the tradition of tusked dugongids with moderately deflected snouts, but with derived characters such as reduced supraorbital processes, nasals, and lacrimals, more anterior suborbital expansion of the jugals, distinct and slightly inturned processi retroversi of the squamosals, deeper mandibles, and absence of accessory mental foramina. The adult cheek dentition is typically reduced to DP5-M3, and the cusp patterns are somewhat more complex than in Halitherium. The upper teeth generally feature partial obstruction of the transverse valley by the metaconule, while the lophs of the lowers tend to be more crescentic. North American Metaxytherium, which may well deserve separate generic status, generally lacked tusks and, at least in the Pliocene, developed quite sharply downturned snouts. This genus has so far been encountered twice in Africa. A partial skull of Metaxytherium sp. indet. is known from the lower Miocene of Gebel Zelten, Libya (Savage pers. comm.). A lower molar and a skullcap from the Pliocene of Dar bei Hamri, Morocco, were referred to "Felsinotherium" cf serresi by Ennouchi (1954). M. serresi (Gervais 1859) is a contemporary form from France (Deperet and Roman 1920), and the identification is reasonable. This record serves principally to underline the not surprising fact that both sides of the Mediterranean shared a common sirenian fauna.

Indeterminable Sirenian Remains At this point may be listed several African occurrences of Tertiary sirenians that, unfortunately, do not yet include diagnostic material. Eocene: Isolated cheek teeth and other remains have been collected at Mogadishu and at Callis and Carcar, Somalia (Savage 1977), and further fragments in the middle Eocene Nautilus Beds at Daban, Somali Republic (MacFayden in Haas and Miller 1952). Upper Eocene beds at Dor el Talha, Libya, have also yielded rib fragments (Savage 1971). Oligocene: A tusk, possibly sirenian, has been found at Bedeil, Somali Republic (Savage 1969). A skeleton has been found in the Fortuna Sandstone at

578

Domning

Sirenia

Djebel ech Cherichira, Tunisia (P. Robinson, pers. comm. to G. Heal). Miocene: Fragments have been collected from both the Hipparion and the pre-Hipparion levels of the Vindobonian Beglia Formation, Bled ed Douarah, Tunisia, by Robinson (pers. comm. to G. Heal; Robinson and Black 1969).

anatomically the most distinct of the three is not the West African but the South American, suggest that the African form is a very recent immigrant from the New World. Hatt (1934) provides a useful summary of the anatomy and distribution of the African species T. senegalensis Link 1795.

Subfamily Dugonginae Simpson 1932

Phylogeny and Evolution

Genus Dugong Lacepede 1799 SYNONYMY. Platystomus Fischer v. Waldheim 1803, nec Platystoma Meigen 1803, Amblychilus Fischer v. Waldheim 1814, Dugungus Tiedemann 1808, Halicore Illiger 1811. TYPE SPECIES. Trichechus dugon Müller 1776. Dugong dugon (Müller) 1776 has, as noted above, no known fossil record, and the few fragmentary fossil sirenians known from the Indopacific region have shed no real light on its history. The dugong, however, is reasonably viewed as a descendant of halitheriines that became isolated in the eastern remnant of Tethys during the Miocene and specialized for feeding on the extensive sea grass meadows of the Indopacific. It is anatomically quite conservative, apart from minor derived characters such as narrow supraorbital processes and strongly inflected Processi retroversi of the squamosals and the obvious dental specializations (adult cheek teeth reduced to M2-3, lacking roots and enamel). Populations in different parts of the Indopacific, once distinguished by specific names, are now all regarded as but a single species. A convenient summary of the literature on Dugong is provided by Husar (1975). Dugongs and manatees (see below) have been traditionally hunted by man (see, for example, Petit 1927), and archaeologists working at coastal sites in tropical Africa should expect to encounter sirenian bones (for illustrations see Kaiser 1974).

Much has already been said about the phylogeny of the sirenians discussed in this chapter; the general relationships are summarized in figure 28.1. It must be noted that sirenian paleontology to date has been purely descriptive, and actual documentation of phylogeny—as opposed to mere considerations of "stage of evolution"—is only beginning to become possible. As the record becomes more completely known, the ecology as well as the cladistic relationships of fossil sirenians can be clarified, and the sirenian record can begin to contribute ideas and data to other areas of paleontological interpretation. But much basic taxonomic work remains to be done; the many described species of genera such as Eotheroides, Halitherium, and Metaxytherium doubtless require some lumping, and the genera themselves may need to be split in order to better reflect their kinship to some of the monotypic forms. Sirenian diversity seems to have been constrained by the lack of diversity in suitable marine food plants and habitats, which made niche partitioning difficult, and by the vagility of sirenians themselves, which minimized opportunities for geographic isolation of populations. The pattern that seems to be emerging is one of repeated parallel evolution of adaptations for a very small number of possible niches, e.g., bottom feeding in forms with strongly deflected snouts such as in Rytiodus, Thalattosiren, Dugong, and some Metaxytherium. Such parallelism, of course, makes the record all the more difficult to unravel. Africa's role in sirenian evolution has been very limited due to that continent's relative emergence throughout the Cenozoic. Apart from whatever importance the south shore of Tethys may have had as the, or a, scene of original entry of the water by sirenian ancestors, the main events of sirenian history seem to have been played out in areas peripheral to Africa—notably Europe and the Caribbean, as the Indopacific record is still inadequate. The newly discovered Eocene to Miocene sirenian-bearing deposits in Africa, however, could fill some of the intriguing gaps in the record preserved in the more thoroughly explored parts of the Tethyan realm.

Family Trichechidae Gill 1872 Genus Trichechus Linnaeus 1758, nec 1766 SYNONYMY. Manatus Brünnich 1772, Oxystomus Fischer v. Waldheim 1803, Neodermus Rafinesque 1815, Halipaedisca Gistel 1848. TYPE SPECIES. Trichechus manatus Linnaeus 1758. What little Tertiary record there is of manatees is restricted to South America. They appear in North America only in the Pleistocene and lack any fossil record at all in Africa. The close similarities among the three living species, however, and the fact that

φ ft

"Ö α>

τί ö φ β Ö φ Λ ο S-H « Β* 13 CO Ö

>>

Λ ft -β Φ α

ce Ö Ο Til

• FH

φ bp ö iS US

Ο CO

CO '"ö ö jzj3 cd

'S .0 Φ s

ϊ>> β öc Φ ω ο ο ,μ >> rC ' ΌΦ

ft I Μ

Ν .S s l Ι ί

fa .2

580

Domning

References Abel, O. 1913. Die eozänen Sirenen der Mittelmeerregion. Erster Teil: Der Schädel von Eotherium aegyptiacum. Paläontogr. 59:289-360. Andrews, C. W. 1902. Preliminary note on some recently discovered vertebrates from Egypt (part III). Geol. Mag. (4) 9:291-295. Bertram, G. C. L., and Bertram, C. K. R. 1973. The modern Sirenia: Their distribution and status. Biol. J. Linn. Soc. Lond. 5(4):297-338. Blainville, Η. M. D. de. 1840. Osteographie, Livr. 7, Des Phoques (G. Phoca, L). Paris. 1844. Osteographie, Livr. 15, Des Lamantins 4(Buffon), (Manatus, Scopoli), ou gravigrades aquatiques. Paris. Collignon, M., and Cottreau, J. 1927. Paleontologie de Madagascar. XIV. Fossiles du Miocene marin. Annls. Paleont. 16(4): 135-171. Dartevelle, E. 1935. Les premiers restes de mammiferes du tertiaire du Congo: la faune Miocene de Malembe. C. R. Congr. Sei. Bruxelles 1:715-720. Delfortrie, E. 1880. Decouverte d'un squelette entier de Rytiodus dans le falun Aquitanien. Actes Soc. Linn. Bordeaux 34:131-144. Deperet, C., and Roman, F. 1920. Le Felsinotherium serresi des sables pliocenes de Montpellier et les rameaux phyletiques des sireniens fossiles de l'Ancien Monde. Arch. Mus. Hist. Nat. Lyon 12(4):l-56. Domning, D. P. 1977. An ecological model for late Tertiary sirenian evolution in the North Pacific Ocean. Syst. Zool. 25(4):352-362. Ennouchi, E. 1954. Un sirenien, Felsinotherium cf serresi, ä Dar bei Hamri. Notes Serv. Geol. Maroc 9 (Notes et Mem. no. 121):77-82. Filhol, H. 1878. Note sur la decouverte d'un nouveau mammifere marin (Manatus coulombi) en Afrique, dans les carrieres de Mokattam pres du Caire. Bull. Soc. Phil. Paris (7)2:124-125. Fondi, R., and Pacini, P. 1974. Nuovi resti di Sirenide dal Pliocene antico della Provincia di Siena. Palaeontogr. Ital. 67(n.s. 37):37-53. Fuchs, Η. 1973. Contributions ä l'etude des sireniens fossiles du bassin de la Transylvanie. (IV) Sur un fragment d'humerus de Cheia Baciului (Cluj). Studia Univ. BabesBolyai, Ser. Geol.-Mineral. 18(2):71-77. Gervais, P. 1859. Zoologie et paleontologie frangaises. 2nd ed. Paris. 1872. Travaux recentes sur les sirenides vivants et fossiles. Jour, de Zool. 1:332-353. Haas, O., and Miller, A. K. 1952. Eocene nautiloids of British Somaliland. Bull. Am. Mus. Nat. Hist. 99:313-354. Hartog, C. den. 1970. The sea-grasses of the world. Verh. kon. Nederl. Akad. Wet., Afd. Natuurk. (2)59(1): 1-275. Hatt, R. T. 1934. A manatee collected by the American Museum Congo expedition, with observations on the Recent manatees. Bull. Am. Mus. Nat. Hist. 66(4): 533-566. Husar, S. L. 1975. A review of the literature of the dugong

Sirenia

CDugong dugon). U.S. Fish & Wildl. Serv., Wildl. Res. Rept. 4:1-30. Kaiser, Η. E. 1974. Morphology of the Sirenia: a macroscopic and x-ray atlas of the osteology of Recent species. New York: S. Karger. Kellogg, R. 1936. A review of the Archaeoceti. Carnegie Inst. Wash. Puhl. 482:1-366. 1966. Fossil marine mammals from the Miocene Calvert Formation of Maryland and Virginia. 3. New species of extinct Miocene Sirenia. Bull. U.S. Nat. Mus. 247(3):65-98. Kretzoi, M. 1941. Sirenavus hungaricus n.g., n.sp., ein neuer Prorastomide aus dem Mitteleozän (Lutetium) von Felsögalla in Ungarn. Ann. Hist.-Nat. Mus. Nat. Hungarici (Min. Geol. Pal.) 34:146-156. 1953. Le plus ancien vestige fossile de mammifere en Hongrie. Földt. Közlöny 83(7-9):273-277. Lartet, E. 1866. Note sur deux nouveaux Sireniens fossiles des terrains tertiaires du bassin de la Garonne. Bull. Soc. Geol. Fr. (2)23:673-686. Lepsius, G. R. 1882. Halitherium schinzi, die fossile Sirene des Mainzer Beckens. Abh. Mittelrhein. Geol. Ver. l:vi + 200. Owen, R. 1875. On fossil evidences of a sirenian mammal (Eotherium aegyptiacum, Owen) from the Nummulitic Eocene of the Mokattam Cliffs, near Cairo. Quart. J. Geol. Soc. London 31(1):100-105. Petit, G. 1927. Nouvelles observations sur la peche rituelle du dugong ä Madagascar. Bull. Mem. Soc. d'Anthrop. Paris (7)8:246-250. Pick, F. K. 1907. Zur feineren Anatomie der Lunge von Halicore dugong. Arch. f. Naturgesch. 73(l)(2):245-272. Reinhart, R. H. 1959. A review of the Sirenia and Desmostylia. Univ. Calif. Publ. Geol. Sei. 36(1): 1-146. 1976. Fossil sirenians and desmostylids from Florida and elsewhere. Bull. Fla. St. Mus. Biol. Sei. 20(4):187-300. Richard, M. 1946. Les gisements de mammiferes tertiaires d'Aquitaine. Mem. Soc. Geol. Fr. 52:xxiv + 380. Robinson, P., and Black, C. C. 1969. Note preliminaire sur les vertebres fossiles du Vindobonien (formation Beglia), du Bled Douarah, Gouvernorat de Gafsa, Tunisie. Notes Serv. Geol. Tunisie 31:67-70. Savage, R. J. G. 1967. Early Miocene mammal faunas of the Tethyan region. Syst. Assoc. Publ. 7:247-282. 1969. Early Tertiary mammal locality in southern Libya. Proc. Geol. Soc. Lond. 1657:167-171. 1971. Review of the fossil mammals of Libya. In "Symposium on the geology of Libya," University of Libya, pp. 215-225. 1975. Prorastomus and new early Tertiary sirenians from North Africa. Amer. Zool. 15(3):824. 1976 (1977). Review of early Sirenia. Syst. Zool. 25(4):344-351. Savage, R. J. G., and White, Μ. E. 1965. Exhibit: Two mammal faunas from the early Tertiary of central Libya. Proc. Geol. Soc. Lond. 1623:89-91. Sickenberg, Ο. 1934. Beiträge zur Kenntnis tertiärer Sirenen. Mem. Mus. Roy. Hist. Nat. Belgique 63:1-352.

Domning

Siegfried, P. 1967. Das Femur von Eotheroides libyca (Owen) (Sirenia). Paläont. Zeits. 41(3/4): 165-172. Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bull. Am. Mus. Nat. Hist. 85:xvi + 350. Spillmann, F. 1959. Die Sirenen aus dem Oligozän des Linzer Beckens (Oberösterreich), mit Ausführungen über "Osteosklerose" und "Pachyostose." Oesterr. Akad. Wiss., math.-nat. Kl., Denkschr. 110(3):l-68.

Sirenia

581

Thenius, E. 1952. Die Säugetierfauna aus dem Torton von Neudorf an der March (CSR). Neues Jb. Geol. Pal., Abh. 96(1):27-136. Zdansky, O. 1938. Eotherium majus sp.n., eine neue Sirene aus dem Mitteleozän von Aegypten. Palaeobiologica 6(2):429-434.

The majority of fossil whale species reported from Africa were collected in Eocene rocks of Egypt and belong in the primitive cetacean suborder Archaeoceti. Most were collected or described in the scientific literature before or around the first part of the Twentieth Century, and some of the primary literature is out of date. Recent publications dealing with specimens from India (Sahni and Mishra 1972,1975; Satsangi and Mukhopadhyay 1975) have probably not referred extensively enough to the literature on north African Tethys fossils. The studies by Dames, Fraas, Stromer, and Andrews have described most of the archaeocetes of the world, the majority of which were found in Egypt. These species have figured prominently in theories on the origin and evolution of the order Cetacea. In marked contrast, the postEocene cetacean record of Africa is scanty and undoubtedly biased by a lack of field work in appropriate rock units. A discussion of the fossil whales, dolphins, and porpoises of Africa must consider that in life these animals, the fossil remains of which are now found in rocks on the African continent, were inhabitants of the marine waters surrounding it and the fresh water on it. Whether these cetaceans were fluvial, near-shore, or pelagic, their distribution was not necessarily influenced by the same dietary, climatic, or geographic constraints as was the dispersal of the terrestrial animals of Africa. Cetacean systematics is sometimes biased by the land mass on which fossils are found, without full consideration of the ancient ocean basin in which they lived. Therefore, distributions of fossil marine mammals (listed in summary works like Simpson 1945; Romer 1966) should more correctly be recorded as Eocene of Tethys instead of Eocene of Egypt, or Miocene of North Atlantic instead of Miocene of Europe. The fossil record of cetaceans is now too incomplete to recognize and document precise, limited distributions. The safest initial assumption is that a species was probably distributed throughout an ocean basin or water mass rather than having a restricted distribution near a continental margin. The orientation of this essay must be termed "gradistic," as the limited fossil evidence is insufficiently documented and interpreted at present to be quantified or subjected to cladistic analyses.

Some Morphological Specializations of Cetacea The following characterization is based on many general and summary works on the history and morphology of the Cetacea, especially those of Miller

Barnes and Mitchell

(1923), Kellogg (1928a,b, 1936, 1938), Slijper (1936, 1962), Simpson (1945), Tomilin (1957,1967), Dechaseaux (1961), Rice (1967), and Trofimov and Gromova (1962, 1968). Living cetaceans are obligative aquatic mammals, having lost efficient means of locomotion on land. Their anatomy and physiology are modified for moving, feeding, and reproducing in water and breathing at the surface. They are one of the most highly modified and specialized of the mammalian orders. The neck is generally short, sometimes with fused vertebrae, and the body is fusiform. Hair is present in modern Cetacea in the form of isolated hairs in patterns on the head and jaws and chin of many species. The presence of hair on early archaeocetes is debatable but entirely possible. Horizontal caudal flukes supported by fibro-cartilage are present on all Recent Cetacea and presumably on extinct species. Fossil skeletons show the change in proportion and shape of the terminal caudal vertebrae that in modern species is associated with the base of the caudal fluke. A single, median dorsal fin is present in most but not all Recent species and may have been secondarily lost in various lineages. The forelimb is flattened into a flipper with the bony digits enclosed within it. Blubber is usually well developed in modern species. Evolutionary trends have been toward the development of a "telescoped" skull, in which the bones of the rostrum extend over and under the cranium. This is likely the result of changes in cranial kinetics associated with specialized feeding habits. The narial opening moved from the typical, anterior position of most mammals to a dorsoposterior position, apparently associated with the need to breathe quickly and efficiently at the air-water interface. In many species the middle ear air sinus extends into various parts of the ventral surface of the skull. Three suborders have long been recognized. The primitive, extinct suborder Archaeoceti has skulls with the bones not telescoped, but retaining the usual mammalian relationships to each other. The external narial opening is at midpoint on the skull or further forward. The intertemporal region is long and narrow, and the lambdoidal crests are large. The full eutherian dental formula is present, with loss of Μ 3 in some species. The posterior cheek teeth are multiple rooted with pronounced cusps. The living suborder Mysticeti, the baleen whales, have vestigial teeth, generally lost in utero, and baleen in the adult. Baleen is a neomorph, comprised of fused fibrous tubules of epidermal origin rooted in and suspended from the palate. The fringe of the baleen becomes frayed and interwoven, forming a sieve which living species use to capture food,

Cetacea

583

mainly fish and small crustaceans. The rostral portions of the maxillae are extended laterally in a thin plate to support the baleen, and the cranial portions do not extend far posteriorly onto the cranium. The occipital shield extends dorsoanteriorly over the cranium and in some species shortens the length of the primitive interorbital constriction, thereby changing the size of the temporal fossae. The external bony naris is anterior to the orbit and the external fleshy nostrils are paired in all mysticetes. The living suborder Odontoceti, the toothed whales, have multiple to single-rooted teeth with complex to simple conical crowns. Some species have greatly reduced functional dentition in the adult. The rostral portions of the maxillae are thick and not greatly extended laterally, and in most families they are extended posteriorly over the front of the cranium nearly to, or touching, the occipital shield. The bony external nares migrate to a dorsal position at the level of the orbits. The occipital shield does not extend far dorsoanteriorly over the cranium. The primitive intertemporal constriction is obliterated by the rostral and frontal bones. The external fleshy nostril is a single opening in living odontocetes. Some of the characters above are not reflected in bones, and their existence in fossil species can only be implied. There are, however, sufficient osteological characters given to separate the three suborders.

The Fossil Cetacea of Africa The fossil cetaceans that have been reported from the continent of Africa are neither numerous nor taxonomically diverse. They do however include specimens and taxa that have been extremely important in studies of cetacean evolution. Most African cetacean fossils are Eocene Archaeoceti. PostEocene fossils are more scanty and not well studied. The potential exists for further exploration to obtain new fossils, and for much interpretation. Part of the reason for this incomplete record of African cetacean evolution results from the nearly continuous history of uplift of the African continent. Marine transgressions have been relatively small in area, shallow, and short-lived. Most African cetacean fossils have been found in near-shore and even estuarine and freshwater facies. This precludes the systematic sampling of truly pelagic cetacean assemblages but does not obviate the finding of pelagic species in near-shore facies. On the other hand, it allows unusually precise correlations with chronologies based upon terrestrial vertebrates. Fossils of African cetaceans occur in Eocene, Oli-

584

Barnes and Mitchell

gocene, Miocene, and Pliocene rocks. Eocene marine transgressions left bands of marine deposits on the African continental margin. There were large embayments from the Atlantic Ocean in Nigeria and from Tethys in northern Africa. Archaeocete fossils have been discovered in the former Atlantic embayment represented by the middle Eocene Ameki Formation (Andrews 1920; Van Valen 1968). Elouard (1966) has reported archaeocete fossils from near Kaolack in Senegal. In Libya and Egypt, the Tethys Sea left a regressive marine sequence of middle through late Eocene rocks containing archaeocete fossils. These fossils record the major evolutionary features of the group. They are associated with other marine organisms and in some cases with terrestrial or freshwater animals, allowing extensive interpretations of paleoecology and chronology. The geology of Eocene rocks in Egypt has been described by Beadnell (1905), Andrews (1906), Kellogg (1936), Said (1962), and Simons (1968). These fossils have been collected and studied periodically since the turn of the twentieth century. Kellogg, Andrews, and Simons summarize the research on earlier collections. Specimens were collected by expeditions from Germany (including Richard Markgraf), the Egyptian Geological Survey (under Beadnell), the British Museum (Natural History) (Andrews 1901a,b, 1907b; Andrews and Beadnell 1902; Anonymous 1901), the American Museum of Natural History (Granger 1908,1910), the University of California (Deraniyagala 1948; Phillips 1948), and Yale University (Simons 1961, 1962, 1964; Simons and Ostrom 1963, 1967; Moustafa 1974). Two additional Eocene localities in Africa have produced cetacean fossils that have not been identified below the level of order. Van Valen (1968:37) quoted a statement by Reyment that in 1965 the latter had seen "many bones of a whale in situ" in the middle Eocene Ameki Formation in Nigeria. The Ameki Formation produced the mandibles and vertebra of the primitive archaeocete Pappocetus lugardi Andrews 1920 (see Kellogg 1936:243-244), and it is imperative that more specimens (such as reported by Halstead and Middleton 1974) be collected to properly place the species in evolutionary context. The fossils reported by Van Valen, Reyment, and Halstead and Middleton demonstrate the necessity for further field work in Nigeria. The second locality is in Libya. Savage (1971:219-220) has reported cetaceans from a late Eocene deposit at Dor el Talha (= Gebel Coquin). The material, which was found with sirenians and late Eocene terrestrial mammals, has not been further identified. During Oligocene time, Tethys retreated northward and the Atlas Mountains were formed. Rocks

Cetacea

representing this age in northern Africa are rare and have often been covered by more recent deposits. No fossil cetaceans of this age are reported in the formally published literature from Africa, nor indeed from any southern continent except Australia and New Zealand. Oligocene fossil cetaceans are reported from marine rocks of North America, Europe, and Asia, where their scarcity has been discussed by several writers (Mead 1975; Orr and Faulhaber 1975; Lipps and Mitchell 1976; Whitmore and Sanders 1976; Barnes 1976). In Miocene time, the Mediterranean Sea in the Tethys area covered parts of northern Africa, but this transgression was not as extensive as the previous Eocene transgression. At the end of the Miocene Epoch, the Mediterranean Sea's connection with the Atlantic Ocean was blocked and evaporation began. Cetaceans have been reported from the early Miocene Marada and Moghara formations, representing the Mediterranean transgression in Libya and Egypt, but these fossils are few. One occurrence of a Miocene ziphiid or beaked whale has been recorded from rocks deposited in a freshwater river system far inland in Kenya (Mead 1975). Pliocene records are few. Cetacean fossils have been reported but not further identified from Pliocene rocks at Gasr es Sahabi in Libya by Savage (1971:221-222). There are several species of cetaceans in collections from Pliocene marine deposits at Langebaanweg in South Africa (see Hendey 1970:103, 1973; 1974:39; 1976:237), but these have not been studied in detail. Hendey has commented on the late Cenozoic record of the southwestern Cape Province as follows: Apart from the Carnivora, the only other mammals which are recorded locally as fossils are Cetacea. Holocene cetacean remains are not uncommon in hominid occupation sites and other deposits adjacent to the present coast and have also been recovered during building operations on Cape Town's reclaimed foreshore area. Heavily mineralized cetacean remains are frequently washed ashore on the beach at Milnerton near Cape Town, in association with other marine fossils and the remains of terrestrial mammals. The latter have included the gomphothere tooth fragment referred to earlier. Most of this material is in private collections and is unstudied. The cetacean remains from Langebaanweg (Hendey 1970a: 103), all of which are from Bed I of the Varswater Formation are also unstudied (Hendey 1974:53).

While Pleistocene records of cetaceans in Africa exist in collections and likely occur in various faunal listings and other compilations, we have been unable to find any certainly identified records readily indexed in the formal world literature. Bones and teeth of uncertain geologic age have

Barnes and Mitchell

been dredged off the west coast of Cape Peninsula, northward as far as Cape Columbine, Saldanha Bay, South Africa, and along the southwestern slope of Agulhas Bank, at a depth of 150 to 200 fathoms. These remains have apparently not yet been studied and published, but a listing of those on exhibit at the South African Museum (Barnard 1954:33) included "vertebrae of moderate sized whales, ear-bones of Right Whales and Fin Whales, portion of lower jaw resembling that of a Killer Whale, portions of the skulls of Beaked Whales (Mesoplodon), portion of a skull closely resembling that of Cuvier's Beaked Whale (Ziphius)."

Systematic Account Order Cetacea Brisson 1762 Suborder Archaeoceti Flower 1883 Family Protocetidae Stromer 1908 Genus Protocetus Fraas 1904 SYNONYMY. Protocetus Fraas 1904a:201. TYPE SPECIES. Protocetus atavus Fraas 1904a. Protocetus atavus Fraas 1904 SYNONYMY. Protocetus atavus Fraas 1904a:201, pis. 10-12. The geologically oldest and the most primitive fossil cetacean known by a skull from Africa is Protocetus atavus. The type species is also the only species of the genus known from Africa. The holotype (no. 11084, Württembergische Naturaliensammlung, Stuttgart) is a well-preserved skull (see Fraas 1904a, pi. 10, figs. 1 - 2 , pi. 11, fig. 1; Kellogg 1936, pi. 34) from near Cairo, Egypt. It was collected from the same stratum as Protosiren fraasi, in the basal portion of the lower Mokattam Formation (=Mokattam Series of Beadnell 1905; Kellogg 1936), the age of which is early middle Eocene (= early Lutetian age). Vertebrae, ribs, a tooth, and part of a second skull have been referred to the species (Fraas 1904a; Stromer 1908b; Pompeckji 1922; Kellogg 1936; Slijper 1936, figs. 64, 199). The skull fragment referred to Protocetus atavus by Stromer (1908b: 108109) is slightly larger than the holotype (Kellogg 1936:237). The total body length can be estimated at approximately 2.5 meters (see also Kellogg 1936:240, 276) based on the vertebral centra lengths and the schematic skeletal reconstruction given by Slijper (1936, fig. 217a). This is small for whales, but about the size of many fossil and Recent porpoises. The primitive characters shown by P. atavus in-

Cetacea

585

clude: the anterior position of the external narial opening, trenchant cheek teeth lacking serrations such as are present on teeth of the Dorudontinae and Basilosaurinae, lack of an enlarged air sinus in the basicranium surrounding the periotic and auditory bulla, short vertebral centra with broad and erect neural spines and large zygapophyses, and apparently an articulation between the innominate bones and sacral vertebrae. The primitive nature of P. atavus, particularly in the dentition, has been used as evidence for its derivation from creodonts (Fraas 1904a; Andrews 1906; Kellogg 1936). Kellogg (1936:276, 279) believed that Protocetus was ancestral to the later Dorudontidae (Dorudon and Zygorhiza, Dorudontinae of this paper), citing as evidence the small body size and relatively unmodified vertebrae with short centra common to all three genera. Genus Pappocetus Andrews 1920 SYNONYMY. Pappocetus Andrews 1920:309. TYPE SPECIES. Pappocetus lugardi Andrews 1920. Pappocetus lugardi Andrews 1920 SYNONYMY. Pappocetus lugardi Andrews 1920: 309, text-fig. 1, pi. 1. This poorly known species was based upon an incomplete mandible (holotype, BM(NH) M-11414, our figure 29.1), and a referred mandible, tooth, and axis vertebra (see Andrews 1920, text fig. 1, pi. 1; Halstead and Middleton 1974, figs. 4, 5), all collected from the Ameki Formation (Reyment 1965; Van Valen 1968:37) in the Ombialla District of Nigeria (Andrews 1920:309). These specimens, from an animal larger than Protocetus atavus, are of middle Eocene (= Lutetian) age (see Kellogg 1936:243; Van Valen 1968:37). They are probably younger geologically than Protocetus atavus, which is of early Lutetian age. Andrews (1920) realized that because of disparate parts, the type mandible of Pappocetus lugardi could not be objectively compared with Protocetus atavus. Kellogg (1936:279-280) regarded the two genera as distinct, but Van Valen (1966:90; 1968:37) questioned the validity of Pappocetus. Subsequently, Sahni and Mishra (1972) described a new species of middle Eocene Protocetus from India, and they considered the two genera "independent but related" (1972:494). They qualified this distinction by pointing out that the mandible fragments referred to their new species, while differing from those of Pappocetus, were only tentatively associated with the type skull of their new species. The type jaw of P. lugardi definitely represents a primitive morphology of protocetid whale. Andrews (1920) cited the "carnivore"-like teeth, which bear

Figure 29.1 Pappocetus lugardi Andrews 1920. Holotype, BM(NH) M-11414, incomplete mandible: (a) left lateral, (b) occlusal, and (c) right lateral views. Referred left dentary, BM(NH) M-11086: (cl) lateral, (e) occlusal, and ( f ) medial views. Scale lines equal 10 cm.

Barnes and Mitchell

cingulae and are present in the full eutherian complement (Ιχ_3, Cj, Ρχ_4, M!_3). More fossils are needed to resolve its uncertain relationships. Van Valen (1968:37) mentioned that in 1965 R. A. Reyment saw bones of a cetacean at the type locality of the Ameki Formation. Dorsal vertebrae and a rib collected near the type locality of Pappocetus lugardi indicated to Halstead and Middleton (1974) that the species had a short body. Genus Eocetus Fraas 1904 SYNONYMY. Mesocetus Fraas 1904a:217, nec van Beneden, 1880, nec Moreno 1892; Eocetus Fraas 1904b:374. TYPE SPECIES. Mesocetus schweinfurthi Fraas 1904a. Eocetus schweinfurthi (Fraas) 1904 SYNONYMY. Mesocetus schweinfurthi Fraas 1904a:217, pi. 1, fig. 3; Eocetus schweinfurthi, Fraas 1904b:374. Eocetus schweinfurthi is known only by the holotype, a skull (Württembergische Naturaliensammlung, Stuttgart, No. 10986). Two lumbar vertebrae described by Stromer (1903b:83-85) were referred to the species by Kellogg (1936:232). All three fossils were collected from the same horizon, in the upper part of the lower Mokattam Formation (= Mokattam Series of Beadnell 1905; Kellogg 1936), which is late middle Eocene in age (Kellogg 1936:232, 280), and they are therefore stratigraphically higher and geologically younger than the holotype of Protocetus atavus. The locality is near Cairo, Egypt, and geographically near the type locality of Protocetus atavus. The skull of Eocetus schweinfurthi (see Fraas 1904a, pi. 10, fig. 3; Kellogg 1936, pi. 33) is much larger than that of Protocetus atavus and is intermediate in size between skulls of the dorudontine Dorudon osiris and the basilosaurine Prozeuglodon isis (Kellogg 1936:280). Like Protocetus atavus, the skull of Eocetus schweinfurthi has the full eutherian dental formula (I1"3, C1, Ρ 1 " 4 , M 1-3 ). The cheek teeth resemble those of Protocetus, but their crowns bear small crenulations on the cutting edges (see Fraas 1904a, pi. 11, figs. 10, 11). These serrations have been interpreted as precursors of the enlarged accessory cusps on the cheek teeth of the Basilosauridae (including Dorudontidae in this chapter, Kellogg 1936:231, 233, 280). Additional advanced characters differentiating E. schweinfurthi from Protocetus include the large size of the skull, its elongate rostrum, more posteriorly positioned external nares, long and narrow intertemporal region,

Cetacea

587

and high sagittal crest (Andrews 1906:xxiii; Kellogg 1936:232-280). Fraas (1904a:219) regarded Eocetus as an intermediate between Protocetus and "Zeuglodon" (=Prozeuglodon and Dorudon), but Kellogg (1936:276, 280), citing the near contemporaneity of Eocetus with Protocetus, suggested that at least two different lines of descent existed, with the protocetid E. schweinfurthi being the oldest member of a morphological series leading to the increasingly derived basilosaurine genera Prozeuglodon, Basilosaurus, and Platyosphys. Van Valen (1968:37) believed that Eocetus is "clearly related to the Basilosauridae and could equally well be referred to the . . . family." The statements by Kellogg and Van Valen, however, rely mainly on the fact that the vertebrae Kellogg referred to Eocetus schweinfurthi have anteroposteriorly elongate centra like those of Basilosaurus and Prozeuglodon. We believe that the referral of these vertebrae to the species is tenuous, and that the many small serrations on the cheek teeth of Eocetus schweinfurthi are not necessarily homologous with the three to four accessory cusps on the teeth of basilosaurids. We retain Eocetus in the Protocetidae, and we caution that the species still is represented with certainty only by the holotype skull. An associated skeleton is needed to provide additional data on the vertebrae of this species. Family Basilosauridae Cope 1868 SYNONYMY. Zeuglodontidae Bonaparte 1849: 618; Hydrarchidae Bonaparte 1850:1; Basilosauridae Cope 1868:144; Stegorhinidae Brandt 1873:334; Dorudontidae Miller 1923:13; Prozeuglodontidae Moustafa 1954:87. Species placed in the genera Dorudon, Zygorhiza, Prozeuglodon, and Basilosaurus have very similar dentition and cranial anatomy. This was recognized by Kellogg (1936) and Moustafa (1954). Most of these species were originally placed in the genus "Zeuglodon" and in the family Zeuglodontidae, both now considered invalid. Miller (1923:13, 40) separated out the species of "Zeuglodon" with short or "normal" vertebral centra and placed them in a new family, Dorudontidae, in contrast to the Basilosauridae, which had vertebral centra "greatly elongated." Kellogg (1936) made diagnoses and generic allocations following Miller's (1923) classification, and he has been followed by most subsequent writers (Simpson 1945; Thenius and Hofer 1960; Dechaseaux 1961; Trofimov and Gromova 1962, 1968; Romer 1966; Simons 1968). Moustafa (1954) reunited the two families, but put them in a new family, the Prozeuglodontidae, "without any implication that the genus [Prozeuglodon] is in any sense the

588

Barnes and Mitchell

type of the family" (1954:87, brackets ours). We agree that it is proper to recognize one family, because of the similarities in skulls and teeth, but believe that the elongated vertebral centra in the larger archaeocetes is not simply a function of large size, as stated by Moustafa (1954:87). For example, vertebral structure, number, and centrum length in species of modern Delphinidae vary widely and are probably a reflection both of phylogenetic relationships and of locomotor function The new family name Prozeuglodontidae, proposed by Moustafa (1954:87), is unnecessary. Article 23, d, 1 of the International Code of Zoological Nomenclature (Stoll et al. 1964) specifies that when a family group is formed by the union of two or more taxa, the oldest valid family group name of any included taxa shall be used. Therefore, Basilosauridae Cope 1868 should be the family name. We do believe that the genera in the previously recognized families Dorudontidae and Basilosauridae form two groups, and we place them in the subfamilies Dorudontinae and Basilosaurinae within the family Basilosauridae. This arrangement is essentially that used by Slijper (1936:540), who reduced Dorudontidae to a subfamily that he included with the subfamily Zeuglodontinae in the family Zeuglodontidae. Subfamily Dorudontinae Miller 1923 SYNONYMY. Dorudontidae Miller 1923:13; Dorudontinae, Slijper 1936:540, as a subfamily of Zeuglodontidae. Genus Dorudon Gibbes 1845 SYNONYMY. Dorudon Gibbes 1845:254-256; Basilosaurus, Gibbes 1847:5-14; Doryodon Cope 1868:144, 155\Durodon Gill 1872:93 (typog. error); Zeuglodon, Dames 1894:204 (part, and see Kellogg 1936:184); Zeuglodon, Smith 1903:322; Zeuglodon, Stromer 1903a:(part); Protocetus, Fraas 1904a:(part); Zeuglodon CDorudon), Stromer 1908a:81-88; Zeuglodon, Abel 1914:204; Zeuglodon, Dart 1923:616618, 627; Prozeuglodon, Kellogg 1928:40. TYPE SPECIES. Dorudon serratus Gibbes 1845. Dorudon intermedius (Dart) 1923 SYNONYMY. ?Zeuglodon osiris Dames 1894:35, 36 (part); Zeuglodon osiris, Stromer 1908b:110,114118 (part); Zeuglodon intermedius Dart 1923:617, 629-632,Dorudon intermedius (Dart 1923), Kellogg 1936:222. Dorudon intermedius is the geologically oldest African species in the genus Dorudon. It was originally based upon characters of an endocast from a skull. The holotype skull (Kellogg 1936, pis. 30-31;

Cetacea

BM[NH] M-10173; our figure 29.2a) and a referred skull figured by Stromer (1908b, as Zeuglodon osiris) are both from the Birket el Qurun Formation of the Fayum, Egypt, although Moustafa (1974:72) listed the species from the Qasr el Sagha Formation. They are of early late Eocene age, and therefore roughly contemporaneous with the basilosaurine Prozeuglodon isis. The skull of D. intermedius, a relatively small archaeocete, is approximately 795 mm (31 inches) long, and its skeleton was probably no more than 4.9 m (16 feet) long. According to Kellogg (1936:287), it differs from the geologically younger Dorudon osiris by having a more highly vaulted cranium and lambdoidal crests that do not project so far posteriorly and are less constricted medially in occipital view. The vertebrae that Kellogg (1936:223) referred to the species had been previously referred by Dames (1894:35-36) to Zeuglodon (=Dorudon) osiris. Kellogg did not discuss the reassignment of them to D. intermedius. We consider this referral tenuous in the absence of associated skeletons of D. intermedius. Dorudon osiris (Dames) 1894 SYNONYMY. Zeuglodon osiris Dames 1894:204; Zeuglodon (Dorudon) osiris (Dames 1894), Stromer 1908b:81-88 (part); Dorudon osiris (Dames 1894), Kellogg 1936:184. Dorudon osiris is the largest species of Dorudon from Egypt. It is represented by the lectotype mandible and premaxillae (Geologisch-Paläeontologisches Inst, und Museum der Universität, Berlin, no. Μ 16) and 21 referred specimens as listed by Kellogg (1936:185-186). All are from the Qasr el Sagha Formation (see Kellogg 1936:184-186; Simons 1968:13-14) of late Eocene age (late Bartonian correlative), and therefore stratigraphically above and geologically younger than D. intermedius from the Birket el Qurun Formation of early late Eocene age. D. osiris is one of three species of Dorudon recognized by Kellogg (1936) from the Qasr el Sagha Formation. It differs from the contemporaneous D. stromeri and the older D. intermedius in three respects: it has a less highly vaulted cranium; the lateral parts of the lambdoidal crests on the cranium project strongly posteriorly and are constricted medially in posterior view; and it is larger in size. Most of the bones in the postcranial skeleton are included in the referred specimens listed by Kellogg (1936), but few if any were found associated with the skulls that comprise the majority of the referred specimens. The vertebrae that Dames (1894:197201) assigned to D. osiris were reassigned to other

Barnes and Mitchell

Cetacea

589

figure 29.2 Skulls of Tethys Basilosauridae: (a) Dorudon intermedins (Dart 1923), holotype, BM(NH) M-10173; b) Dorudon osiris (Dames 1894), referred specimen, BM(NH) M-10228. Scale lines equal 10 cm. ipecies by Kellogg (1936:184). Skeletons associated vith skulls are therefore needed before the postcralial osteology of D. osiris can be objectively studied md compared. Mitchell and El-Khashab could not locate some specimens of D. osiris specifically looked for in the üairo Geological Museum (CGM), 2 0 - 2 1 July 1977. rhese included the partial skull CGM-10018 and the ;ndocast therefrom (Kellogg 1936:186), and the porion of the ramus of the left mandible anterior to the :rack containing canine and incisor alveoli, illusrated by Andrews (1906, fig. 77), specimen CGML0207. (See also discussion under D. elliotsmithii, )elow.) Flouard (1966) described vertebrae and a :heek tooth from near Kaolack in Senegal which he dentified as Zeuglodon cf osiris. The fossils may inleed represent Dorudon osiris but more study is nec;ssary for positive identification. The tooth (Elouard .966, fig. 2) is not a premolar, but a posterior molar see Stromer 1903a:346; Kellogg 1936, pis. 19, 21, !2).

lorudon zitteli (Stromer) 1903 SYNONYMY. Zeuglodon zitteli Stromer 1903b:

70; Protocetus zitteli (Stromer 1903), Fraas 1904a 216, 217, 219; Dorudon zitteli (Stromer 1903) Kellogg 1936:212. Dorudon zitteli is known from the late Eocenc Qasr el Sagha Formation. It therefore was contemporaneous with D. osiris and D. stromeri and geologically younger thanD. intermedius from the earl} late Eocene Birket el Qurun Formation. It is about the size of D. stromeri and D. intermedius and has ε skull similar to theirs in general conformation. Like them, it differs fromZ). osiris by being smaller. Kellogg (1936:214) summarized the unique characters of D. zitteli as: a different curvature of the posterioi margin of the supraorbital process of the frontal, ε bilobed root on P 1 , unusually large auditory bullae a high coronoid process of the mandible, and a scap ula with a narrow prescapular fossa. The species is represented by the holotype, ε weathered skull with vertebrae and ribs (Paläonto logische Sammlung, Alte Akademie, Munich, no 1902.XI.60.), and four confidently or tentatively re ferred specimens. Kellogg's (1936:214) distinctions separating D. zitteli from D. osiris and D. stromer were not strongly made, and more specimens may b


32

0 0 0

3 0 3

0 3 0

3 3 3

2 2 1

2 2 0

0 0 0

0 0 0

0 0 0

3 2 0 0 3 3 2

3 3 3 3 0 0 2

0 3 3 3 0 0 3

3 3 0 3 0 0 3

1 2 3 3 0 0 2

0 2 3 3 0 0 2

0 0 0 3 0 0 2

0 0 2 3 0 0 2

0

0

0

0

1

0

0

0

0

1

0

0

0

0

2

0

0

0

0

0

3

3

0

0

2S

_

3Μ

.53 3Β

0 0 0

0 0 0

0 0 0

0 0 0

0 2 1 3 0 0 2

0 2 1 1 0 0 2

0 2 0 1 0 0 2

0 0 0 0 0 0 0

0 0 0 0 0 0 0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

α>

Species Protocetidae Protocetus atavus Pappocetus lugardi Eocetus schweinfurthi

Basilosauridae Dorudon intermedius Dorudon osiris Dorudon zitteli Dorudon stromeri Dorudoni?) sensitivus Dorudoni?) elliotsmithii Prozeuglodon isis

Archaeoceti undet. "Zeuglodon " cf "Z."

brachyspondylus

?Acrodelphidae Schizodelphis

aff. S. sulcatus

Ziphiidae Gen. et sp. undet. Odontoceti undet. "Delphinus"

vanzelleri

Note: Disparate fossil specimens create problems in comparing taxa and some species may be synonyms. Symbols are: 0, unknown; 1, questionably referred; 2, confidently referred; 3, holotype or lectotype.

Barnes and Mitchell

endocasts are needed for the majority of the species in order to determine their interrelationships and even their validity. This is particularly true for Dorudon zitteli, D. stromeri, D. sensitivus, and-D. elliotsmithii. The identity of "Zeuglodon" brachyspondylus, known only from vertebrae, must be learned. Kellogg pointed out the similarity between these vertebrae and those of Zygorhiza kochii from North America. Numerous skulls and skeletons of "Prozeuglodon" and Dorudon were seen but not collected in the Birket el Qurun Formation by the University of California expedition in 1947 (Phillips 1948:667; Deraniyagala 1948:2, 16, pi. 1). The total known cetacean record in Oligocene through Pleistocene marine rocks in Africa is pitifully meager. It has been shown herein that although suitable sedimentary deposits are rare, they do exist and cetaceans have been reported from some of them. It is an understatement to say that research on Neogene fossil cetaceans in Africa is a wide open field. We thank the following persons for providing Mitchell with access to and assistance with collections: R. Said, R. A. Eissa, and B. El-Khashab of the Geological Museum, Cairo; and A. J. Sutcliffe, A. Gentry, A. Currant, and J. Hooker of the British Museum (Natural History). We thank D. E. Savage and J. H. Hutchison for assisting Barnes with collections at the University of California, Berkeley. R. E. Fordyce helped with literature searches, and V. M. Kozicki took the photographs from which L. Reynolds made the prints. Ms. P. Zeadow and Ms. Daphene Cowan aided in manuscript preparation, and Ms. M. Butler prepared the archaeocete restorations.

References Abel, O. 1900. Untersuchungen über die fossilen Platanistiden des Wiener Beckens. Denkschr. Acad. Wiss. Vienna 68:839-874, pis. 1 - 4 (for 1899). Abel, O. 1905. Les odontocetes du Bolderien (Miocene superieur) d'Anvers. Mem. Mus. Roy. d'Hist. Nat. de Belgique 3(2):1-155. Abel, O. 1914. Die Vorfahren der Bartenwale. Denkschr. Akad. Wiss. Vienna 90:155-224, pis. 1-12 (for 1913). Allen, G. M. 1921. Fossil cetaceans from the Florida phosphate beds. Jour. Mamm. 2(3):144-159, pis. 9-12. Andrews, C. W. 1901a. Preliminary note on some recently discovered extinct vertebrates from Egypt. Part I. Geol. Mag., n.s., decade 4, 8:400-409. Andrews, C. W. 1901b. Preliminary note on some recently discovered extinct vertebrates from Egypt. Part II. Geol. Mag., n.s., decade 4, 8:436-444. Andrews, C. W. 1904. Further notes on the mammals of the Eocene of Egypt. Part III. Geol. Mag., n.s., decade 5, 1(5):211-215. Andrews, C. W. 1906. A descriptive catalogue of the Ter-

Cetacea

599

tiary Vertebrata of the Fayüm, Egypt. London: British Museum (Natural History), xxxvii + 324 p., 26 pis. Andrews, C. W. 1907a. Note on the cervical vertebra of a Zeuglodon from the Barton Clay of Barton Cliff (Hampshire). Quart. Jour. Geol. Soc. London 63:124-127. Andrews, C. W. 1907b. The recently discovered Tertiary Vertebrata of Egypt. Smithsonian Report 1906: 295-307. Andrews, C. W. 1908a. Note on a model of the skull and mandible of Prozeuglodon atrox Andrews. Geol. Mag., ser. 5, 5:209-212, pi. 9. Andrews, C. W. 1908b. Model of the skull and mandible of Prozeuglodon atrox And. Proc. Zool. Soc. London 1908:203. Andrews, C. W. 1920. A description of new species of zeuglodont and of leathery turtle from the Eocene of southern Nigeria. Proc. Zool. Soc. London 22:309-319, pis. 1 - 2 (for 1919). Andrews, C. W. 1923. Note on the skulls from which the endocranial casts described by Dr. Dart were taken. Proc. Zool. Soc. London 1923:648-654. Andrews, C. W., and H. J. L. Beadnell. 1902. A preliminary note on some new mammals from the upper Eocene of Egypt. Cairo: Egypt Surv. Dept. Pub. Works Min., 9 pp., 4 figs. Andrews, R. C. 1921. A remarkable case of external hind limbs in a humpback whale. Amer. Mus. Novit. 9:1-6. Anonymous. 1901. Geological discoveries in Egypt. London Times (4 Oct.). Anthony, R. 1926. Les affinites des cetaces. Ann. Inst. Oceanogr. Monaco, Paris, ser. 2, 3:93-135, pi. 1. Barnard, Κ. H. 1954. A guide book to South African whales and dolphins. South African Museum, Guide no. 4, 33 pp. Barnes, L. G. 1976. Outline of eastern North Pacific fossil cetacean assemblages. Syst. Zool. 25(4):321-343. Beadnell, H. J. L. 1905. The topography and geology of the Fayüm province of Egypt. Cairo: Egypt Survey Dept., Pub. Works Min., 101 pp. Beddard, F. E. 1900. A book of whales. New York: G. P. Putnam's Sons, xv + 320 p. Boyden, Α., and D. Gemeroy. 1950. The relative position of the Cetacea among the orders of Mammalia as indicated by precipitin tests. Zoologica 35:145-151. Cabrera, A. 1922. Manual de Mastozoologia. Madrid and Barcelona: Manuales—Gallach, cxx; 440 + 12 pp. pi. Carter, J. T. 1948. Comparison of the microscopic structure of the enamel in the teeth of Zeuglodon osiris Dames, and of Prosqualodon davidi Flynn. In T. Flynn, Description of Prosqualodon davidi Flynn, a fossil cetacean from Tasmania. Trans. Zool. Soc. London 26: 192-193 + pis. 5-6. Case, E. C. 1934. A specimen of a long-nosed dolphin from the Bone Valley Gravels of Polk County, Florida. Univ. Michigan Contrib. Mus. Paleontol. 4(6):105-113, pis. 1-2.

Cope, E. D. 1868. An addition to the vertebrate fauna of the Miocene period, with a synopsis of the extinct Ceta-

600

Barnes and Mitchell

cea of the United States. Proc. Acad. Nat. Sei. Philadelphia 19(4): 138-156. Dal Piaz, G. 1901. Di alcuni resti di Cyrtodelphis sulcatus dell'arenaria di Belluno. Pt. 1. Palaeont. Italica 9:1877:287-292, pi. 34. Dal Piaz, G. 1903. Sugli avanzi di Cyrtodelphis sulcatus dell'arenaria di Belluno. Pt. 1. Palaeont. Italica 9:187219, pis. 28-31. Dal Piaz, G. 1905. Sugli avanzi di Cyrtodelphis sulcatus dell'arenaria di Belluno. Pt. 2. Palaeont. Italica 11: 253-279, pis. 18-21. Dames, W. 1894. Über Zeuglodonten aus Aegypten und die Beziehungen der Archaeoceten zu den übrigen Cetaceen. Palaeontol. Abhandl., Jena, neue folge, 1(5): 189-222 (reprint numbered 1-36), pis. 30-36. Dart, R. A. 1923. The brain of the Zeuglodontidae (Cetacea). Proc. Zool. Soc. London 42:615-648. Dechaseaux, C. 1961. Cetacea, p. 831-886. In J. Piveteau, ed., Traite dePaleontologie, vol. 6, pt. 1. Paris: Masson et Cie. Deraniyagala, P. E. P. 1948. Some scientific results of two visits to Africa. Spolia Zeylanica 25(2): 1-42, pis. 1-14. filouard, P. 1966. Decouverte d'un archeocete dans les environs de Kaolack. Notes Africaines 109:8-10. Emlong, D. 1966. A new archaic cetacean from the Oligocene of northwest Oregon. Univ. Oregon Mus. Nat. Hist. Bull. 3:1-51. Flower, W. H. 1883. Cetacea. In Encyclopaedia Britannica, 9th ed., vol. 15, pp. 391-400. Edinburgh. Fourtau, R. 1918. fitudes sur les vertebres fossiles du Wadi Natroun. Bull. Inst, figypte 6(1):195-196. Fourtau, R. 1920. Contribution ά l'etude des vertebres miocenes de l'ßgypte. Cairo: Egypt. Surv. Dept., 121 pp. + 3 pis. Fraas, E. 1904a. Neue Zeuglodonten aus dem unteren Mitteleocän vom Mokattam bei Cairo. Geol. Palaeont. Abhandl., neue folge, 6(3):199-220, pls. 10-12. Fraas, E. 1904b. (Abstract No. 1048). Geol. Centralbl., Leipzig 5(8):374. Geinitz, Η. B. 1884. Nachträge zu den Funden in den phosphatlagern von Helmstedt, Büddenstedt u.a. Sitz.-Ber. Abhandl. naturwiss. Ges. Isis 1883:105-111. Gervais, P. 1853. Sur quelques ossements fossiles de phoques et de cetaces du Languedoc et de l'Aquitaine. Bull. Soc. Geol. France, ser. 2, 10:311-313. Gervais, P. 1861. Sur differentes especes de vertebres fossiles observees pour la plupart dans le midi de la France. Mem. Acad. Sei. Lett. Montpellier 5:117-132, pl. 4. Gibbes, R. W. 1845. Description of the teeth of a new fossil animal found in the Green-sand of South Carolina. Proc. Acad. Nat. Sei. Philadelphia 2:254-256, pl. 1. Gibbes, R. W. 1847. On the fossil genus Basilosaurus, Harlan, (Zeuglodon, Owen,) with a notice of specimens from the Eocene Green Sand of South Carolina. J. Acad. Nat. Sei. Philadelphia, n.s., 1(2):5-15, pis. 1 - 5 (1848?). Gidley, J. W. 1913. A recently mounted Zeuglodon skeleton in the United States National Museum. Proc. U.S. Nat. Mus. 44:649-654, pis. 81-82.

Cetacea

Gill, T. 1872. Arrangement of the families of mammals. Smithsonian Misc. Coll., No. 230, ll(l):i-vi, 1-98. Gimbrede, L. de A. 1962. The Hurricane lentil: source of many Claiborne fossils. Jour. Paleontol. 36(5): 11161120. Goodwin, R. F. 1952. Foetal fructose in various mammals. Nature 170:750. Granger, W. 1908. A preliminary notice of the Fayüm collection. Nat. Hist. 8:8-14. Granger, W. 1910. [Report on the exploration of the Fayüm.] Trans. New York Acad. Sei. 19:303-304. Gregory, W. K. 1910. The orders of mammals. Bull. Amer. Mus. Nat. Hist. 27:1-524. Halstead, L. B., and J. A. Middleton. 1972. Notes on fossil whales from the Upper Eocene of Barton, Hampshire. Proc. Geol. Assoc. (London) 83(2):185-190. Halstead, L. B., and J. A. Middleton. 1974. New material of the archaeocete whale, Pappocetus lugardi Andrews, from the middle Eocene of Nigeria. Jour. Mining and Geol. 8(1 + 2):81-85. Halstead, L. B„ and J. A. Middleton. 1976. Fossil Vertebrates of Nigeria. Part II. 3. Eocene Vertebrates. 3.4. Archaeocete whale: Pappocetus lugardi Andrews, 1920. The Nigerian Field 41(3):131-133, figs. 4-6. Hamilton, W. R. 1973. A lower Miocene mammalian fauna from Siwa, Egypt. Palaeontology 16(2):275-281. Hendey, Q. B. 1970. A review of the geology and palaeontology of the Plio/Pleistocene deposits at Langebaanweg, Cape Province. Ann. S. African Mus. 56:75-117. Hendey, Q. B. 1973. Fossil occurrences at Langebaanweg, Cape Province. Nature 244:13-14. Hendey, Q. B. 1974. The late Cenozoic Carnivora of the south-western Cape Province. Ann. S. African Mus. 63:i-ii, 1-369. Hendey, Q. B. 1976. The Pliocene fossil occurrences in "E" Quarry, Langebaanweg, South Africa. Ann. S. African Mus. 69:215-247. Ishihara, Υ., T. Saito, Y. Ito, and M. Fujino. 1958. Structure of sperm and sei whale insulins and their breakdown by whale pepsin. Nature 181:1468-1469. Kellogg, A. R. 1924. A fossil porpoise from the Calvert Formation of Maryland. Proc. U.S. Nat. Mus. 63(14): 1-39, pis. 1-18. Kellogg, A. R. 1925. On the occurrence of remains of fossil porpoises of the genus Eurhinodelphis in North America. Proc. U.S. Nat. Mus. 66(26):l-40, pis. 1-17. Kellogg, A. R. 1928a. The history of whales: their adaptation to life in the water. Q. Rev. Biol. 3(l):29-76. Kellogg, A. R. 1928b. The history of whales: their adaptation to life in the water (concluded). Q. Rev. Biol. 3(2): 174-208. Kellogg, A. R. 1936. A review of the Archaeoceti. Carnegie Inst. Washington Publ. 482:i-xv + 1-366, pis. 1-37. Kellogg, A. R. 1938. Adaptation of structure to function in whales. Carnegie Inst. Washington Publ. 501:649-682. Keyes, I. W. 1973. Early Oligocene squalodont cetacean from Oamaru, New Zealand. New Zealand Jour. Marine and Freshwater Res. 7(4):381-390.

Barnes and Mitchell

Kleinenberg, S. Ε. 1958. Κ voprosu ο proiskhozhdenii kitoobraznykh [The origin of the Cetacea]. Doklady Akad. Nauk S.S.S.R. 122(5):950-952. Kleinenberg, S. E., and Α. V. Yablokov. 1958. Ο morfologii verkhnikh dykhatel'nykn putei kitoobraznykh [The morphology of the upper respiratory passages in the cetaceans], Zool. Zhurnal 37, edit. 7:1091-1099. Kuhn, Ο. 1935. Archäoceten aus dem norddeutschen Alttertiär. Zentralblatt für Mineralogie, Geologie und Paläontologie, Abteilung B: Geologie und Paläontologie: 219-226. Lipps, J. H., and E. D. Mitchell. 1976. Trophic model for the adaptive radiations and extinctions of pelagic marine mammals. Paleobiology 2(2): 147-155. Lucas, F. A. 1900. The pelvic girdle of Zeuglodon, Basilosaurus cetoides (Owen), with notes on other portions of the skeleton. Proc. U.S. Nat. Mus. 23:327-331, pis. 5-7. Maglio, V. J. 1969. A shovel-tusked gomphothere from the Miocene of Kenya. Breviora, Mus. Comp. Zool., Harvard Coll. 310:1-10. Makino, S. 1948. The chromosomes of Dall's Porpoise, Phocoenoides dallii (True), with remarks on the phylogenetic relation of the Cetacea. Chromosoma 3(3):220-231. Mchedlidze, G. A. 1976. Osnovnyye cherty paleobiologicheskoi istorii Kitoobraznykh [Basic features of the paleobiological history of the Cetacea]. Akademia Nauk Gruzinskoi S.S.R., Inst. Paleobiol., "Metsniereba" Press, 136 pp., 32 pis. [In Russian, English summary.] Mead, J. G. 1975. A fossil beaked whale (Cetacea: Ziphiidae) from the Miocene of Kenya. Jour. Paleont. 49(4):745-751. Miller, G. S., Jr. 1923. The telescoping of the cetacean skull. Smithsonian Misc. Coll. 76(5): 1-70. Mitchell, E. D. 1966. Faunal succession of extinct North Pacific marine mammals. Norsk Hvalfangst-Tidende 1966 (3):47-60. Mitchell, E. D. 1970. Pigmentation pattern evolution in delphinid cetaceans: an essay in adaptive coloration. Canadian J. Zool. 48(4):717-740. Mossman, H. W. 1937. Comparative morphogenesis of the fetal membranes and accessory uterine structures. Publ. Carnegie Inst. Washington 482:130-246. Moustafa, Y. S. 1954. Additional information on the skull of Prozeuglodon isis and the morphological history of the Archaeoceti. Proc. Egyptian Acad. Sei. 9:80-88, pi. 1 (for 1953). Moustafa, Y. S. 1974. Critical observations on the occurrence of Fayum fossil vertebrates. Ann. Geol. Surv. Egypt 4:41-78. Müller, J. 1849. Über die fossilen Reste der Zeuglodonten von Nordamerica, mit Rücksicht auf die europäischen Reste aus dieser Familie. Berlin: G. Reimer, iv + 38 pp., 27 pis. (folio). Nemoto, Τ. 1963. New records of sperm whales with protruded rudimentary hind limbs. Sei. Repts. Whales Res. Inst. 17:79-81. Norris, K. S. 1968. The evolution of acoustic mechanisms

Cetacea

601

in odontocete cetaceans, p. 297-324. In Ε. T. Drake, ed., Evolution and Environment. New Haven: Yale Univ. Press, 470 pp. Ogawa, T., and T. Kamiya. 1957. A case of the cachalot with protruded rudimentary hind limbs. Sei. Repts. Whales Res. Inst. 12:197-208. Ohsumi, S. 1965. A dolphin (Stenella caeruleoalba) with protruded rudimentary hind limbs. Sei. Repts. Whales Res. Inst. 19:135-136, pis. 1-2. Orr, W. N., and J. Faulhaber. 1975. A middle Tertiary cetacean from Oregon. Northwest Sei. 49:174-181. Osborn, H. F. 1924. Andrewsarchus, giant mesonychid of Mongolia. Amer. Mus. Novit. 146:1-5. Parry, D. A. 1949. The anatomical basis of swimming in whales. Proc. Zool. Soc. London 119:49-60. Phillips, W. 1948. Recent discoveries in the Egyptian Faiyum and Sinai. Science 107(2791):666-670. Pompeckj, J. F. 1922. Das Ohrskelett von Zeuglodon. Senckenbergiana 4(3-4):43-100, pi. 2. Raven, H. C., and W. K. Gregory. 1933. The spermaceti organ and nasal passages of the sperm whale (Physeter catodon) and other odontocetes. Amer. Mus. Novit. 677: 1-18.

Reyment, R. A. 1965. Aspects of the geology of Nigeria. Ibadan: Ibadan Univ. Press, 145 pp. Rice, D. W. 1967. Cetaceans. In S. Anderson and J. K. Jones, Jr., eds.,Recent mammals of the world: a synopsis of families, pp. 291-324. New York: Ronald Press Co. Romer, A. S. 1966. Vertebrate paleontology, 3rd ed. Chicago and London: Univ. Chicago Press, viii + 468 pp. Sahni, Α., and V. P. Mishra. 1972. A new species of Protocetus (Cetacea) from the middle Eocene of Kutch, western India. Palaeontology 15(3):490-495. Sahni, Α., and V. P. Mishra. 1975. Lower Tertiary vertebrates from western India. Monogr. Palaeontol. Soc. India 3:1-48 + 6 pis. Said, R. 1962. The geology of Egypt. New York: American Elsevier, 377 pp. Satsangi, P. P., and P. K. Mukhopadhyay. 1975. New marine Eocene vertebrates from Kutch. J. Geol. Soc. India 16(l):84-86. Savage, R. J. G. 1969. Early Tertiary mammal locality in southern Libya. Proc. Geol. Soc. London 1657:167-171. Savage, R. J. G. 1971. Review of the fossil mammals of Libya. In C. Gray, ed., Symposium on the Geology of Libya, pp. 215-225. Faculty of Science, Univ. of Libya. Seeley, H. G. 1876. Notice on the occurrence of remains of a British fossil zeuglodon (Z. Wanklyni, Seeley) in the Barton Clay of the Hampshire Coast. Quart. Jour. Geol. Soc. London 32, pt. 4 (128):428-432. Seeley, H. G. 1881. Note on the caudal vertebra of a cetacean discovered by Prof. Judd in the Brockenhurst beds, indicative of a new type allied to Balaenoptera (Balaenoptera juddi). Quart. Jour. Geol. Soc. London 37(148): 709-712. Simons, E. L. 1961. [Untitled.] Soc. Vert. Paleo. News Bull. 61:13-14.

602

Barnes and Mitchell

Simons, Ε. L. 1962. An expedition to the Egyptian desert. Yale Sei. Mag. Simons, E. L. 1964. [Untitled.] Soc. Vert. Paleo. News Bull. 70:14-15. Simons, E. L. 1968. Early Cenozoic mammalian faunas, Fayum Province, Egypt. Part 1. African Oligocene mammals: introduction, history of study, and faunal succession. Peabody Mus. Nat. Hist. Yale Univ. Bull. 28:1-21.

Simons, E. L., and J. H. Ostrom. 1963. [Untitled.] Soc. Vert. Paleo. News Bull. 67:18-19. Simons, E. L., and J. H. Ostrom. 1967. [Untitled.] Soc. Vert. Paleo. News Bull. 80:8-10. Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bull. Am. Mus. Nat. Hist. 85:i-xvi, 1-350. Slijper, E. J . 1936. Die Cetaceen. Vergleichend—anatomisch und systematisch. Capita Zool. 6-7:i-xvi, 1-590. Slijper, E. J. 1958. Walvissen. Amsterdam: Centen's Uitgeversmaatschappij, 524 pp. Slijper, E. J. 1962. Whales. London: Hutchinson & Co., 475 pp. Smith, G. E. 1903. The brain of the Archaeoceti. Proc. Roy. Soc. London 71:322-331. Stoll, N. R., et al., eds. 1964. International Code of Zoological Nomenclature adopted by the XV International Congress of Zoology. London: International Trust for Zoological Nomenclature. 176 pp. Stromer, Ε. 1903a. Bericht über eine von den Privatdozenten Dr. Max Blanckenhorn und Dr. Ernst Stromer von Reichenbach ausgeführte Reise nach Aegypten. Einleitung und ein Schädel und Unterkiefer von Zeuglodon osiris Dames. Sitz, mathemtisch-physikalische Classe der kgl. Bayerischen Akad. Wissensch. 32(3):341-352. Stromer, E. 1903b. Zeuglodon-Reste aus dem oberen Mitteleocän des Fajüm. Beitr. Paläontol. Geol. Osterr. Ung. 15(2 + 3):59-100, pls. 8 - 1 1 . Stromer, E. 1907. Fossile Wirbeltier-Reste aus dem Uadi Färegh und Uadi Natrün in Ägypten. Abh. Senkenberg. Naturf. Gesell. 29:97-132, pl. 20. Stromer, E. 1908a. Die Urwale (Archaeoceti). Anat. Anz. 33(4-5):81-88, pl. 1. Stromer, E. 1908b. Die Archaeoceti des ägyptischen Eozäns. Beitr. Paläontol. Geol. Osterr.-Ung. Orients. 21:106-177, pls. 4 - 7 . Struthers, J. 1881. On the bones, articulations, and muscles of the rudimentary hind-limb of the Greenland right-whale (Balaena mysticetus). Jour: Anat. Physiol., London 15:141-176, 301-321, pls. 14-17. Struthers, J. 1893. On the rudimentary hind-limb of a great fin-whale (Balaenoptera musculus) in comparison with those of the Humpback whale and the Greenland right whale. Jour. Anat. Physiol. 27:291-335, pls. 17-20. Szalay, F. S. 1969a. The Hapalodectinae and a phylogeny of the Mesonychidae (Mammalia, Condylarthra). Amer. Mus. Novit. 2361:1-26. Szalay, F. S. 1969b. Origin and evolution of function of the

Cetacea

mesonychid condylarth feeding mechanism. Evolution 23:703-720. Szalay, F. S., and S. J . Gould. 1966. Asiatic Mesonychidae (Mammalia, Condylarthra). Bull. Am. Mus. Nat. Hist. 132(2):129-173, pls. 9 - 1 2 . Tarlo, L. Β. H. 1964. A primitive whale from the London Clay of the Isle of Sheppey. Proc. Geol. Assoc. 74: 319-323 (for 1963). Thenius, E., and H. Hofer. 1960. Stammesgeschichte der Säugetiere. Eine Übersicht über Tatsachen und Probleme der Evolution der Säugetiere. Berlin: Springer-Verlag, 322 pp. Thompson, D'A. W. 1889. On the systematic position of Zeuglodon. Stud. Mus. Zool. Univ. College Dundee 1(9): 1-8.

Tomilin, A. G. 1957. Kitoobraznye [Cetacea], In Zveri SSSR i prilezhashchikh stran [Mammals of the U.S.S.R. and adjacent countries], vol. 9, Moscow: Izdatel'stvo Akademi Nauk S.S.S.R. Tomlin, A. G. 1967. Cetacea. In V. G. Heptner, ed., Mammals of the U.S.S.R. and adjacent countries, vol. 9. Jerusalem: Israel Program for Scientific Translations, 717 pp. Trofimov, Β. Α., and V. I. Gromova. 1962. Otryad Cetacea (Kitoobraznye). In Y. A. Orlov, ed., Osnovi Paleontologii vol. 13, pp. 171-182. Moscow: Gosudarstvennoe Nauchno-Tekhnicheskoe Izdatel'stvo Literatury po Geologii i Okhrane Nedr. Trofimov, Β. Α., and V. I. Gromova. 1968. Order Cetacea. In V. I. Gromova, ed., Fundamentals of paleontology, vol. 13, Mammals, pp. 225-241. Jerusalem: Israel Program for Scientific Translations. True, F. W. 1908. On the classification of the Cetacea. Proc. Amer. Philos. Soc. Philadelphia 47(189):385-391. Van Beneden, P. J. 1883. Sur quelques ossements de cetaces fossiles, recueillis dans les couches phosphatees entre l'Elbe et le Weser. Bull. Acad. Roy., Belgique, ser. 3, 6:27-33. Van Valen, L. 1966. Deltatheridia, a new order of mammals. Bull. Am. Mus. Nat. Hist. 132(1):1-126, pls. 1-8.

Van Valen, L. 1968. Monophyly or diphyly in the origin of whales. Evolution 22(1):37-41. Weber, M. C. W. 1904. Die Säugetiere. Einführung in die Anatomie und Systematik der Recenten und Fossilen Mammalia. Jena: Gustav Fischer, 866 pp. Whitmore, F. C., and Α. E. Sanders. 1976. Review of the Oligocene Cetacea. Syst. Zool. 25(4):304-320. Yablokov, Α. V. 1964. Konvergentsiya ili parallelizm ν razvitii Kitoobraznykh. [Convergence or parallelism in the evolution of cetaceans.] Paleont. Zhur. 1964 (1): 97-106. Yablokov, Α. V. 1965. Convergence or parallelism in the evolution of cetaceans. Int. Geol. Rev. 7(8):1461-1468. Zhemkova, Z. P. 1965. Ο proiskhozhdenii Kitoobraznykh 0Cetacea) [On the origin of whales (Cetacea)]. Zool. Zh. 44(10): 1546-1552. [In Russian, English summary.]

603

30 Patterns of Faunal Evolution Vincent J. Maglio

From the preceding chapters it is clear that the socalled "African fauna" is in many ways no more African than Eurasian. Its development was closely linked with mammalian evolution in northern continents and its history was characterized by slow accretion punctuated by quantum shifts in composition. Yet throughout the later Cenozoic Africa gave as much to the north as it received in return. The exact sequence of events that led to the various modern Old World faunas is not entirely clear, and perhaps never will be, but from continued studies of the kind discussed in this volume, at least the broad outlines of these events are emerging. In spite of tremendous strides over the last several decades, the record still remains limited and certainly grossly underrepresented in some cases. Cooke (1972) has pointed out that for the entire African Tertiary we know only about 150 fossil genera and fewer than 250 species, compared to 256 genera and 740 species living in the continent today. The record improves drastically in the Pleistocene, where another 150 fossil genera are seen. Clearly we must approach any discussion of Cenozoic African faunas with due skepticism, keeping in mind that probably upwards of 80% or more of the continent's former diversity remains undiscovered. The inadequacy of the record is seen principally on the species level. But many genera also are unknown, and even on the family or ordinal levels we must certainly be lacking knowledge, especially of earlier Tertiary groups that were either endemic to the continent or that, after entering from Eurasia, failed to survive. We need only recall the order Pholidota for a dramatic example. This group probably existed in Africa since the late Oligocene or early Miocene (Patterson, in press) and yet remains unrepresented as fossil except in upper Pleistocene deposits (Klein 1972). The Embrithropoda, a uniquely African order, would have remained completely unknown were it not for a single Oligocene locality in Egypt. In the midst of this problem we may still find it instructive to analyze major faunal events that shaped the biological profile of Africa. Beginning with an original stocking of primitive placental mammals, there followed a sequence of new immigrations from Eurasia, plus in situ evolution of endemic groups, each subsequent step further complicating the faunal array. Several authors have examined various aspects of this history. Cooke (1968, 1972) discusses major fossil localities and faunas through the Cenozoic within the framework of paleogeographic and paleoecologic studies. Coryndon and Savage (1973) analyze

604

Maglio

the extent of faunal communication between African and Eurasiatic plates across the closing Tethys Seaway that separated them and give data on major periods of mammalian dispersion into and out of Africa. It would be pointless to repeat here what is said by these authors, and this chapter will attempt only to pinpoint the most important episodes in this history and to summarize overall patterns in the evolution of modern African mammals. Most of the data used here were drawn directly from preceding chapters in this volume and from sources cited therein. The reader is referred to these references so that specific bibliographic citations need not be repeated and can be kept to a minimum here.

Faunal Successions In table 30.1 are listed "typical" faunas of Cenozoic epochs in Africa. These are typical only in the sense that the genera represented have their primary temporal distributions as shown, and a fauna of any particular age will include many of the genera shown for that epoch. Actual assemblages will depend largely on geographic region, ecologic setting, and the presence or absence of competing forms. (For tables, see pp. 613-619.) It can be seen that the entire Cenozoic record is punctuated by "sudden" appearance and disappearance of taxa on all levels from order down to genus. Some of these represent true evolutionary or dispersal events, whereas others are artifacts of the record. Without discussing details of fauna or geology, both discussed admirably by Cooke (1972), I will briefly review the major features in the origin of the modern African assemblage.

Mesozoic The mammalian record in Africa properly begins in the Triassic, where forms transitional between therapsid reptiles and true mammals can be found. Two undoubted mammals have been described from southern Africa. Erythrotherium parringtoni (Crompton 1964) from beds of late Triassic age in Lesotho appears to be closely related to Morganucodon of similar age in England, and a common ancestor has been suggested for both. Similarly, Megazostrodon rudnerae from upper Triassic deposits of Lesotho is very similar to European contemporaries (Crompton and Jenkins 1968). The only other described Mesozoic mammal from Africa is an edentulous jaw from the late Jurassic of Tendaguru, Tanza-

Patterns ofFaunal Evolution

nia, described by Dietrich (1928) as Brancatherulum tendagurense, of uncertain affinities. It is unlikely that any of these early mammalian records had anything to do directly with the origin of later mammals in the continent, but they do demonstrate a close faunal tie between Africa and Europe during this time interval, a tie that was to persist probably at least until the end of the Paleocene.

Paleocene-Eocene Cretaceous and Paleocene records of mammals in Africa are totally lacking and we can only speculate as to the course of events that was to shape the structure of later faunas so characteristic of the Eocene and Oligocene. It is likely that the Paleocene and earlier Eocene witnessed immigration into Africa of at least three groups from the north—one or more primitive condylarthran stocks that were to give rise to later African subungulates, a prosimian primate stock, and creodont carnivores. During the middle and late Eocene we catch glimpses of a record, still grossly incomplete and confined entirely to northern Africa from Egypt to Senegal (see Savage 1969). These faunas are limited to only nine families, of which five are marine mammals (Cetacea and Sirenia). These are the earliest records of the orders and consist of already highly specialized forms, suggesting a long prior history, perhaps from late Paleocene times. Of the remaining four families, one is the hyaenodontid carnivore Apterodon and the others are members of endemic orders, probably with a common origin in the middle Eocene. These are gomphotheriid proboscideans and, only distantly related to them, Moeritherium and Barytherium. Without an earlier Paleocene record we can only guess that the whales derived from some creodont stock of which no trace remains and that sirenians, moeritheres, barytheres, and proboscideans arose in the early to middle Eocene from earlier subungulate invaders. The hyaenodont undoubtedly represents a late Eocene migrant from Europe. Just what was going on in the interior of the continent is not known, and it is intriguing to wonder what new groups wait to be discovered there.

Oligocene By the early Oligocene a great change had occurred in African faunas. Here, too, the record is underrepresented and confined to North Africa. Only two families persist from the Eocene, gomphotheres and hyaenodonts, and to these 14 new families are added. Eight of the latter are rodents, insectivores, or bats, most of endemic origin, derived from some-

Maglio

what earlier invasions from Eurasia. Of the larger mammals, creodonts are abundant, with four genera and nine species known from the Fayum alone. Primates now appear in the record, represented by three endemic families, Parapithecidae, Pongidae, and Hylobatidae, all apparently derived from earlier prosimian stock. The only artiodactyls are anthracotheres, represented by two genera with close relatives in Eurasia. Two new endemic orders make their appearance here, the Embrithropoda and Hyracoidea, each with a single family. The former filled the large herbivore niche but never amounted to more than a single genus. Hyracoids, filling a medium-size browser niche, appear in full radiation showing the greatest diversity they will ever achieve, with seven genera known only from northeastern Africa. Both of these orders clearly evolved in the continent and their origins must have occurred much earlier, in the later half of the Eocene. Of earlier groups that do not continue into the Oligocene record, cetaceans and sirenians almost cer-

Patterns

of Faunal Evolution

605

tainly persisted in some form, although not known as fossil, and two groups, Moeritherioidea and Barytherioidea, became extinct without issue.

Earlier Miocene By far the most dramatic faunal upheaval on the continent occurred in the late Oligocene-early Miocene (figures 30.1 to 30.4). Fossil localities are concentrated in East Africa, with several in North Africa and one in the Namib. Only 14 families present earlier are also recorded here. Twenty-nine new families and 79 new genera make their appearance, and of these, 12 families and 31 genera are micromammals. In terms of overall generic resemblances for this fauna, Savage (1967, p. 277) has shown a greater link between Africa and Asia during the Burdigalian than between Africa and Europe. One order, Embrithropoda, failed to survive from the Oligocene, and two new ones are recorded. One, Tubulidentata, is probably autochthonous and of African subungulate origin; the other, Perissodac-

Figure 30.1 Relative generic diversity in the Cenozoic for families of artiodactyls in Africa. Large arrowheads indicate times when families first entered the continent. Small arrowheads indicate periods of later immigration. Families lacking arrowheads are autochthonous. (Ε) Europe, (A) Asia.

606

Maglio

Patterns ofFaunal

Evolution

Figure 30.2 Relative generic diversity in the Cenozoic for families of carnivores in Africa. Symbols as in figure 30.1.

tyla, certainly entered from Eurasia and is represented by two families, Chalicotheriidae and Rhinocerotidae. The former was a chance invasion and the family subsequently became extinct here. The latter was far more successful, with four genera penetrating the continent independently. Pliohyracidae persist but are drastically reduced, only three genera being represented; the extant family Procaviidae is now also present, derived from some earlier hyracoid stock, perhaps in late Oligocene times. Proboscidea have by this time begun a radiation into several basic groups, with Mammutidae certainly derived from gomphotheres late in the Oligocene. Creodont carnivores remain diverse with four new genera added to two that carry through from earlier times. The first fissiped groups enter at this time from Eurasia and include canids, viverrids, and felids. But together these families barely equal contemporary creodonts in generic diversity. Primates remain about as abundant as earlier. Hylobatids are still rather rare and pongids comprise two genera, of which Dryopithecus is the more

diverse. Monkeys (of uncertain family reference) are recorded for the first time, as are Lorisidae, although the latter may have existed earlier. Among artiodactyls, anthracotheres remain with three genera, one probably representing a new immigrant from Asia. Archaic suids entered at about this time with four Eurasiatic genera recorded and two additional genera probably representing in situ evolution from primitive Eurasiatic stocks. Tragulids and palaeomerycids also expanded into Africa at this time, with similar genera occurring in Africa and Eurasia. At least two endemic giraffes are seen in lower Miocene deposits, representing the Palaeotraginae and Sivatheriinae; both appear to have derived from African ancestors. Of the Bovidae only Walangania is recorded in East Africa and slightly later Protragocerus and Eotragus appear in North Africa, the latter two also known in Europe at about the same time.

Later Miocene By the late Miocene another 18 families have appeared in Africa. Eleven of these are micromam-

Maglio

mals, of which five are bats known in Europe much earlier. They were perhaps present but unrecorded in the African early Miocene or even Oligocene. Of the primates, Hominidae are now present although rare, and hylobatids appear not to have survived in Africa. Mustelid and hyaenid carnivores have now entered the continent, and among the hyraxes the family Procaviidae has completely replaced the earlier pliohyracids. Elephants of a primitive type had only recently emerged from earlier gomphotheres, but the latter also persist in the form of short-jawed Anancinae. Rhinocerotids are still a diverse group, with a new form related to Recent genera emerging from some unknown earlier stock. From Eurasia Hipparion populated the continent during middle Miocene times as the first African equid. Artiodactyls for the first time are beginning to dominate the African landscape. Hexoprotodont Hippopotamidae, recently evolved possibly in Africa, are now seen both here and in Asia, and more advanced suids of the subfamily Suinae are

Patterns

ofFaunal

Evolution

607

characteristic. Nine distinct tribes of Bovidae are now recognized, mostly with Asiatic ties. At least nine genera are recorded by latest Miocene times, compared to only one in the early Miocene. Of these, five are unknown outside the continent, suggesting some in situ evolution of recently arrived stocks. Four new giraffid genera are seen here for the first time, at least two probably evolved in Africa from earlier members of the family. Anthracotheres and palaeomerycids are now rare and fail to survive the epoch.

Pliocene By Pliocene time faunal changes on the family level had tapered off, with only three new families appearing. But a major revolution was occurring on the generic level as 81 new genera are recorded for the first time. Thus 76% of the land-mammal Pliocene fauna is new, and this represents the greatest single faunal change since the revolution of the Oligocene-Miocene transition. About 53% of these genera are endemic to Africa.

Figure 30.3 Relative generic diversity in the Cenozoic for families of primates and perissodactyls in Africa. Symbols as in figure 30.1.

608

Maglio

Patterns ofFaunal Evolution

Figure 30.4 Relative generic diversity in the Cenozoic for families of varied subungulate orders and for micromammals. Symbols as in figure 30.1.

Of groups evolved in Africa, monkeys definitely assignable to the Cercopithecidae are now widespread, as are at least two types of hominids. Although pongids are lacking from the record, they certainly were present. Two new endemic genera of orycteropodids have evolved and a large deinothere has replaced the smaller Miocene form. Two new genera of procaviid hyraxes are seen for the first time, and three new genera of elephant have emerged from a Mio-Pliocene radiation of the family. Suine genera number about six, all endemic but of uncertain relationship. Fifteen bovid genera have been added to previous ones. Of these all but five are unknown outside the continent although most have close relatives in Asia, suggesting immigration during the late Miocene. N e w groups recently arrived in Africa include the camel, clearly of Asiatic origin, the modern Giraffa, a chalicothere from Asia, the genus Equus, a Eurasiatic saber-toothed carnivore, and probably an agriotherine bear.

Earlier Pleistocene The early Quaternary as a whole saw relatively minor changes in the African fauna on the family level, only four new ones appearing at this time. Three of the latter, Leporidae, Rhizopodidae, and Cervidae, entered from Europe, and the fourth, the bat family Myzopodidae, is endemic and now confined to Madagascar. Deinotheres, chalicotheres, and gomphotheres finally vanish from the African scene at the end of this period. The major changes seen in African faunas are on the generic level. About 53 new genera appear here, but since 89 Pliocene genera persisted into the Quaternary, the total character of the fauna is less drastically altered than in the Mio-Pliocene transition. About half of these new appearances were immigrants from the north and half were products of in situ evolution. Of the 16 carnivore genera first recorded here, at least nine seemingly entered from Eurasia, where they are also present. Two new suid genera, eleven bovids, a number of monkeys, bats,

Maglio

and rodents also penetrated the continent at this time.

Later Pleistocene During the last half of the Pleistocene very little occurred to alter the complexion of African mammalian faunas. Only one family new to the continent appears in the record. This group, the Pholidota, although not recorded earlier, must certainly have been present since earlier Tertiary times. On the generic level, again only minor changes took place, with the disappearance of older forms being as important as the appearance of new ones. Most of these latter evolved from earlier groups in Africa and therefore do not occur outside the continent.

Endemism On the basis of overall faunal comparisons with Eurasia many authors have recently concluded that Africa was relatively isolated during Oligocene and earliest Miocene time (van Couvering 1972, p. 264; Coryndon and Savage 1973, p. 123). Such isolation must have followed a period of limited faunal exchange with continents to the north during the Paleocene and early Eocene, for by late Eocene and early Oligocene times a number of specialized and uniquely African terrestrial groups had appeared, descended from primitive Eurasiatic stocks. Even though present along the southern shores of Tethys, most of these groups remained confined to the continent, giving Africa its greatest endemism of the Cenozoic. In figures 30.5 and 30.6 the degree of endemism for terrestrial mammals is shown. Family endemism· reached 81% by early Oligocene times and for genera the figure was 84%; these values may be higher still when a fossil record for deeper continental regions becomes available. Of the three nonendemic families, one is a bat, leaving only two, Anthracotheriidae and Hyaenodontidae, for which a sweepstakes dispersion across the Tethys Seaway must be postulated. A sudden influx of mammals from northern continents during the Oligocene-Miocene transition dramatically marked the end of Africa's isolation. Family endemism dropped to 33% and generic endemism to 62%. More than half of the new families are micromammals that evolved rapidly to new family status, accounting for much of the observed endemism. If only larger mammals are considered, however, endemism drops to about 25%, nearly one-fourth that of the Oligocene.

Patterns ofFaunal Evolution

609

Subsequent patterns of familial evolution involved minor reduction in endemism, reaching a low of 20% by the Pliocene. During the later Pleistocene and early Holocene family endemism rose slightly again to 23%. But rather than resulting from evolution or immigration of new groups, this rise was due principally to extinction in Eurasia of families such as hippopotamids and giraffids, giving the impression of an increase in African endemics. This has also been the reason why modern Ethiopian mammals appear to represent a relict Pliocene fauna by European standards. Generic-level endemism reached its lowest value of 46% in the middle Miocene, probably as a result of continued relatively unhampered immigration from the north. Between the Pliocene and Recent, generic endemism rose again, reflecting progressive in situ evolution of numerous stocks that entered earlier. Toward the end of the Pleistocene and in the Holocene endemism reached its highest level since Oligocene times, and presumably followed development of the Sahara Desert, which served as a major barrier against north-south dispersion (Coryndon and Savage 1973, p. 134). This idea is supported by the fact that North Africa, partially isolated from the remainder of the continent, has continued to exchange faunal elements with Eurasia to the present day (Cooke 1972, p. 122).

Faunal Turnover In figures 30.5 and 30.6 the entire Quaternary is plotted as a single point for turnover and extinction curves (shaded areas) in order to make time intervals for Cenozoic subdivisions more nearly equal. Even so, later periods tend to be shorter than earlier ones, but their better record compensates somewhat for these inequities. Broken lines give values plotted for subunits within the Pleistocene, but it should be noted that the short time intervals involved are mainly responsible for the lower values here. Viewing these data in terms of faunal replacement, we see on the generic level extremely high rates of turnover in which 80 to 90% of the fauna was renewed between Eocene and Oligocene times and again between the Oligocene and early Miocene. As discussed above, the essential difference between these two transitions was a turnover caused by local evolution and involving mainly autochthonous and endemic groups in the former event but encompassing mostly immigrants and their slightly modified endemic descendants in the latter. By middle and late Miocene times less than onehalf of the African fauna was renewed in each sue-

1 N 3 0 3 H

3 N 3 0 0 i S I 3 1 d Ί

I

IN

I

3

D o o n o

3 N 3 0 0 I W

O O l I d 1

1

W

I

0 0 3

3

ο in

CO φ

•s ε

ε CO CO

ο •«t

ο

CO D

CO 0) ο

Ü .0

c

CO kCO > Μ Ο

(3 υ

Ο .S*

-c Ό -3 £ § Jce et s α Η S J Propalaeochoerus Listriodon Kubanochoerus Lopholistriodon Xenochoerus Sanitherium Nyanzachoerus Notochoerus Metridiochoerus Kolpochoerus Stylochoerus Potamochoeroides Phacochoerus Sus Potamochoerus Hylochoerus Tayassuidae Percarichoerus Hippopotamidae Hexaprotodon Hippopotamus Tragulidae Dorcatherium Camelidae Camelus Cervidae Cervus Dama Megaloceros Palaeomerycidae Climacoceras Propalaeoryx Canthumeryx Giraffidae Zarafa Palaeotragus Giraffokeryx Samotherium Okapia Giraffa Prolibytherium cf Helladotherium Sivatherium Bovidae Taurotragus Tragelaphus Eo tragus Protragocerus Mio tragus Syncerus Ugandax Pelorovis Simatherium

9

? XX

XX XX — X X X X X X X XX — X X X X XXX XX XXX X X X X X X X X X XX X X X X X XX X -

X

X XXX X _ χ - X X X X X ? X X X X X 9 X X X -

-

X X X X

XX X X X X X — X XXX XX —

φ .2g β φ >> Ό I f f ? 3 > Ό 8 >> Ό Ο Ο 'S Ό CS .BP ~ S α s ω « § J OH W S J ο Basilosauridae Dorudon Prozeuglodon Archaeoceti indet. "Zeuglodon" Zaphiidae gen. indet.

ο Acrodelphidae Schizodelphis Odontoceti indet. "Delphinus" Balaenopteridae Balaenoptera

X X X

| S . ^ ® >> Ό & χ Β Ό «β a ω a §

2

*** φ Oh χ CL, ϊ ί CO ω 5

ΐ .φ λ J

X X X

Table 30.2 Distribution of endemic land mammal families in the Old World. Data are modified from Simpson 1 9 4 5 and the present volume.

Pleistocene Pliocene Miocene Oligocene Eocene

Table 30.3

African endemics

Gurasiatic endemics

Present in both

Total Old World families

10 10 22 12 3

9 14 16 43 40

44 42 39 4 1

63 66 77 59 44

Total families in Africa

Percent Old World families in Africa

Percent Old World families endemic to Africa

54 51 61 16 4

86 79 79 27 9

15 15 29 20 7

Cenozoic distribution of land mammal families in Africa. Φ

Φ

Φ

Total number First appearances Last appearances Endemics

cφ υ Ο

c Φ ο ο w

BS

4 4 2 3

16 14 2 13

β Φ υ ο

β Φ ο Ο

•>§>

_φ Ό Ό

43 29 6 14

49 12 5 11

Ν

α Φ u Ο

Φ

C Φ Ο

ο

cd hJ 50 6 2 10

φ α 8 ο

φ ε

51 3 1 10

54 4 1 11

>>

'S Μ

Φ

ε φ Ό Ό

54 1 1 11

Φ

S φ υ Ο Φ

ε φ -Μ π) J 53 0 2 11

Maglio

Patterns ofFaunal

Evolution

619

Table 30.4 Cenozoic distribution of land mammal genera in Africa. φ

Φ

Total number First appearances Last appearances Endemics

s υ ο Ο

c φ u ο

o> c Φ υ Ο W

ÖD

4 4 2 3

31 29 23 26

s φ υ Ο

Φ

cβ W

Ό

87 79 57 54

56 26 30 26

Cooke, Η. Β. S. 1968. Evolution of mammals on southern continents. II. The fossil mammal fauna of Africa. Quart. Rev. Biol. 43(3):234-264. 1972. The fossil mammal fauna of Africa. In Keast, Α., F. C. Erk, and B. Glass, edsEvolution, mammals and southern continents, Albany: State Univ. of N.Y. Press, pp. 89-139. Coryndon, S. C. and R. J. G. Savage. 1973. The origin and affinities of African mammal faunas. In Organisms and continents through time. London: Syst. Assoc. Publ. no. 9, Special Papers in Palaeontology 12:121-135. Crompton, A. W. 1964. A preliminary description of a new mammal from the upper Triassic of South Africa. Proc. Zool. Soc. Lond. 142:441-452. Crompton, A. W. and F. A. Jenkins, Jr. 1968. Molar occlusion in late Triassic mammals. Biol. Rev. 43:427-458. Dietrich, W. O. 1928. Brancatherulum n.g.—ein Proplacentalier aus dem obersten Jura des Tendaguru in Deutsch-Östafrika. Zentr. Mineral. Geol. Palaeontol. 1927B:423-426. Edwards, W. E. 1967. The late-Pleistocene extinction and diminution in size of many mammalian species. In Martin, P. S. and Η. E. Wright, Jr., eds., Pleistocene extinctions. The search for a cause, New Haven: Yale Univ. Press, pp. 141-154. Klein, R. G. 1972. The late Quaternary mammalian fauna of Nelson Bay (Cape Province, South Africa): its impor-

Ο

C Φ

υ ο CS

Φ

S φ

Ο

J

47 21 21 24

S φ U

ο φ

ο

Ό

References

C φ υ

φ

107 81 18 57

ε Οη >> «CS

142 53 33 72

Χ!

Τ3

Φ

C φ

υ

Ο

Εφ Φ -*->

(Β

J 119 10 12 63

121 14 11 65

e φ υ Φ Ρί 256 146

?

184

tance for megafaunal extinctions and environmental and cultural changes. Quatern. Res. 2:134-142. Leakey, L. S. B. 1965. Olduvai Gorge 1951-1961. Cambridge: Cambridge Univ. Press. Martin, P. S. 1966. Africa and Pleistocene overkill. Nature 212:339-342. 1967. Prehistoric overkill. In Martin, P. S. and H. E. Wright, Jr., eds., Pleistocene extinctions. The search for a cause, New Haven: Yale Univ. Press, pp. 75-120. Patterson, B. 1975. The fossil aardvarks (Mammalia: Tubulidentata). Bull. Mus. Comp. Zool., Harvard 147: 185-237. Savage, R. J. G. 1967. Early Miocene mammal faunas of the Tethyan region. In Adams, C. G. and D. V. Ager, eds., Aspects of Tethyan biogeography. London: Syst. Assoc. Publ. no. 7, pp. 247-282. 1969. Early Tertiary mammal localities in southern Libya. Proc. Geol. Soc. Lond. 1657:167-171. Simpson, G. G. 1945. Principles of classification and a classification of mammals. Bull. Am. Mus. Nat. Hist. 85: 1-360. —.. 1965. The geography of evolution. Philadelphia: Chilton Books, 249 pp. Van Couvering, J. A. 1972. Radiometric calibration of the European Neogene. In Bishop, W. W. and J. A. Miller, eds., Calibration of hominoid evolution, Edinburgh: Scottish Academic Press, pp. 247-271.

Index

Aardvark. See Orycteropodidae; Orycteropus; Tubulidentata Aardwolf. See Proteles Aberdares, 6 Aboukir, Algeria: Bovidae, 556 Abu Hugar, Sudan: Bovidae, 548, 550 Acacia, 8, 20, 230, 336 Aceratherium, 371, 373-374, 376; acutirostratus, 371, 373-374, 376; incisivum, 373-374, 376; lemanense, 374, 376 Acheulian: Cave of Hearths, 207; and Cervidae, 497; Chemoigut Basin, 32; and Homo erectus, 225-227; Hopefield, 40; Kanjera, 35; Kapthurin Beds, 30, 204; Laetolil Beds, 158; Lake Ndutu, 205; the Maghreb, 199; Olduvai Gorge, 199, 225-226, 230; Olorgesailie, 108; Peninj group, 34; Sidi Abderrahman, 35; Sterkfontein, 172; Swartkrans cave system, 182, 189; Vaal River Basin, 39 Achtiaria, 514 Acinonyx, 2, 260, 262\jubatus, 260; in savanna biome, 7 Acomys: in savanna biome, 7; subspinosus, 11 Acrodelphidae, 591-592, 596 Adapis, 116 Adaptive grade and structural-functional zones: of Australopithecus, 218-222; of Homo, 222; of Ramapithecus, 217-218 Addax, 12, 552; nasomaculatus, 10, 552, 554 Adi Ugri, Eritrea: Deinotheriidae, 321 Aegodontia, 564, 565, 567 Aegyptopithecus, 115, 122, 123, 124, 127, 130-132, 143·, Aeolopithecus compared, 139\Apidium compared, 113; Oligopithecus compared, 128-129; relationships of, 116; zeuxis, 124, 143 Aeolopithecus, 124, 128, 129, 139, 143, 144; chirobates, 124, 129, 139 Aepyceros, 8, 9, 554, 558, 569; melampus, 555, 558 Aepycerotinae, 554 Aethomys, 7 Aetiocetidae, 595 Aetiocetus, 595 Afar, Ethiopia, 228; Bovidae, 541, 569; (Alcelaphinae, 558; Bovinae, 546, 548; Hippotraginae, 552); Hippopotamidae, 492 Africanomys, 82, 83 Africanthropus njarascensis, 210 Afrochoerus, 475; nicoli, 475 Afrocricetodon, 78, 80, 81 Afrocricetodontinae, 75, 78, 79, 80, 81, 85 Afrocyon, 253, 254 Afropterus gigas, 65, 66 Afrosmilus, 260-261, 262 Agorophiidae, 595 Agorophius, 595 Agrange, Chad: Anthracotheriidae, 424 Agriotherium, 37, 255, 608 Agulhas Bank, South Africa: Cetacea, 585 Αϊη Beida, Algeria: Equidae, 412 Αϊη Boucherit, Algeria, 35, 36; Bovidae, 560; Equidae, 404 Αϊη Brimba, Tunisia, 36; Bovidae, 560, 563

Αϊη el Bey, Algeria: Equidae, 404; Suidae, 464 Αϊη el Hadj Baba, 36; Equidae, 394, 416 Αϊη Hanech, Algeria, 35-36; Bovidae, 554, 563; Equidae, 404, 409; Proboscidea from, 358 Αϊη Jourdel, Algeria: Bovidae, 545, 556; Cercopithecidae from, 106; Equidae, 404 Αϊη Tit Mellil, Morocco; Suidae from, 469 Albert National Park, 12 Alcelaphinae, 541, 554-558, 564; Laetolil, 565; Langebaanweg, 565; Olduvai Gorge, 569; Sterkfontein, 228; Swartkrans, 231; Waldi Natrun, 568 Alcelaphini, 8, 541, 554-558, 563, 567 Alcelaphus, 8, 9, 10, 11, 554, 556, 557, 565; buselaphus, 554, 556; howardi, 555; lichtensteini, 554, 556 Alcicephalus, 517 Alengerr Beds, Kenya, 30; Deinotheriidae, 321; Rhinocerotidae, 373, 374 Algae, 227 Algeria: Bovidae, 541, 554; Carnivora, 249 (Canidae, 254; Mustelidae, 255; Ursidae, 255); Cervidae, 496, 497; Equidae (Equus, 404, 408, 412; Hipparion, 394, 399); fossil discoveries, 23; geological development, 22; Giraffidae, 509 (Giraffinae, 522; Palaeotraginae, 516, 517, 518; Sivatheriinae, 523, 524, 525); Miocene fossils, 27; PliocenePleistocene fossils, 35-36; Proboscidea, 353, 354, 358, 359; Suidae, 455; Tubulidentata, 273 Ali Bacha, Algeria: Equidae, 412 Allenopithecus, 102 Allotheria, 46 Alouatta, 141 Amado, Ethiopia: Hippopotamidae, 492 Amblychilus, 578 Amblypoda, 330 Amblysomus hamiltoni, 64 Amebelodon, 340, 343, 344 Ameki Formation, Nigeria, 25; Cetacea, 584, 585, 587 Amerhippus, 404 Amirian: Bovidae, 549; Hominidae, 159, 197, 198 Ammodorcas, 559, 567 Ammotragus lervia, 562 Amphechinus, 60; rusingensis, 60 Amphichoerus, 443 Amphicyon, 254 Amphicyoninae, 253, 255, 265 Amphilestidae, 46; relationships of, 5253 Amphimerycidae, 536 Amphipithecus, 112; mongaungensis, 120 Amphitherium, 54 Amphitragulus, 497, 564 Anagalidae, 57 Anancinae, 339, 342, 344, 345-349, 360, 607 Anancus, 339, 343, 345-349, 350, 360; arvernensis, 345, 348; osiris, 35, 36, 345-348, 359, 360 Anaptomorphidae, 91 Anasinopa, 252, 253 Anatolia: Hominidae, 218; Tubulidentata, 275

622

Anchitherium, 288 Ancylotherium, 369; hennigi, 368, 369; pentelicum, 369 Andrewsarchus, 593 Andrewsimys, 78 Anfatian Stage, 203 Anglocetus beatsoni, 595 Angola: Cercopithecidae, 103; wildlife of, 12 Anomaluridae, 2, 5, 6, 74, 78, 81; distribution, 69 Anomaluroidea, 74, 80, 85, 86 Antarctica: Cetacea, 596 Anteaters: scaly, see Pholidota; true, see Myrmecophagidae Antelopes, 9, 402, 415, 568. See also individual species. Antherurus, 79 Anthracokeryx, 430 Anthracotheriidae, 25, 26, 27, 297, 311, 423-433, 605, 606, 607, 609; general characteristics, 423-424; Hippopotamidae compared, 483, 486, 489; relationships, 430-433 Anthracotherium, 430 Anthropoidea, 100-117, 120-144, 165 Antidorcas, 8, 10, 556, 559-560, 562, 563, 567, 568; australis, 559, 568; bondi, 550, 559, 568; marsupialis, 559; recki, 559-560 Antilocapra, 504 Antilocapridae, 498, 503-504, 536 Antilope, 559, 561, 565, 567; cermcapra, 559, 561; maupasii, 550; selenocera, 550; sivalensis, 554; subtorta, 561 Antilopinae, 228, 231, 541, 558-561, 564, 569 Antilopini, 8, 541, 559-561, 565, 567 Ants and termites, 275, 276, 277 Aonyx, 5, 255, 256; congica, 5; capensis, 256 Aotus trivirgatus, 113 Aper aethiopicus, 467 Apes, 1, 2, 5, 6, 120-144, 151; diagnoses, 127-141; habitats, 142; history of study, 121-127; paleoenvironments, 143; pongid distributions, 141-143; and Ramapithecus compared, 152. See also Anthropoidea Aphelops, 372 Aphronorus, 251 Apidium, 113-114, 122; dental formula, 100; Fayum, 112, 120, 123, 124; moustafai, 113, 124; Oligopithecus compared, 128; Parapithecus compared, 115, phiomense, 113, 114, 122; relationships, 101, 115-116, 124 Apodemus, 84 Apternodontidae, 63 Apterodon, 251, 252, 604; saghensis, 252 Aquitanian: Bovidae, 563; Deinotheriidae, 321; Rhinocerotidae, 374; Sirenia, 577; Suidae, 443 Arabia: Bovidae, 565; geological development, 22 Archaeoceti, 582, 583, 584, 585-591, 598; morphology and habitus, 596-597; phylogenetic position, 592- 595 Archaeopteropus, 66 Archaeosiren, 575 Archaeotrigon, 54 Archidiskodon·. africanavus, 352, 353; an-

Index

drewsi, 357; broomi, 356, 357; exoptatus, 353, 355, 356;griqua, 355, 356; hanekomi, 356, 357; Imperator, 357; loxodontoides, 353; meridionalis, 352, 353, 357; milletti, 353;planifrons, 352, 353; planifrons nyanzae, 357; recki, 355; sheppardi, 356, 357; subplanifrons, 357; transvaalensis, 356, 357; vanalpheni, 353; yorki, 357 Arctocebus, 93, 95, 98; calabarensis, 93, 97; pusillus, 263 Arctocyonidae, 56, 250 Argentina: Carnivora, 263; Triassic deposits, 53 Arid Zones, 4, 7; biomass, 10; inhabitants, 9-10; vegetation, 18-20 Aristida, 18, 20 Armadillo. See Dasypodidae; Dasypus Arsinoitherium, 282; andrewsi, 280, 281, 282; zitteli, 279, 280, 281, 282 Artiodactyla, 6, 7, 11, 113, 605, 606, 607; Anthracotheriidae, 423-433; Bovidae, 540-569; Camelidae, 537-538; Cervidae, 496-497; Fayum, 311; Gelocidae, 536; Hippopotamidae, 483-495; numbers of, 2; Palaeomerycidae, 497-506; relationship with Cetacea, 592; Suidae and Tayassiudae, 435-479; Tragulidae, 536-537 Arvicanthis, 84 Arvicolinae, 84 Asellia vetus, 65 Asia: Anthracotheriidae, 423, 430, 431, 433, 486; Anthropoidea, 102, 112, 127, 144; Bovidae, 548, 552, 564, 567; Carnivora, 255, 262, 263, 264-265; Cercopithecidae, 102, 112; Cetacea, 584; Chalicotheriidae, 368, 369; cultural associations of Hominidae, 225; Deinotheriidae, 315, 321, 322, 323; Equidae, 389, 396, 416, 417, 418; faunal tie with Africa, 112, 605, 606, 607, 608, 611, 614; Gelocidae, 536; Giraffidae, 516, 526, 528-529; Hippopotamidae, 486, 489; Hominidae, 216, 222; Hyracoidea, 284, 289, 308, 309, 310; Lorisidae, 98; Mesozoic mammals, 53, 54; Neanderthal population, 211; Palaeomerycidae, 498, 503, 506; Pholidota, 268-270; Proboscidea, 339,344, 359, 360, 364; Rhinocerotidae, 374; Rodentia, 69, 74, 76, 80-82, 85-86; Suidae, 436, 444, 445, 447, 448, 451, 452, 453; Tayassuidae, 476; Tragulidae, 536, 537; Tubulidenta, 274 Asinus, 379, 402, 404, 412-413, 417418; burchelli, 408 Asses. See Equus Atelerix, 61; albiventris, 61; algirus, 61; major, 60, 61; sclateri, 61 Ateles, 125, 141 Aterian Industry, 211, 212 Aterir Beds, 30; Rhinocerotidae, 375 Atheria, 46 Atherurus, 5 Atilax, 257; paludinosus, 257 Atlantic Ocean, geological development, 22, 23 Atlantoxerus, 82\getulus, 82; huvelini, 82; tadlae, 82 Atlas Mountain Range, 17, 18, 20, 22, 23, 255

Aulaxinuus libycus, 102 Australia: Cetacea, 584; prehistoric extinctions, 11 Australopithecus, 158, 166, 167-185, 189, 199; adaptive grade and structuralfunctional zone, 218-222; africanus, 37-38, 163, 164, 165-167, 170-171, 173, 224, 225, 288 (and A. crassidens compared, 179, 180, 181, 182; and A. robustus compared, 184, 185; characteristics, 170-171, 173; distribution, 171-173; fossil occurrences, 220-222, 223; and Homo compared, 186, 188, 189, 191, 192, 193, 194; paleoenvironment, 228, 229); boisei, 169, 173-179, 219, 223, 224, 225 (and A. africanus compared, 172; and A. crassidens compared, 182; and A. robustus compared, 184, 185; characteristics, 173-176, 179, 180, 181; distribution, 176-179; Hadar Formation, 165; and Homo compared, 188; Koobi Fora Formation, 193; paleoenvironment, setting, 228-229, 230); coexistence with Homo, 222-223, 230; crassidens, 179-183, 189, 219, 222,223 (characteristics, 179-181, 184, 185; distribution, 181-183; paleoenvironment, 229); extinction, 223; and Hominidae compared, 161, 163, 165; and Homo compared, 186, 187, 188, 191, 192, 193, 218, 219, 222; and H. erectus compared, 194, 195, 200; paleoenvironment, 227-229; and Ramapithecus compared, 147, 150, 151, 152, 219; robustus, 169, 172, 183-185, 188, 220, 223, 225 (paleoenvironment, 229) Austria: Sirenia, 576; Suidae, 447 Austrolagomys, 84 Aye-aye. See Daubentonia Baard's Quarry, Langebaanweg, South Africa: Bovidae, 552; Equidae, 392, 393, 406 Babirussa, 452, 475 Baboons. See Papio Bacitherium, 536, 564 Badger, honey. See Mellivora Balaena mysticetus, 596 Balaenoptera, 594, 596, 597 Bankies, South Africa: Equidae, 411 Barbary, 249; Carnivora, 262 Baringo, Kenya, 32-33; Bovidae, 541; Carnivora, 255; Equidae, 389, 391, 395; Giraffidae, 517, 520, 529; Hippopotamidae, 486, 487, 489; Hominidae, 156, 158, 172; Palaeomerycidae, 500; Suidae, 465 Bartonian age, 25; Cetacea, 588; Moeritherioidea, 334 Barypoda, 330 Barytheria, 330, 336, 337 Barytheriidae, 329-330; relationships, 330 Barytherioidea, 330; extinction, 605 Barytherium, 25, 327, 328, 329-330, 604; grave, 329-330 Basilosauridae, 587-591, 593, 597 Basilosaurinae, 585, 588, 590-591, 593, 596, 598 Basilosaurus, 587, 588, 591, 596; cetoides, 595, 596

Index

Bathyergidae, 2, 7, 10, 11, 70, 71, 74, 79, 80, 81, 82, 85; distribution, 69 Bathyergoidea, 71, 74, 77-78, 79, 80, 84, 86; distribution, 69 Bathyergoides, 78, 79, 80 Bathyergoididae, 71, 74 Bathyergus suillus, 11 Bats. See Chiroptera; Microchiroptera; and specific families Bdeogale, 5, 257; crassicauda, 257,jacksoni, 257; nigripes, 257;puisa, 257; tenuis, 257 Bears. See Agriotherium; Ursidae Beatragus, 554, 558, 569; antiquus, 558; hunteri, 554-555, 558, 568 Beaufort Series, 20-21 Bedeil, Somalia: Sirenia, 577 Beglia Formation, Tunisia, 26, 27; Bovidae, 563, 568; Carnivora (Canidae, 254; Creodonta, 253; Felidae, 262; Hyaenidae, 257, 260; Mustelidae, 255); Rodentia, 83; Sirenia, 578 Bel Hacel, Algeria; Bovidae, 554; Proboscidea, 358 Beni Mellal, Morocco: Bovidae, 563, 568; Carnivora, 255, 260; Chiroptera, 65, 66; Insectivora, 60, 61; Lagomorpha, 84; Rodentia, 70, 76, 79, 80, 82, 83, 85 Benicerus, 568; theobaldi, 563 Benue Rift, 22 Berberomys, 84 Bethlehem, Jordan: Equidae, 395; Proboscidea, 364 Bienotherium, 53 Biotic zones, 4 - 1 1 Birds: aquatic, 227; Miocene fossils, 26 Birket el Qurun Formation, Cetacea, 588, 589, 591, 593, 598, 599 Blackbuck, Indian. See Antilope cervicapra Blastomerycini, 497 Blastomeryx, 497, 504 Blaubank Valley, South Africa, 227 Bled ed Douarah, Tunisia, 26; Anthracotheriidae, 433; Equidae, 391; Proboscidea, 342, 345 Blesbok. See Damaliscus dorcas Bloembos, South Africa: Bovidae, 552 Bloembosch, South Africa: Equidae, 406 Bloubok. See Hippotragus leucophaeus Bochianga, Chad: Camelidae, 538 Bololo, Congo, 27 Bolt's Farm, Transvaal: Bovidae, 559; Cercopithecidae, 103; Chiroptera, 65, 66; Hyracoidea, 311; Insectivora, 57, 60, 61, 63, 64; Proboscidea, 355; Suidae, 470, 473 Bone, Algeria: Hippopotamidae, 489 Bongo. See Tragelaphus eurycerus Bontebok. See Damaliscus dorcas Bontequaggas. See Equus burchelli Boocercus, 6, 543 Böodontia, 564, 565, 566, 567 Border Cave, South Africa: Bovidae, 559; Homo sapiens, 213, 214, 215 Bos, 548, 549, 565, 566, 567; makapaani, 562; primigenius, 549 Boselaphini, 540, 546-547, 552, 554, 564, 565, 566, 567 Boselaphus: probubalis, 556; tragocamelus, 546 Boskop race, 213

Bothriodon, 430 Bothriogenys, 299, 424-427, 428, 430, 431, 433; africanus, 427, 431; andrewsi, 427; fraasi, 427; gorringei, 299, 427; parvus, 427; rugulosus, 427 Botswana, 12, 13, 14 Bou Hanifa Formation. See Oued el Hammam Bovidae, 2, 6, 8, 9, 10, 35, 36, 37, 536, 540-569; adaptive radiation, 112, 402, 493; character evolution, 568; classification, 540-541, 564-567; earliest in Africa, 563-564; extinctions, 11, 568; faunal evolution, 567-568, 606, 607, 608; Fayum, 311; Fort Ternan, 500; fossil localities, 541-543; general characteristics, 498, 503-504, 540; and Giraffidae compared, 511, 523; Kromdraai, 185; Makapan Limeworks, 172; Miocene, 26, 27, 30, 32; origins, 563564; paleoenvironments, 227, 228, 229; South Africa, 232; Sterkfontein, 191; Swartkrans, 182, 200; systematics, 543-564 Bovinae, 8, 540, 543-549, 564 Bovini, 8, 540, 547-549, 553, 566, 567 Bovoidea, 504, 564 Brachyodus, 424, 430, 431; fraasi, 424 Brachypotherium, 371-373, 374, 376; aurelianense, 372; brachypus, 372; goldfussi, 372-373, 376; heinzelini, 371, 372, 373, 376; lewisi, 371, 372, 376; snowi, 372, 373, 376; stehlini, 372 Brachystegia, 20 Bramatherium, 529 Brancatherulum tendaguruense, 22, 46, 54, 604; age and associated fauna, 54; classification, 54 Branisamys, 78 Broken Hill, Zambia, 40; Bovidae, 556, 560; Carnivora, 262; Equidae, 414; Giraffidae, 524; Hominidae, 154, 159 Brontotheriidae, 289 Brunhes-Matuyama Epoch, 199 Brunhes Normal Epoch, 199, 203 Bubalus, 548, 566; andersoni, 548; baini, 548; nilssoni, 548; vignardi, 549 Budorcas, 567; taxicolor, 561 Buel Haderaut, Libya, 25; Sirenia, 576 Buffalo. See Bovinae; Bovini; Syncerus Buffalo Cave, South Africa: Bovidae, 562 Bugti, Pakistan: Carnivora, 253; Chalicotheriidae, 368, 369; Deinotheriidae, 315, 322; Suidae, 447 Bukwa, Uganda, 28; Anthropoidea, 138, 142; Deinotheriidae, 321; Hyracoidea, 302, 308; Rhinocerotidae, 372, 374, 376; Rodentia, 70; Tragulidae, 537 Bularchus arok, 548-549 Bulla Regia, Tunisia: Insectivora, 63 Bunohyrax, 287, 289, 294-297, 304, 312; affinis, 288, 295, 296; characteristics, 285, 299, 302; fajumensis, 291, 292, 293, 294, 295-296, 297, 299; and Geniohyus compared, 290, 291, 292; major, 294, 296-297, 298, 299 Bunolagus, 10 Bunolistriodon, 444, 445; lockharti, 444 Bunolophodon, 341, 342 Bunomastodontidae, 340 Burdigalian: Anthracotheriidae, 431; Bovidae, 563, 564; Deinotheriidae, 315,

623

320, 321, 323; faunal evolution, 605; Proboscidea, 342; Rhinocerotidae, 372, 374; Sirenia, 576, 577; Suidae, 442, 443, 448, 449 Burma, 112; Anthracotheriidae, 423, 430, 433 Bushbabies. See Galago Bushbucks. See Tragelaphus Bushman Rock Shelter, Transvaal: Homo sapiens, 213, 215 Bushmen, 13, 213, 215 Bushpig, 7. See also Potamochoerus porcus Butselia, 63 Buxton, 37; Hominidae, 154 Cairo, Egypt: Cetacea, 585, 587 Callis, Somalia: Sirenia, 577 Callithrix, 113, 128 Cambrian Period, 20 Camel, bactrian. See Camelus bactrianus Camelidae, 537-538, 612 Camelopardalis: affinis, 518, 529 Camelus: bactrianus, 537, 538; sivalensis, 538; thomasi, 538 Canidae, 2, 7, 10, 11, 253-255, 263, 606 Caninae, 255 Canis, 2, 7, 253, 254; adustus, 253, 254; africanus, 254; antiquus, 254; aureus, 253, 254; gallaensis, 253; lamperti, 253; lupus, 254; mesomelas, 253, 254; simensis, 10, 253 Canthumeryx, 498, 500-502, 503, 505506; sirtensis, 500-502, 506, 513 Cape Columbine, South Africa: Cetacea, 585 Cape of Good Hope, South Africa: Equidae, 409 Cape Mountain Range, 17, 18, 20 Cape Province: Bovidae, 567, 568, (Antilopinae, 559; Bovinae, 543; Hippotraginae, 553); Cetacea, 584, 585; Equidae, 403, 409-410, 412; Hominidae, 37; Proboscidea, 360 Capra, 562, 563; walie, 10, 557 Capreolus, 497, 514; matheronis, 497 Caprinae, 541, 558, 561-563, 564, 565, 567, 568 Caprini, 541, 562-563, 564, 565, 567 Capsian Stage: Equidae, 412 Carboniferous Period, 22 Carcar, Somalia: Sirenia, 577 Carcharodon megalodon, 576 Cardioderma, 65; cor, 66 Caribbean: Sirenia, 573, 578 Carnivora, 5, 7, 11, 25, 26, 27, 37, 159, 185, 219, 227, 249-265, 270, 277, 584, 606, 608; Canidae, 253-255; Creodonta, 250-253; distribution, 263; Felidae, 260-262; Hyaenidae, 257-260; Mustelidae, 255-256; origins, 263-265; Phocidae, 262-263; Ursidae, 255; Viverridae, 256-257 Casablanca, Morocco: Cervidae, 497 Castor fiber, 376 Catarrhini, 2, 100, 112, 114, 116, 117, 120 Cave breccias, Transvaal: Hominidae, 181-182; Insectivora, 57; Suidae, 466, 469 Cave of Hearths, South Africa: Hominidae, 159, 206, 207, 210

624

Cave Sandstone, Lesotho, 47 Cavia, 286 Caviomorpha, 69, 70, 78, 79; relation with Phiomorpha, 86-87 Ceboidea, 100, 101, 112, 113; Parapithecidae distinguished from, 112-113 Cebus, 113 Cenchrus, 20 Central Africa, 36-37 Cephalophinae, 6, 8, 11, 13, 540, 549550 559 564 Cephalophini, 540, 549-550, 566, 567 Cephalophus, 549-550; caeruleus, 549, 550; monticola, 6, 549, 550; parvus, 549; pricei, 545, 558; sylvicultor, 6 Ceratotherium, 7, 35, 375, 376;praecox, 37, 371, 375, 376; simum, 36, 371, 375, 376; simum germanoafricanum, 375 Cercamonius, 116 Cercocebus, 103, 104; dentition, 102 Cercopithecidae, 1, 2, 5, 6, 101-112, 608; origins, 112, 116; Parapithecidae distinguished from, 112-113; relationships, 116; taxonomic characteristics, 101-102

Cercopithecinae, 5, 102; taxonomic characteristics, 101-102 Cercopithecini, 101, 102 Cercopithecoidea, 37, 100, 101, 112, 113, 116, 121; Fayum, 120; Kromdraai, 185, 220; Lothagam, 227; Makapan Limeworks, 222; Swartkrans, 181, 219; Sterkfontein, 221 Cercopithecoides, 108, 109, 110; molletti, 109; williamsi, 108, 109 Cercopithecus, 102, 117; aethiops, 7; patas, in savanna biome, 7; talapoin, 115, 116 Cervidae, 11, 126, 496-497, 498, 523, 536, 608 Cervoidea, 504 Cervus: algericus, 497; elaphus, 496-497, 500; elaphus barbarus, 496, 497; pachygenys, 497 Cetacea, 23, 25, 27, 263, 574, 582-599, 604, 605; Acrodelphidae, 591; Basilosauridae, 587-591; morphological specializations, 582-583; morphology and habitus of Archaeocetes, 596; origins and phylogenetic position, 592-595; Protocetidae, 585-587; problems in paleontology, 597-599; Ziphiidae, 592; Zoogeographie significance, 595-596 Cetotheriopsis, 595 Chad Basin, 12, 23, 37; Anthracotheriidae, 424, 429, 433; Equidae, 399; Giraffidae, 522, 528; Proboscidea, 345, 350, 351, 352, 358 Chalicotheriidae, 27, 227, 288, 289, 368370, 606, 608 Chalicotherium, 368-369,grande, 368, 369;pilgrimi, 368, 369; rusingense, 368, 369; wetzleri, 369 Chalouf, Isthmus of Suez: Sirenia, 576 Chari Tuff, 33, 179; age, 200 Cheetah. See Acinonyx Cheirotherium, 577 Chelif Valley, Algeria, 36 Chelmer, Rhodesia, 40; Bovidae, 559 Chemeron Formation, Kenya, 30, 37; Cercopithecidae, 109; Chalicotheriidae, 368, 369; Deinotheriidae, 323; Equidae,

Index

395; Hominidae, 158, 172; Proboscidea, 352, 356; Rhinocerotidae, 375; Suidae, 456, 460, 464 Chemoigut Beds, Kenya, 30-32; artifacts, 225; Deinotheriidae, 323; Hominidae, 178 Cherengani Hills, 27 Cherichera, Tunisia: Deinotheriidae, 321; Proboscidea, 342, 343, 345 Chersenotherium, 517 Chesowanja, Kenya: Hominidae, 158; Suidae, 475 Chevrotains. See Hyemoschus; Tragulidae; Tragulus Chilotheridium, 374, 376; pattersoni, 371, 374, 376 Chilotherium, 371, 374 Chimpanzee. See Pongidae China: Bovidae, 563, 565; Carnivora, 260, 262; Equidae, 389, 396-397; Giraffidae, 529; Hippopotamidae, 483; Hominidae, 194, 196; Hyracoidea, 288, 289, 307, 308, 311; Suidae, 444, 446; Triassic, 53 Chinji, India: Bovidae, 565; Carnivora, 260 Chios Island, Greece: Rodentia and Lagomorpha, 71 Chiroptera, 1, 2, 4-5, 6, 10-11, 13, 56, 65-67, 604, 607, 608, 609, 612; Emballonuridae, 65, 66, 67; Hipposideridae, 65, 66, 67; of Madagascar, 11; Megadermatidae, 65, 66, 67; Molossidae, 66, 67; Myzopodidae, 65, 66, 67; Pteropodidae, 65, 66, 67; Rhinolophidae, 65, 66, 67; Vespertilionidae, 65, 66, 67 Chiwondo Beds, Malawi, 37; Suidae, 457, 458, 460 Chleuastochoerus, 444 Chlorotalpa spelea, 64 Choerolophodon, 339, 344, 345 Choeromorus, 443 Choeropsis, 6, 483; of Madagascar, 11 Choerotherium, 443 Chrysochloridae, 2, 4, 7, 10, 64-65 Chrysopogon, 20 Chrysotricha hamiltoni, 64 Civets. See Nandinia\ Osbornictis; Viverrinae Civettictis, 256; civetta, 256 Climacoceras, 497, 498, 499-500, 501, 502, 503, 505, 506; africanus, 499, 506 Climate of Africa, 4, 17-18; role in extinctions, 12, 613 Colobinae, 5, 102, 108-109, 110, 112; dentition, 111; taxonomic characters, 101-102

Colobus, 102, 108, 109, 110, 111; badius, 108; flandrini, 108; guereza, 108 Comiphora, 20, 230 Conde Soreadon, Tunisia: Proboscidea, 359 Condylarthra, 251, 277, 592; and Hyracoidea compared, 312 Congo, 12, 23; Anthracotheriidae, 424; Deinotheriidae, 315; Proboscidea, 343, 359; Rhinocerotidae, 371, 372; Sirenia, 575 Connochaetes, 8, 9, 554, 556, 569; africanus, 556; gnou, 10, 554, 555-556; gnou laticornutus, 556; taurinus, 554, 556, 569; taurinus prognu, 556

Conohyus, 448, 451 Constantine, Algeria, 35-36; Cervidae, 497 "Continental Intercalate," 22 Contrebandiers Cave, Morocco: Homo sapiens, 212-213 Cooper's Farm, South Africa: Hyracoidea, 311 Copopods, 597 Coriphagus, 251 Cornelia, South Africa, 40, 207; Bovidae, 541, 569 (Alcelaphinae, 555, 556, 557; Antilopinae, 561); Equidae, 399; Giraffidae, 527; Hippopotamidae, 491; Suidae, 462, 469, 470, 475, 476 Cornelian Faunal Span, 189; Bovidae, 541, 569; Hominidae, 159 Creodonta, 26, 249, 250-251, 252, 264, 265, 585, 592, 604, 605, 606; distribution, 263, 265; Hyaenodontidae, 277; relationship with Cetacea, 592 Cretaceous Period, 22, 23, 27 Cricetidae, 2, 5, 7, 10, 11, 81 Cricetodon, 83; atlasi, 83; ibericus, 83 Cricetodontidae, 70, 75, 77, 78, 80, 85, 86; North Africa, 82-83 Cricetomys, 5, 13, 79 Criotherium, 565, 567 Crocidura, 61, 62-63; bicolor, 61, 62, 63; bottegi, 62; flavescens, 62; hindei, 61, 63; hirta, 63; maurisca, 62; occidentalis, 62; taungensis, 61, 63 Crocidurinae, 61 Crocodilia, 26, 37, 53, 227, 491, 597 Crocuta, 258-259; crocuta, 258, 260 Crossarchus, 5, 257; alexandri, 257; ansorgei, 257;gambianus, 257; obscurus, 257 Cryptochloris, 10 Cryptoprocta: ferox, 257; Madagascar, 11; spelea, 257 Cryptoproctinae, 11, 257 Ctenodactylidae, 2, 10, 82-83; distribution, 69; North African, 82 Ctenodactyloidea, 76, 85, 86; distribution, 69 Cubango-Kalahari Basin, 23 Cucumis humifructus, 277 Cucurbitaceae, 277 Cultivation, and utilization of wild mammals, 13 Cultural associations of Hominidae, 223226

Cuvieroninae, 339 Cyclotis, 6 Cynictis, 257; penicillata, 257; selousi, 257; sengaani, 257 Cynocephalus atlanticus, 106 Cynodontia, 49; postcranial skeleton, 51; size, 52 Cyrenaica, Libya: Cervidae, 497 Cyrtodelphis, 591, 596; sulcatus, 591 Dahomey Gap, 6 Dakkamys, 83 Dama, 497; dama, 497; schaeferi, 497 Damalavus, 568; boroccoi, 562 Damaliscus, 8, 554, 556-557; antiquus, 557; dorcas, 8, 554, 556; dorcas dorcas, 11; lunatus, 9, 554; niro, 556-557, 558, 568 Damalops, 558,565;palaeindicus, 557, 567

Index

Dar bei Hamri, Morocco: Sirenia, 577 Dar-es-Soltan, Morocco: Hominidae, 210, 211-212 Dassies. See Procavia capensis Dasypodidae, 274, 276, 277 Dasypus, 276 Daubentonia, 11 Deer. See Cervidae; Ceruus ; Dama; Megaloceros; Moschus Deinotheriidae, 315-329, 336, 364, 608, 612; classification, 320; diagnosis, 316; evolutionary changes, 316-317; mode of life, 325-328; relationships, 328329 Deinotherium, 26, 317, 320, 323-324, 325, 327, 328; bavaricum, 320; bozasi, 323-324, 326, 328; cuvieri, 320, 321; giganteum, 320, 323, 324, 325, 327, 328; gigantissimum, 324; hobleyi, 321; hopwoodi, 323, 324; indicum, 323, 324; levius, 324; orlovii, 320 Delphinidae, 588 Delphinus sulcatus, 591 "Delphinus" vanzelleri, 592 Dendrohyrax, 5, 308, 309, 312 Dendromurinae, 7, 75, 81, 82, 83, 85 Dendropithecus, 125, 127, 129, 131, 139, 140-141, 144; macinnesi, 93, 125, 134, 138, 140-141, 142, 144 Denen Dora Member, Hadar, Ethiopia: Hominidae, 164 Dera Bugti, Pakistan: Anthracotheriidae, 431, 433 Desmostylia, 333 Developed Oldowan, 172, 177, 178, 181, 191, 200, 226; of Olduvai, 225, 226, 230; of Swartkrans, 225 Devonian Period, 20 Dhok Pathan Formation, India, 147-148; Bovidae, 552, 565, 566; Deinotheriidae, 315 Diamantohyus, 441, 446, 447; africanus, 447 Diamantomyidae, 71, 77, 78-79 Diamantomys, 71, 77, 78, 79 Diamond Fields of Southwest Africa: Rodentia, 70, 77 Dicerorhinus, 30, 374, 376; africanus, 374; leakeyi, 371, 373, 374, 376;primaevus, 374; sansaniensis, 374; schleiermacheri, 374; sumatrensis, 376; tagicus, 374 Diceros, 7, 374-375, 376; bicornis, 11, 371, 374-375, 376; douariensis, 374375; pachygnathus, 374 Dichobunidae, 452 Dicoryphochoerus, 449, 450, 451-452, 462; haydeni, 452; robustus, 452; titan, 452 Dicotylidae, 435 Dicrocerus, 496, 497 Diet: of Anthracotheriidae, 424; of apes, 143; of Cetacea, 597; of Giraffidae, 530; of Sirenia, 573, 574, 578; of Suidae, 446, 450-451, 453, 477, 479 Dietary hypothesis, 169 Dik-diks. See Madoqua Dinocerata, 330 Dinofelis, 37, 262; obeli, 262 Dinopithecus, 104-105; brumpti, 105; ingens, 104-105 Dinosaurs, 53

Diplomesodon, 63; fossorius, 61, 63;pulchellum, 63 Dipodidae, 10, 75-76, 83; distribution, 69 Dissacus, 592, 593; navajovius, 592 Dissopsalis, 252, 253, 265 Djebel Bel Hacel, Algeria, 36 Djebel Mallah, Tunisia, 36; Proboscidea, 345 Djebel M'Dilla, Tunisia: Proboscidea, 345 Djebel Mrhile, Tunisia, 26 Djebel Semene, Tunisia: Proboscidea, 345 Djerid area, Tunisia: Bovidae, 547 Djourab depression, Tunisia, 36 Docodonta, 46; relationships, 52 Dogs, hunting. See Lycaon Dolichohippus, 379, 402, 403, 404-408, 416 Dolichopithecus ruscinensis, 108 Doliochoerus, 442 Dolphins. See Cetacea Dor el Talha, Libya, 25; Barytheriidae, 329, 330; Carnivora, 253; Cetacea, 584; Moeritherioidea, 334; Proboscidea, 339, 340, 341; Sirenia, 577 Dorcabune, 537; porrecticornis, 552; triquetricornis, 550 Dorcatherium, 142, 536-537; chappuisi, 537; crassum, 537; libiensis, 537; majus, 537; minus, 537; naui, 537; parvum, 537;peneckei, 536;pigotti, 537; puyhauberti, 537; rogeri, 537; songhorensis, 537 Dorcatragus, 558 Dormice. See Muscardinidae Dorobo Tribe, 13 Dorudon, 585, 587, 588-590; elliotsmithii, 589, 590, 593-594, 598, 599; Fayum, 23, 598, 599; intermedius, 588, 589, 591, 593, 598; osiris, 587, 588589, 590, 591, 593-594; relationships, 593-594, 595; sensitivus, 590, 593-594, 598, 599; serratus, 588, 593; stromeri, 588, 589, 590, 591, 593-594, 599; zitteli, 589-590, 591, 593-594, 599 Dorudontidae, 585, 587-588 Dorudontinae, 588-590, 591; characteristics, 585, 597; relationships, 593, 595, 596, 598-599 Doryodon, 588 Douaria, Tunisia: Giraffidae, 523; Rhinocerotidae, 374 Dremotheriidae, 527, 528 Dremotherium, 497, 504, 516, 528, 564 Drills. See Mandrillus Dromedary, 537, 538 Dryopithecinae, 120, 125, 126, 127-139, 151; characteristics, 127, 160-161, 162; relationships, 144, 152 Dryopithecus, 127-128, 132, 181, 218, 606; and Aegyptopithecus compared, 130-131; africanus, 131, 132, 134; characteristics, 128, 129, 148; and Dendropithecus compared, 140; fontani, 132, 144; gordoni, 127, 135; laietanus, 144; major, 127, 135; mogharensis, 111; and Proconsul compared, 132, 134; punjabicus, 147; and Ramapithecus compared, 151, 152; sivalensis, 126; vancouveringi, 127, 136 Dugong, 574, 576, 578; dugon, 573, 587 Dugongidae, 574, 575-578 Dugonginae, 574, 578

625

Dugungus, 578 Duikers. See Cephalophinae; Cephalophini; Cephalophus Dwyka Series, 20 East Africa, 14, 27-30, 31-35, 605, 606. See also individual countries and localities East African Plateau, 17 East Indies: Sirenia, 573 East Rudolf. See East Turkana East Turkana, Kenya, 32, 33, 230; Bovidae, 541, 545, 549, 551, 553, 554, 560; Carnivora, 249, 254, 255, 257, 262; Cercopithecidae, 102, 104, 106, 108, 109, 110; Deinotheriidae, 323, 324; Equidae, 399, 400, 402; Giraffidae, 519, 520, 521, 524, 525; Hippopotamidae, 491-492; Hominidae, 158, 172, 179; paleoenvironment, 230; Proboscidea, 353, 355, 356; Suidae, 459, 464, 465, 472, 473, 474, 476; Tubulidentata, 273. See also Koobi Fora Formation Eastern Rift Valley, 92, 176; geological development, 27 Ecological features of Africa, 2-11 Echinosoricinae, 60 Ectopotamochoerus, 462; dubius, 464 Edentata, 270, 368 Egerkingen, 253 Egypt, 23, 25, 26, 35, 604; Anthracotheriidae, 424, 428; Bovidae, 562; Carnivora, 249; Cervidae, 497; Cetacea, 582, 584, 588, 591; domestication of wild animals, 12; Embrithropoda, 603; Equidae, 399, 412; Giraffidae, 509, 517, 525; Primates, 121-124; Proboscidea, 345; Sirenia, 574, 575, 576 Eidolon helvum, 5 Ekora Beds, Kenya, 32; Deinotheriidae, 323; Equidae, 390, 391, 396, 398, 401; Proboscidea, 354; Rhinocerotidae, 371, 375; Suidae, 457 El Haserat, Libya, 26 Eland. See Taurotragus Elandsfontein, South Africa, 40, 207; Bovidae, 541, 568, 569, (Alcelaphinae, 555, 556, 558; Antilopinae, 558-559, 560; Bovinae, 543, 548, 549; Hippotraginae, 550, 552, 553, 554); Carnivora, 254, 255, 262; Giraffidae, 526; Hominidae, 206-207, 209, 216; paleoenvironment, 231; Pholidota, 268; Proboscidea, 353; Rhinocerotidae, 375; Suidae, 462, 473 Elasmotherium, 328 Elephantidae, 338, 349-358; classification, 339, 342; evolution, 344, 360; feeding habits, 336 Elephantinae, 351, 364 Elephantoidea, 315, 325, 327, 328, 329 Elephantomys, 59; langi, 57 Elephants. See Proboscidea; and individual species Elephantulus, 59; antiquus, 57, 58, 59; brachyrhynchus, 59; broomi, 57, 59; dentition, 58; edwardi, 58, 59; fuscipes, 59; fuscus, 59; fuscus leakeyi, 57, 59; intufi, 59; rozeti, 59; rufescens, 59; rupestris, 59; vandami, 10

626

Elephas, 352, 354-375; africanavus, 352, 355, 356, 358; antiquus, 356, 362, 364; atlanticus, 353; ekorensis, 354-355, 356, 362, 364; exoptatus, 352, 353; iolensis, 35, 36, 354, 356-357, 362; maximus, 336, 354, 362; meridionalis, 355, 358; moghrebiensis, 358; namadicus, 356, 362, 364; planifrons, 353, 355, 358, 364; pomeli, 353, 354, 356, 357; recki, 34, 36, 353, 354, 355-356, 363, 364; zulu, 353, 355, 356 Elgon Mountain, Uganda, 6 Eliomys, 82, 83 Elisabethfeld, Southwest Africa: Palaeomerycidae, 502; Hyracoidea, 310 Ellobius, 84 Emballonuridae, 65, 66, 67 Embrithropoda, 279-282, 312, 603, 605 Endemism, 609 Enhydriodon, 255-256 Entelodon, 486 Environment of Africa, 17-20; climate, 4, 17-18; role in evolution, 11-12; vegetation, 17-20 Eocene Epoch, 23-25; Anthracotheriidae, 423, 430, 433; Barytheriidae, 329; Carnivora, 249, 251, 252, 253, 257, 263, 264; Cetacea, 582-585, 587-593, 595597; Chalicotheriidae, 368; Embrithropoda, 280; Equidae, 388; extinctions during, 612; geological developments, 22, 23; Hyracoidea, 287, 311, 312; Moeritherioidea, 334, 335; faunal evolution, 604, 605, 609, 611, 613; Pecora, 564; Pholidota, 270; Primates, 101, 112, 113, 116, 128, 129; Proboscidea, 337, 339, 340, 341, 364; Rodentia, 76, 84-85, 87; Sirenia, 573578; Suidae, 442, 448; Tragulidae, 536; Tubulidentata, 270, 275 Eocetus, 587, 594, 596; schweinfurthi, 587, 590, 591, 593, 598 Eomoropidae, 368 Eosiren, 575 Eotheria, 46 Eotherium, 575; aegyptiacum, 575; libyca, 576; majus, 575, 576; markgrafi, 576; stromeri, 576 Eotheroides, 25, 574, 575-576, 577, 578; abeli, 575; aegyptiacum, 575 Eotragus, 502, 503, 563, 564, 606; artenensis, 564 Eozostrodon: classification, 46; dentition, 48-49; morganucodon, 46, 48; oehleri, 53; postcranial skeleton, 51; watsoni, 53 Epimachairodus, 262 Epimeriones, 83 Epiphiomys, 78 Epixerus, 5 Epoicotheriidae, 270 Epoicotherium, 64 Epomophorus, 4, 5, 7, 66 Epomops, 4 Eptesicus hottentotus, 65 Equidae, 26, 182, 379-418; evolution of, 413-418; geographic distribution, 382387 Equus, 7, 10, 36, 379, 390, 392, 400, 402-413, 415-418, 608; africanus, 412; broomi, 406; cawoodi, 405, 406; distribution, 10, 403-404; food selection, 9;

Index

fowleri, 405, 406, 407; gigas, 405; harrisi, 405, 406, 407; helmei, 405, 406; identification, 402-403; kuhni, 405, 406, 407; louwi, 405; lylei, 406, 408, 409; mauritanicus, 408, 409, 417; platyconus, 408, 409;plicatus, 406, 407; poweri, 405, 406;primigenius, 390; sandwithi, 405, 406; simplex, 408, 409; simplicissimus, 408, 409; sivalensis, 404, 416; stenonis, 404, 405; westphali, 405; zietsmani, 405, 406 Equus (Asinus), 379, 402, 408; asinus, 403, 408, 412; asinus africanus, 412, 414, 416, 418; hydruntinus, 412; tabeti, 408, 409 Equus (Dolichohippus), 379, 402, 403, 404-408; capensis, 403, 405-407, 409, 410, 416\grevyi, 403, 404, 405, 407408, 416; numidicus, 35, 403, 404, 405, 407, 408, 416, 417; oldowayensis, 403, 404-405, 407, 416; simplicidens, 403, 404, 407, 408, 416 Equus (Equus), 402-403, 407, 408, 412, 416; caballus, 403, 406, 407, 408 Equus (Hippotigris), 379, 402, 403, 407, 408-412; burchelli, 403-414, 417, 418; burchelli antiquorum, 408-409; burchelli mauritanicus, 408, 412, 416, 417; quagga, 10, 12, 403, 406, 407, 408, 409-412, 416-417; quagga quagga, 411; quagga wahlbergi, 408; zebra, 10, 11, 403, 404, 405, 407, 408, 409-410, 417 Eragrostis, 20 Eremitalpa, 64;granti, 10 Erinaceidae, 7, 10, 60-61, 64, 65, 67; distribution, 60; origin, 61 Erinaceinae, 60 Erinaceus, 60, 61, 65; europaeus, 61; frontalis, 10; sansaniensis, 60 Erinaceus (Atelerix): albiventris, 61; algirus, 61; broomi, 60, 61; major, 60, 61; sclateri, 61 Erythrocebus, 7, 102 Erythrotherium, 46-51, 53, 604; classification, 46; dentition, 48-49; parringtoni, 21, 46-53; postcranial skeleton, 51 Erythrozootes, 64; chamerpes, 63 Es-Skhul, Neanderthal population, 211 Ethiopia, 30, 609; Bovidae, 548, 567; Cervidae, 497; Equidae (Equus, 404, 407; Hipparion, 388, 389, 390, 394, 395, 397, 398, 399); geological development, 22; Giraffidae, 520, 521, 525; Hominidae, 155; Lorisidae, 91; Rhinocerotidae, 371 Ethiopian Highlands, 10 Eumeryx, 497, 516, 528 Euoticus, 91, 97, 98 Euphausiidae, 597 Eurasia: Bovidae, 565, 567, 568 (Bovinae, 547, 548; Caprinae, 561); Equidae, 390, 391, 396; faunal tie with Africa, 603-609, 611, 614; Pongidae, 126, 144; Proboscidea, 360, 364 (Gomphotherioidea, 345, 351, 354, 356; Mammutoidea, 358); Rhinocerotidae, 372, 374, 376; Suidae, 452; Tayassuidae, 435 Eurhinodelphidae, 596 Eurhinodelphis, 592, 596

Europe: Anthracotheriidae, 423, 430, 431, 433; Bovidae, 563, 564, 565, 566, 567, 568, 569; Carnivora, 251, 252, 253, 263, 264, 265 (Creodonta, 253; Felidae, 262; Viverridae, 257); Cercopithecidae, 102, 112; Cervidae, 496; Cetacea, 582, 584, 591, 596; Chalicotheriidae, 368, 369; cultural associations of Hominidae, 225; Deinotheriidae, 315, 322, 323, 324, 328; Equidae, 389, 391; faunal ties with Africa, 604-608, 611, 614; Giraffidae, 516, 528; Hominidae, 203; Hippopotamidae, 487, 492; Hyracoidea, 284, 289, 307, 308, 311; Mesozoic mammals, 53, 54; Neanderthal population, 211; Palaeomerycidae, 498, 503, 506; Pholidota, 270; Pongidae, 127, 139, 140, 144; Proboscidea, 339, 342, 352, 358, 360, 362, 363, 364; Rhinocerotidae, 371, 372, 374, 376; Rodentia, 74, 75, 76; Sirenia, 574, 576, 577, 578; Suidae, 441-445, 447, 448, 452, 453; Tayassuidae, 476; Tragulidae, 536, 537; Triassic deposits, 53 Euryboas, 260; namaquensis, 260 Eurygnathohippus, 388, 390; cornelianus, 399 Eutriconodonta, 46 Extinctions, 11-12, 609, 612-614; Anthracotheriidae, 423; Barytheriidae, 605; Bovidae, 568; Cervidae, 497; climatic influence, 12, 613; Equidae, 411, 415; Giraffidae, 509; Hominidae, 223; human influence, 12, 613; Moeritheriidae, 605; Proboscidea, 364 Falconidae, 286 Farming, impact on species proportions, 82 Faunal history: and patterns of evolution, 603-614; and prehistoric extinctions, 11-12 Fayum, Egypt, 23, 25, 26; Anthracotheriidae, 427, 430, 431, 433; Barytheriidae, 329, 330; Carnivora, 252, 253, 264, 605; Cetacea, 590; Chiroptera, 65, 66; Embrithropoda, 279, 280; Hyracoidea, 287, 288, 296, 299; Insectivora, 56, 57; Moeritherioidea, 333, 334, 335; Parapithecidae, 112-113, 114, 115; Pongidae, 120, 121-124, 128, 129, 131, 141, 143, 144; Rodentia, 70, 71, 74, 76, 77, 80, 84; Sirenia, 575 Felidae, 5, 6, 7, 10, 11, 26, 257, 258, 260-262, 264, 265, 606; distribution, 2; domestication, 12; numbers of, 263 Felis, 260, 262, 264; aurata, 5; chaus, 260; lybica, 260; margarita, 260 Felis (Leptailurus), 260; brachyura, 260; serval, 5, 260 Felis (Lynx), 260; caracal, 10, 260 Felis (Microfelis), 260; nigripes, 7, 10, 260 Felis (Profelis), 260; aurata, 260 Felsinotherium, 577; serresi, 577 Fennec, 254 Fennecus, 253, 255; zerda, 253 Ferecetotherium, 595 Ferryville, Tunisia: Gomphotherioidea, 345; Mammutoidea, 359 Fish, 26, 27, 30, 35, 37, 227 Fish Hoek, South Africa: Homo sapiens, 213, 215

Index

Florisbad, South Africa, 40; Bovidae, 541 (Alcelaphinae, 555, 556, 557; Antilopinae, 559; Hippotraginae, 551, 553); Giraffidae, 524\Homo sapiens, 213, 215; paleoenvironment, 231; Suidae, 469 Florisbad Faunal Span, 40, 214; Bovidae, 541, 569; Hominidae, 159, 206 Forests, 4-6, 7; equatorial, 231; evergreen, 20, 231; fossil mammal records in, 268; gallery, 20, 25; inhabitants of, 4-6; Lowland Forest Zone, 4-6, 20; montane, 20; relic, 4, 6; tropical, 17, 20; vegetation of, 20 Fort Ternan, Kenya, 28, 30, 32, 613; Anthracotheriidae, 424; Bovidae, 541, 563, 564, 567, 568 (Alcelaphinae, 558; Antilopinae, 561; Bovinae, 547; Caprinae, 563); Carnivora (Credonta, 253; Hyaenidae, 258, 259; Viverridae, 257); Deinotheriidae, 321, 322; Giraffidae, 515, 516, 517, 518; Hippopotamidae, 486; Lorisidae, 92, 93, 95; Palaeomerycidae, 499-500; Pongidae, 127, 138, 143, 144; Proboscidea, 360; Ramapithecus, 147, 148, 149, 150; Rhinocerotidae, 371, 375; Rodentia, 70, 71, 79, 80, 81, 85, 86; Tayassuidae, 477; Tragulidae, 537; Tubulidentata, 273 Fortuna Sandstone, Tunisia: Sirenia, 577-578 Fouarat, Algeria, 35, 36 Foum el Kranga, Tunisia: Proboscidea, 345 Fox: Bat-eared, see Otocyon; Common (red), see Vulpes\ Simenian, see Canis simensis", Silver, see Vulpes chama France: Anthracotheriidae, 430; Bovidae, 546, 564 (Bovinae, 549; Caprinae, 563); Carnivora, 252, 261 (Felidae, 261; Hyaenidae, 260); Deinotheriidae, 315, 328; Hominidae, 203; Proboscidea, 342; Rhinocerotidae, 372; Sirenia, 575, 577; Tragulidae, 537; Tubulidentata, 270, 273, 275 Fucotherium, 577 Funisciurus, 5 Gabon forest region, 6 Gafsa, Tunisia; Anthracotheriidae, 424; Carnivora, 257 Galagidae, 2, 5, 6, 11, 90, 98; locomotion, 96; teeth, 95 Galaginae, 90, 91-95, 97, 98 Galago, 7, 91, 93, 94, 95, 96, 97, 98; allem, 93, 98; crassicaudatus, 93, 94, 96, 98; demidovii, 93, 97, 98; elegantulus (Euoticus), 97, 98; inustus, 98; senegalensis, 96, 98·, zanzibaricus, 98 Galagoides demidovii, 5, 97, 98 Galagos. See Galagidae; Galago Galana Boi Beds, Kenya, 33 Galerix, 60; africanus, 60 Galidiinae, 11, 257 Gamble's Cave, Kenya: Suidae, 466 Game farming, 13-14 Game reserves, 12; and tourism, 14 Gandakas, Pakistan: Ramapithecus ,149, 150, 151 Gao, Mali, 25 Gara Ziad, Morocco, 27

Garet el Gindi, Egypt: Embrithropoda, 282 Garet el Moluk, Egypt, 35; Giraffidae, 525 Garet el Tir, Algeria: Proboscidea, 358 Garet Ichkeul, Tunisia, 36; Bovidae, 554; Equidae, 404; Giraffidae, 525 Gaudeamus, 77 Gazella, 8, 9, 10, 415, 550, 552, 558, 559561, 563, 567, 568; arista, 560; atlantica, 560; bondi, 559; cuvieri, 560; dama, 560, 567; decora, 560; domestication, 12; dorcas, 8, 560, 567\gazella praecursor, 560; gracilior, 560; granti, 560, 567; helmoedi, 561; hennigi, 560; janenschi, 560, 569; kohllarseni, 554, 560-561; leptoceros, 560;praegaudryi, 561;praethomsoni, 560; rufifrons, 560, 567; rufina, 560; setifensis, 560; soemmerringi, 560, 567; superba, 552; thomasi, 560; thomsoni, 560, 567; tingitana, 560; vanhoepeni, 560, 569; wellsi, 559 Gazelles. See Antilopini; Gazella Gazellospira, 567 Gebel Bou Gobrine, Tunisia, 26; Proboscidea, 340, 341 Gebel el Qatrani Formation, Fayum: age, 124; Anthracotheriidae, 424, 427; Carnivora, 252; Embrithropoda, 279, 280; Hyracoidea, 284, 290, 291, 302, 304; Lower Fossil Wood Zone, 25 (see also Lower Fossil Wood Zone); Moeritherioidea, 334; Parapithecidae, 113; Pongidae, 131, 139, 143; Proboscidea, 340, 341; Upper Fossil Wood Zone, 25 (see also Upper Fossil Wood Zone) Gebel Tebaga, 36 Gebel Zelten, Libya, 26; Anthracotheriidae, 424, 428, 429, 430, 431; Bovidae, 563, 568; Carnivora, 253; Cercopithecidae, 112; Deinotheriidae, 321, 322, 323, 324; Giraffidae, 509; Hyracoidea, 310; Palaeomerycidae, 498, 500-501, 502, 503; paleoenvironment, 452-453; Proboscidea, 342, 343; Rhinocerotidae, 373; Sirenia, 577; Suidae, 441, 442, 444, 445, 446, 447, 456; Tragulidae, 537 Geese, domestication of, 12 Geiseltal, 253 Geiadas. See Theropithecus gelada Gelocidae, 564 Gelocus whitworthi, 536 Gemsbok. See Oryx gazella Genets. See Viverrinae Genetta, 256; abyssinica, 10, 256,genetta, 256; maculata, 256; servalina, 256; tigrina, 256; victoriae, 5, 256 Geniohyidae, 288, 289, 290, 291 Geniohyinae, 290, 312 Geniohyus, 290-293, 294, 312; characteristics, 285, 286, 294, 295, 296, 302, 308; classification, 287, 289, 295; diphycus, 288, 290, 291, 292-293; fajumensis, 287, 295;gigas, 288, 296, 297; magnus, 290, 293-294, 299; major, 287, 296; micrognathus, 287, 295, 296; minutus, 287, 295, 296; mirus, 287, 290, 291-292, 293, 295, 296; pygmaeus, 293; subgigas, 288, 291, 295, 296 Geogale, 64; aletris, 63, 64; aurita, 64

627

Geogalinae, 63 Geographical features of Africa, 17 Geological development of Africa, 20-23 Geraru, Ethiopia: Hippopotamidae, 492 Gerbillinae, 75, 81, 82, 83, 84, 86; distribution, 69, 75 Gerbillus, 81, 83 Gerbils, 10 Gerenuk. See Litocranius. Germany: Cetacea, 595, 596; Deinotherium, 324; Hominidae, 203; Rhinocerotidae, 376; Tragulidae, 537 Gerontochoerus, 457, 460 "Gerontochoerus" euilus, 458 Getuloxerus tadlae, 82 Geziret el Qorn Island, Fayum: Hyracoidea, 287 Ghana, meat yield from wild animals, 13 Gibbons. See Pongidae Gigantohyrax, 308, 309, 310, 312; maguirei, 289, 310 Gigantopithecus, 127, 128, 129, 132, 150, 151; bilaspurensis, 144; blacki, 144; and Dryopithecus compared, 132 Giraffa, 518-523, 524, 525, 608; attica, 529; camelopardalis, 8, 509, 518-520, 521-522, 525, 529; capensis, 519; characteristics, 500, 505, 514, 528; gracilis, 509, 520-522, 529; habitat, 530-jumae, 509, 518-519, 521-522, 529; andPalaeotragus compared, 515-516; priscilla, 529; punjabiensis, 529; pygmaea, 509, 520-522^ 529; pygmaeus, 521; sivalensis, 529; stillei, 509, 520-521, 522, 529; taxonomic status, 523 Giraffes. See Giraffa·, Giraffidae; Giraffinae Giraffidae, 7, 9, 26, 27, 37, 509-531, 606, 607, 608, 609; classification, 497, 536; cranial appendages, 498; evolutionary trends and relationships, 487, 528531; feeding habits, 370; geographic distribution, 510-511; teeth, 501 Giraffinae, 518-523 Giraffoidea, 498, 503, 504, 505, 564 Giraffokeryx, 501, 505, 509, 512, 516, 529; punjabiensis, 505, 516, 531 Gizeh, Egypt: Proboscidea, 345 Glacial times, paleoenvironments during, 231-232 Gladysvale, South Africa: Insectivora, 61 Gliridae, 82, 83, 85 Glirinae, 76 Gliroidea, 69, 76 Glossopteris, 20 Goa, Mali: Moeritherioidea, 334 Goats. See Bovidae; Capra; Caprini Gobiocerus mongolicus, 564 Gomphotheriidae, 30, 37, 328, 336, 339349, 360, 364, 604, 606, 607, 608, 612; characteristics, 350, 351; classification, 359; and Elephantidae compared, 349; expansion into Eurasia, 338; origin, 341 Gomphotheriinae, 339-340, 345 Gomphotherioidea, 339-358 Gomphotherium, 26, 27, 341-344, 364; angustidens, 328, 342-343, 360; angustidens kisumuensis, 342, 343; angustidens libycus, 342; angustidens pontileviensis, 342; and Barytherium compared, 329; classification, 339; evo-

628

Gomphotherium, (Continued) lution, 345; kisumuensis, 342; and Palaeomastodon compared, 340; productum, 342; and Stegotetrabelodon compared, 350 "Gomphotherium": ngorora, 344, 360; pygmaeus, 342, 343-344 Gondwana surface, 23 Gondwanaland, geological development, 20, 22, 23 Gorgopithecus, 105; major, 105 Gorilla, 2, 108, 151; dentition, 169; gorilla, 5; relationships, 127; size, 128, 132, 135 Graphiurinae, 76; distribution, 70 Graphiurus, 5 Great Karroo, South Africa: Equidae, 403 Great Namaqualand, South Africa: Equidae, 403 Great Rift Valley, 23 Greece, 148; Giraffidae, 529; Hyracoidea, 289; Ramapithecus, 148, 149 Griquatherium, 523; cingulatum, 525, 526; haughtoni, 525, 526 Grombalia, Tunisia: Proboscidea, 345 Grysbok. See Raphicerus Guenons, 101 Guinea forest region, 6 Gulo, 255 Gundis, 2 Guomde Formation, Kenya, 33 Guyot, Morocco: Suidae, 464 Gymnurechinus, 60, 61, 142; camptolophus, 60; leakeyi, 60; songhorensis, 60 Hadar Formation, Ethiopia, 33; Bovidae (Alcelaphinae, 558; Caprinae, 562; Hippotraginae, 552); Cercopithecidae, 102, 106, 110; Deinotheriidae, 323, 324; Equidae, 397, 398, 400; Giraffidae, 551; Hippopotamidae, 492; Hominidae, 158, 163-165, 173, 194, 224; paleoenvironment, 228; Suidae, 456, 459, 465 Hadropithecus, 151 Hagfet el Tera, Libya: Equidae, 412 Halianassa, 577 Halianassinae, 574 Halicore, 578 Halipaedisca, 578 Halitheriinae, 574, 575-578 Halitherium, 575, 576-577, 578; bellunense, 576; christoli, 576 Halysiren, 577 Hamada Damous, Tunisia, 36; Proboscidea, 345; Suidae, 457 Hand axes, 469; from Skoonspruit, 40 Hapalodectes, 592 Hapalodectinae, 592 Haramiyidae, relationships, 53 Hare. See Leporidae. Harounian transgressive stage, 212 Harpagolestes, 593; orientalis, 593 Hartebeest. See Alcelaphini; Alcelaphus', Beatragus Harts River, South Africa, 37 Haua Fteah Cave, Libya, 159; Equidae, 412; Homo sapiens, 210 Hecubides, 253 Hedgehog. See Erinaceidae; Erinaceus frontalis

Index

Heliophobius, 79 Heliosciurus, 5 Helladotherium, 509, 523, 529, 530; duvernoyi, 523; olduvaiense, 524 Helogale, 256-257; brunnula, 256; dybouiskii, 256; hirtula, 256; ivori, 256; macmillani, 256; mimetra, 256;parvula, 256; percivali, 256; undulata, 256-257; varia, 257; vetula, 257; victorina, 257 Helvetian Stage, Suidae from, 443, 444 Hemibos, 548, 566 Hemiechinus, 61 Hemigalago, 97, 98 Hemigalinae, 11, 257 Hemionus, 404 Hemitragus, 547, 563 Hempstead Beds, England: Anthracotheriidae, 430 Henchir Beglia, Tunisia, 26; Proboscidea, 345 Herbivores: forest, 6; savanna, 7, 8 - 9 Herpestes, 5, 256-257; cauui, 256; dentifer, 256; granti, 256; ichneumon, 256; ignitus, 256; nigratus, 256; ochracea, 256; pulverulentus, 256; ratlamuchi, 256; ruddi, 256; sanguineus, 256 Herpestides, 257 Herpestinae, 5, 7, 10, 11, 257, 265; domestication, 12 Hesperochippus, 404 Heterocemas, 498 Heterocephalus, 10, 79, 81 Heterohyrax, 308, 309, 312 Hexaprotodon, 36, 483, 489, 493; general features, 484, 485; harvardi, 487, 488, 489; iravaticus, 489; karumensis, 492; liberiensis, 483, 485, 488, 493-494; protamphibius, 491; sivalensis, 489 Hipparion, 27, 30, 36, 109, 227, 311, 379, 388-402, 413, 415, 418, 607; afarense, 389, 390, 392, 397-399, 400, 401, 402, 416; africanum, 389-402, 413, 415; albertense, 37, 389, 390, 392-393, 399; albertense baardi, 390, 393, 394, 399; albertense serengetense, 392, 399; albertensis, 392, 399; ambiguum, 399; baardi, 389, 390, 392, 393, 402; erassum, 399; and Equus compared, 402, 404; ethiopicum, 389, 399, 416; fauna, 562, 563, 565, 578;gracile, 390, 394; hipkini, 390; hippidiodum, 389; libycum, 35, 36, 389, 390, 392, 394, 397, 398, 399-402, 416; libycum ethiopicum, 399, 400, 401, 415; libycum steytleri, 399, 415; massoesylium, 399; matthewi, 391; mediterraneus, 391, 394; namaquense, 389, 390, 392, 393-394, 402; primigenium, 26, 389-402, 414, 415416;primigenius, 300, 394; sitifense, 36, 389, 390, 392-397, 401, 415-416; sitifense gromovae, 394-395; steytleri, 389-390, 399; turkanense, 389, 390, 393, 395-401 Hippohyus, 442, 447-451, 453 Hippopotamidae, 9, 11, 26, 30, 36, 37, 483-495; classification, 328; evolutionary trends, 423, 433, 492-495, 607, 609; general features, 483-485; Koobi Fora Formation, 224; Lukeino, 227; and Merycopotamus compared, 424, 433

Hippopotamus, 35, 433, 483, 493; aethiopicus, 491, 492; amphibius, 376, 424, 483, 485, 486-488, 491-495; amphibius antiquus, 492; crusafonti, 487; dubius, 576; general features, 483-485; gorgops, 491, 492, 493; hipponensis, 489; imagunculus, 489, 491, 492; kaisensis, 487, 491, 492; lemerlei, 492; medius, 577;pantanelli, 487; primaeuus, 487;protamphibius, 491, 492; protamphibius andrewsi, 487; siculus, 487 Hippopotamus, pygmy. See Hexaprotodon liberiensis Hipposideridae, 65, 66, 67 Hipposideros, 65 Hippotigris. See Equus (Hippotigris) Hippotraginae, 540, 550-554, 564, 565, 568, 569 Hippotragini, 8, 540, 552-554, 567 Hippotragoides broomi, 552 Hippotragus, 8, 552-554; bohlini, 554; cordieri, 547, 554, 558; equinus, 8, 552, 553, 554;gigas, 553-554, 568; leucophaeus, 11, 12, 552, 553, 559, 568; niger, 8, 552, 553-554; problematicus, 552 Hiwegi Formation, Rusinga, 28; Tragulidae, 537 Hluhluwe (game reserve), 12 Hog. See Suidae; Forest, see Hylochoerus; River, see Potamochoerus Holocene Era. Bovidae (Antilopinae, 560; Hippotraginae, 550, 553); Cervidae, 497; Cetacea, 584; Equidae, 379, 413, 417 (Equus, 409); Giraffidae, 522; faunal evolution, 609, 612, 614; Suidae, 436; Tubulidentata, 270 Holsteinian interglacial stage, Homo sapiens from, 203 Homa Bay, Kenya: Deinotheriidae, 323 Homa Mountain, Kenya: Proboscidea, 356; Suidae, 472, 473 Homalodotherium, 370 Hominidae, 30-31, 33, 34, 37-38, 40, 100, 120, 147, 151, 152, 154-232, 607, 608; coexistence of hominid taxa, 222223; cultural associations, 223-226; dentition, 149; descriptions, 159-217; extinctions, 223; paleoenvironmental settings, 227-232; stratigraphic and geographic setting, 154-159 Hominoidea, 30, 120, 156, 158; characteristics, 139; classification, 100, 139, 169, 449; and cercopithecid origins, 116; Ramapithecus, 147-152 Homo, 35, 158, 163, 172, 185-217, 219; adaptive grade and structural-functional zone, 222; africanus, 169; and Australopithecus compared, 165, 167, 169, 179, 218, 219, 222; coexistence with Australopithecus, 222-223, 230; cultural associations, 224, 225; definition, 193; and Dendropithecus compared, 141; distribution, 189-194; features, 186-189; habilis, 173, 186, 190-193, 223, 225 (paleoenvironmental setting, 229, 230); kanamensis, 189; paleoenvironmental setting, 229-232; and Prohylobates compared, 112; and Ramapithecus compared, 147, 150, 151, 152; sp. nov. indet., 186-194

Index

Homo erectus, 182, 194-201; and Acheulian industry, 225-226; and Australopithecus compared, 200, 201; and coexistence of Hominid taxa, 223; description, 194-196; distribution, 196-200; and early Homo compared, 185, 186, 187, 188, 189; and Homo sapiens compared, 194-196, 200-204, 208; and Homo sp. nov. indet. compared, 186, 187, 189, 192, 193; paleoenvironmental settings, 228-231; soloensis, 194, 216-217 Homo (Pithecanthropus): dubius, 192; modjokertensis, 192, 193 Homo sapiens, 185, 201-205, 222, 227, 494; afer, 213-217, 223 (paleoenvironmental settings, 231); capensis, 213, 223; characteristics, 201-205; distribution, 203-205; extinctions, 223; and Homo erectus compared, 194-196, 200-204, 208; neanderthalensis, 205, 210-213; origins, 203, 205, 218; paleoenvironmental settings, 231-232; rhodesiensis, 205, 206-210, 215, 217, 223, 232; sapiens, 201 Homoioceras, 549; singae, 548 Homotherium, 260, 262; ethiopicum, 262 Honanotherium, 529 Hopefield. See Elandsfontein Horses. See Equidae Hsanda Gol Formation, Mongolia: Bovidae, 564 Humbu Formation, Tanzania, 34; Hominidae, 178 Hungary: Deinotheriidae, 315, 320, 327; Homo sapiens, 203; Ramapithecus, 150; Sirenia, 573, 574 Hunting: impact on fauna, 12, 614; and tourism, 14; and utilization of wild mammals, 13 Hyaena, 2, 7, 27, 226, 258, 260, 451; algeriensis, 258; bellax, 260; brunnea, 10, 258; hyaena, 259; namaquensis, 260 Hyaenictis, 260; preforfex, 260 Hyaenidae, 11, 257-260, 264, 265, 607; numbers, 263 Hyaenodon, 253 Hyaenodontidae, 25, 249, 250, 251, 253, 264, 604, 609; characteristics, 250; numbers, 263 Hyaenodontinae, 251, 252-253 Hyainailouros, 253; fourtaui, 253; nyanzae, 253; sulzeri, 253 Hybomys, 5 Hydaspicobus auritus, 552 Hydaspitherium, 529 Hydrarchidae, 587 Hydrodamalinae, 574 Hydrodamalis, 574; gigas, 573 Hydropotes, 504 Hydrurga, 263 Hyemoschus, 6, 536, 537; aquaticus, 537 Hyena. Brown, see Hyaena brunnea. Spotted, see Crocuta crocuta. Striped, see Hyaena hyaena Hylobates, 112, 130, 138; and Aeolopithecus compared, 139; concolor, 130; lar, 130 Hylobatidae, 120, 139-141, 605, 606, 607; classification, 100, 125, 127, 138; and Dendropithecus affinities, 134, 144 Hylobatinae, 125

Hylochoerus, 6, 436, 466-467, 477, 479; antiquus, 466; characteristics, 437, 438, 440; euilus, 459, 460; evolution, 452; grabhami, 465; and Kolpochoerus compared, 462, 463, 464; meinertzhageni, 436, 437, 440, 455, 459, 466-467; and Metridiochoerus compared, 471, 472, 473, 474; and Notochoerus compared, 458, 459; and Nyanzachoerus compared, 455, 456; and Phacochoerus compared, 468, 469 Hyoboops, 424, 428-429, 431, 433; africanus, 428, 429, 431; moneyi, 424, 428, 429, 431; palaeindicus, 428, 429, 431 Hyosus, 442, 447, 448, 4 4 9 - 4 5 1 Hyotheriinae, 442-444, 449, 452, 453 Hyotherium, 442-444, 452; aurelianensis, 444; dartevellei, 441, 442, 443-444; jeffreysi, 447; kijwium, 442, 444; a n d K u banochoerus compared, 446; meissneri, 443; palaeochoerus, 443; palaeocology 452, 453; pascoei, 444; phylogeny and classification, 448, 449, 450, 451; soemmeringi, 443, 453; typum, 443; water housi, 443; and Xenochoerus compared, 447 Hyparrhenia, 220 Hyppohyus, 450; grandis, 450; sivalensis, 450 Hypsignathus, 4 Hyracidae, 308 Hyracoidea, 5, 6, 7, 10, 26, 27, 30, 284312, 605, 607, 608; characteristics, 2, 284-286; evolutionary trends, 270, 311; Gebel Qatrani Formation, 25; history of research, 286-289; relationships, 311-312, 337, 339, 573; Swartkrans cave system, 182; systematic descriptions, 289-311 Hyrax, 308, 310; kruppii, 286 Hyraxes. See Hyracoidea Hystricidae, 5, 11, 70, 79, 81, 182; distribution, 69 Hystricognathi, 71, 84, 85 Hystricognathy, definition of, 70 Hystricoidea, 74, 86 Hystricomorph, definition of, 70 Hystrix, 81, 82, 84 Ibero-Maurusian industry, of Maghreb, 211 Ibexes. See Capra. Ethiopian, see Capra walie Ichneumia, 257; albicauda, 257 Ictidosaurs, 51, 53 Ictitherium, 260 Ictonyx, 7, 256; kalaharicus, 256; orangiae, 256; striatus, 255, 256 Ileret Ridge, East Turkana, 33; Giraffidae, 524; Hominidae, 179; Proboscidea, 354, 356. See also East Turkana; Koobi Fora; Lake Turkana Immigration, 609, 611, 614; Anthracotheriidae, 430, 431, 433; Cervidae, 496497; during Eocene Era, 604; Equidae, 413, 417, 418; during Miocene Era, 8 0 82 Impala. See Aepyceros In-Tafidet, Mali, 25 Incamys, 78 India: Anthracotheriidae, 429; Bovidae, 565, 566, 567, 568; Cetacea, 582, 585,

629

592, 593, 598; Equidae, 389, 404; geological development, 22; Giraffidae, 524, 530; Ramapithecus, 147, 148, 149; Suidae, 442, 445, 447, 451, 455 Indian Ocean, geological development, 22 Indocetus, 593; ramani, 592-593, 595, 598 Indopacific region: Sirenia, 573, 578 Indoredunca theobaldi, 552 Indraloris, 91 Indratherium, 524, 530 Indridae, 11 Iniidae, 597 Insectivora, 2, 4, 6, 10, 11, 27, 56-65, 592, 604; Chrysochloridae, 64-65; Erinaceidae, 60-61, 65; Macroscelididae, 57-60, 65; Ptolemaiidae, 56-57, 65; Soricidae, 61-63, 64, 65; Tenrecidae, 63-64, 65 Intertropical Convergence Zone, 231 Iran: Carnivora, 262; Chalicotheriidae, 369 Irhoud-Ocre, Morocco: Insectivora, 59, 63; Rodentia, 82, 83 Irhoudia, 83 Irrawaddy Beds of Burma: Hippopotamidae, 489 Ischyromyidae, 76, 77, 84, 85 Isimila, Tanzania: Bovidae, 557 Isoberlinia, 20 Israel, 199; Deinotheriidae, 316; Homo sapiens, 210; Neanderthal population, 211 Italy: Carnivora, 256; Cetacea, 591; Hippopotamidae, 487, 489; Sirenia, 575, 576 Ituri Forest, Zaire: Giraffidae, 509, 518, 528 Jackals. See Canis. Asiatic, see Canis aureus. Black-backed, see Canis mesomelas Jaculus, 81 Jamaica: Sirenia, 574 Jaramillo Normal Event, 200 Java: Hippopotamidae, 489; Hominidae, 187, 192, 193, 196, 199 Jebel Irhoud, Morocco: Hominidae, 159, 210-211, 212; Rodentia, 70 Jordan: Bovidae, 565, 566 Julbernardia, 20 Jurassic Period: Brancatherulum tendaguruense, 46, 54; distribution of deposits, 54; faunal evolution during, 604; geological developments during, 22, 27 Kabarnet Trachyte Formation, Kenya, 30, 156 Kabua, Kenya: Suidae, 469 Kabuh Formation, Java, 192; Hominidae, 196 Kabwe, Zambia: 201, Homo sapiens, 206, 209, 210, 216 Kabylie, Algeria: Proboscidea, 343 Kada Hadar Member, Ethiopia: Hominidae, 164, 173; see also Hadar Formation Kagua, Kenya: Suidae, 472, 473 Kaiso Formation, Uganda, 34; Bovidae, 569 (Bovinae, 545, 548; Hippotraginae, 551); Chalicotheriidae, 368, 369;

630

Kaiso Formation, (Continued) Equidae (Equus, 404; Hipparion, 392, 393); Giraffidae, 521 (Palaeotraginae, 518); Hippopotamidae, 489, 491; Proboscidea (Gomphotherioidea, 351, 352, 356; Mammutoidea, 359); Suidae, 454, 456, 459, 460, 463, 469 Kalam, Omo Basin: Proboscidea, 356 Kalem Beds, Omo Basin: Hippopotamidae, 491-492 Kanam, Kenya, 35; Bovidae (Antilopinae, 560; Bovinae, 545; Hippotraginae, 550); Cercopithecidae, 102, 104, 108; Deinotheriidae, 323; Hippopotamidae, 489; Hominidae, 189; Proboscidea, 351; Rhinocerotidae, 374, 375; Suidae, 454, 474 Kanapoi, Kenya, 30, 32, 33, 34, 35; Carnivora (Hyaenidae, 260; mustilds, 255); Cercopithecidae, 103, 104; Chiroptera, 66; Deinotheriidae, 323, 324; Equidae, 391, 396; Giraffidae, 519, 520, 529; Hippopotamidae, 489, 491, 492; Hominidae, 158, 172, 218, 224; Proboscidea, 352; Rhinocerotidae, 371, 375; Rodentia, 81; Suidae, 36, 37, 454, 456, 457, 458, 459, 460 Kangaroo, 453 Kanisamys, 81 Kanjera, Kenya, 35; Bovidae (Alcelaphinae, 557; Antilopinae, 559, 560; Bovinae, 545, 549; Hippotraginae, 550, 553); Cercopithecidae, 106, 107; Suidae, 466-467, 473, 476 "Kansupithecus," 112, 120 Kaolack, Senegal: Cetacea, 584, 589; Moeritherioidea, 334 Kaparaina Basalt, Kenya, 30, 156 Kaperyon Beds, Kenya: Proboscidea, 350 Kapthurin Beds, Kenya, 30; Homo sapiens, 204 Karari-Chari Tuffs, East Turkana, 225 Karbarsero Beds, Ngorora Formation, Kenya: Giraffidae, 516 Kariandusi, Kenya, 35 Kariba-Luangwa-Lower Zambezi trough, 22 Karmosit Beds, Kenya, 30; Bovidae, 558 Karroo, 22; Equidae, 410; vegetation, 231 Karroo Cycle, geological development during, 22 Karroo System, 20, 22 Karugamania, Congo, 28 Karungu, Kenya, 27, 28; Deinotheriidae, 321; Hyracoidea, 302; Insectivora, 57; Pongidae, 134, 135, 140, 141, 142; Proboscidea, 342; Rhinocerotidae, 371, 372, 373; Rodentia, 70; Tragulidae, 537 Katanga, Congo: Suidae, 469 Kavirondo Gulf, Lake Victoria, 28, 35; Carnivora, 257; Deinotheriidae, 322; Lorisidae, 91 Kavirondo Rift Valley, 92, 142 Kayenta Formation, North America, 53 Kazinga Channel, Uganda, 34: Hippopotamidae, 489; Suidae, 456, 457 KBS Tuff, East Turkana, 33, 179, 193, 228, 230; age, 200; Equidae, 400 Kebibat, Morocco: Homo erectus, 196197; Homo sapiens, 203, 204, 205, 210, 212-213 Kedung Brubus, Java: Hominidae, 196

Index

Kekenodon, 595 Kelba, 250-251; dentition, 56; quadreemae, 56 Kendek et Ouaich: Rodentia, 83 Kenya, 12, 14, 28, 30, 112; Anthracotheriidae, 424, 428; Carnivora, 252, 253 (Felidae, 260; Viverridae, 257); Cetacea, 583, 596; Equidae (Equus, 404, 407, 408; Hipparion, 389, 390, 394, 395, 399); Giraffidae, 509 (Giraffinae, 518-522; Palaeotraginae, 514-518; Sivatheriinae, 525); Hominidae, 155, 156; Hyracoidea, 289, 308, 309, 312; Insectivora, 65; Lorisidae, 91; Pongidae, 120, 124, 127, 134-135, 138, 141, 142; Proboscidea, 351; Rhinocerotidae, 371; Rodentia, 77, 78, 79; Suidae, 466; Tubulidentata, 270, 271, 273 Kenya Rift, 30; geological development, 27; Hominidae, 156 Kenyalagomys, 84, 142 Kenyameryx, 498 Kenyamyidae, 71, 77, 78 Kenyamys, 77, 78, 80, 85 Kenyapithecus, 126-127; africanus, 126, 127, 134; wickeri, 127, 147, 148 Kerio Valley, Kenya, 30 Khenchela, Algeria: Proboscidea, 359 Khenzira, Morocco: Cervidae, 496 Kiahera Formation, Rusinga, 28; Tragulidae, 537 Kiangata Agglomerate, Rusinga, 28 Kibish Formation, Omo, 33; Bovidae, 548; Homo sapiens, 214, 216-217; Proboscidea, 356 Kich.ech.ia, 257 Kikagati, Uganda, 34; Proboscidea, 355 Kilimanjaro Mountain, 6 Kipcherere Basin, Kenya, 30 Kirald: Deinotheriidae, 320 Kirimun, Kenya: Rhinocerotidae, 373, 374; Rodentia, 70, 79 Kisingere, Kenya, 27-28 Klasies River Mouth Cave, South Africa: Homo sapiens, 214 Klein Zee, South Africa, 37; Carnivora, 260; Insectivora, 57, 59 Klipspringer. See Oreotragus Knysna forest, 20 Koalas, 90 Kobus, 8, 550-551, 552, 569; ancystrocera, 551, 552; ellipsiprymnus, 8, 550, 551, 568, 569; kob, 550, 551; leche, 8, 550, 551; megaceros, 8, 550; sigmoidalis, 551, 552, 568, 569; venterae, 551 Koffiefontein, South Africa: Equidae, 411 Koiropotamus majus, 465 Kokkoth, Kenya, 35 Kolinga, Chad: Proboscidea, 350, 351, 352, 359 Kolpochoerus, 462-466, 467, 473, 477, 479; afarensis, 462, 465-466; limnetes, 463-464, 465, 467; olduvaiensis, 463, 464-465; paiceae, 462-463; phacochoeroides, 464; sinuosus, 462, 463 Kolpohippus plicatus, 405 Kom Ombo, Egypt: Bovidae, 549; Equidae, 412 Komba, 91, 93, 94, 95, 98; minor, 93, 94, 95, 96; robustus, 93, 95, 96 Koobi Fora Formation, East Turkana, 33; artifacts in, 224, 225-226, 230; Bo-

vidae (Alcelaphinae, 555, 556, 558; Bovinae, 545, 549; Hippotraginae, 551); Cercopithecidae, 103-104; coexistence of hominid taxa in, 223, 230; Equidae, 399, 400, 402; Giraffidae, 525; Hippopotamidae, 492; Hominidae, 165, 187, 193, 195, 219 (Australopithecus, 172-173, 176, 179, 228-229, 230; Homo, 192, 228-229, 230; Homo erectus, 186, 196, 199-200); paleoenvironmental setting, 228-229, 230; Rhinocerotidae, 375; Suidae, 469, 472, 473, 474, 476 (Kolpochoerus, 463, 464; Metridiochoerus, 472, 473, 474; Notochoerus, 460). See also East Turkana Koro Toro, Chad, 36 Koru, Kenya, 27, 28; Chalicotheriidae, 368; Chiroptera, 65; Deinotheriidae, 321; Insectivora, 57, 59, 64; Lorisidae, 91, 92, 93; Palaeomerycidae, 498; Pongidae, 127, 132, 133, 134, 135, 138, 140, 141, 142; Rodentia, 70; Tubulidentata, 273,274 Kosti, Sudan: Suidae, 465 Koulä, Chad, 36; Proboscidea, 352 Kranskraal, South Africa: Bovidae, 555 Kraterohippus elongatus, 408, 409 Kritimys, 84 Kromdraai, South Africa, 37, 38; artifacts in, 225; Bovidae, 541, 558; Carnivora, 262; Cercopithecidae, 103, 104, 105, 109; Hominidae, 167, 172, 183, 184-185, 220; Hyracoidea, 311; Insectivora, 61; paleoenvironmental setting, 229; Rodentia, 70; Suidae, 470 Kruger National Park, 12 Kubanochoerus, 444, 445, 449, 455, 456, 477;gigas, 446\jeanneli, 441, 442, 445-446, 453; khinzikebirus, 442, 446; massai, 441, 442, 446, 453, 456; robustus, 445, 446 Kubi Algi Formation, Kenya, 33; Cercopithecidae, 103; Hippopotamidae, 492; Proboscidea, 352, 355; Suidae, 456, 458, 460 Kudu. Greater, see Tragelaphus strepsiceros. Lesser, see Tragelaphus imberbis Kuehneotheriidae, relationships of, 53 Kulu Beds, Rusinga, 28 Kushalgar, Punjab: Suidae, 447 Kvabebihyrax, 289, 307, 308, 312 "Kyoga surface," 27 Laetolil, Tanzania, 34; Bovidae, 541, 565, 569 (Alcelaphinae, 558; Antilopinae, 559, 560; Bovinae, 545, 549; Hippotraginae, 554); Carnivora, 262; Cercopithecidae, 103, 110; Chalicotheriidae, 368, 369; Equidae, 392, 393, 412; Giraffidae, 519, 521, 530; Hominidae, 158, 163, 165, 224; Proboscidea, 353; Suidae, 460, 466, 476; Tubulidentata, 273 Lagomerycidae, 498, 503 Lagomeryx, 497, 498 Lagomorpha, 2, 7, 84; relation to Macroscelidea, 57 Lagostomus, 71, 79 Lake Albert, 28, 34 Lake Baringo, 30-32 Lake Chad: Hippopotamidae, 489; Proboscidea, 350; Suidae, 458 Lake Eyasi: Hominidae, 206, 207, 210

Index

Lake Ichkeul: Rodentia, 70, 84 Lake Karär, 36; Cervidae, 497 Lake Manyara Park, 9 Lake Natron, 34, 109 Lake Ndutu: Homo, 205, 226 Lake Nyasa: Insectivora, 59 Lake Rudolf. See Lake Turkana Lake Turkana, 30, 32; Equidae, 395; Giraffidae, 519; Hippopotamidae, 489; Hyracoidea, 288. See also East Turkana; Turkana Langebaanweg, South Africa, 37, 254; Agriotherium, 255; Bovidae, 541, 565, 566, 567, 569 (Alcelaphinae, 558; Antilopinae, 559, 560; Bovinae, 546; Caprinae, 562; Hippotraginae, 552, 554); Carnivora (Felidae, 262; Hyaenidae, 257, 259, 260; Phocidae, 263); Cetacea, 584; Equidae, 389, 392, 393, 394; Giraffidae, 520, 526, 529; Pholidota, 268; Proboscidea, 357, 359; Rhinocerotidae, 371, 375; Suidae, 456, 457; Tayassuidae, 477 Langental, Southwest Africa: Palaeomerycidae, 502; Rhinocerotidae, 371, 372 Lanthanotherium, 60 Last Glacial episode, paleoenvironment in Africa during, 231 Last Interglacial, 232 Leakeymys ternani, 81 Leakitherium, 253 Leba, Angola: Cercopithecidae, 103, 104 Lechwe. See Kobus. Nile, see Kobus megaceros Leecyaena, 260; forfex, 260 Lemniscomys, 84 Lemuridae, 11, 90, 97 Lemuroidea, 275 Leopard. See Panthera pardus Leporidae, 10, 84, 608 Leptadapis, 116 Leptailurus, 5, 260 Leptobos, 549, 566, 567, 568;falconeri, 519; syrticus, 549 Leptodon, 286, 307\graecus, 286 Leptomanis, 270 Leptorycteropus, 271, 273, 274, 276; guilielmi, 270 Lepus, 84 Lesotho, 21, 22, 47, 604 Liberia forest region, 6 Libya, 23, 25, 26; Anthracotheriidae, 424, 428, 429, 433; Barytheriidae, 329, 330; Bovidae, 547; Carnivora, 249; Cetacea, 584; Equidae, 408, 412, 418; Giraffidae, 513, 522; Hyracoidea, 289, 309, 312; Sirenia, 575; Suidae, 441-442 Libycochoerus, 445; massai, 445 Libycosaurus petrocchii, 430 Libyhipparion, 388; ethiopicum, 389, 399; steytleri, 399 Libypithecus, 102, 108-109; markgrafi,, 108 Libytherium, 517, 524, 525; maurusium, 524; olduvaiense, 524 Limeworks Quarry, Makapansgat: Hyracoidea, 289; Suidae, 470 Limnocyoninae, 251 Limnopithecus, 124, 125, 126, 127, 137139, 144; and Dendropithecus compared, 140, 141; evansi, 125, 138; lege-

tet, 124, 125, 127, 138-142, 144; macinnesi, 125, 127, 137, 138, 140; and Propliopithecus compared, 129 Lion, 7, 11, 12 Lipotyphla, 2, 64 Listriodon, 444-445, 446, 448, 451, 452, 453; akatidogus, 442, 445; akatikubas, 442, 445;guptai, 445; intermedius, 445; jeanneli, 445; latidens, 445; lockharti, 445; mongoliensis, 444; pentapotamiae, 444, 445; splendens, 444, 445; theobaldi, 444 Listriodontinae, 442, 444-446, 452, 453 Litocranius, 8, 10, 559, 560, 567; walleri, 8 Littorina Cave, Morocco: artifacts, 226; Hominidae, 197, 203, 205 Lizards, 126 Localities of fossil mammals, 23-40 Locusts, 277 Longirostromeryx, 497 Loperot, Kenya, 28; Anthracotheriidae, 428; Cercopithecidae, 111; Cetacea, 592; Deinotheriidae, 321; Hippopotamidae, 486; Hyracoidea, 310; Proboscidea, 344; Rhinocerotidae, 371, 374; Rodentia, 79 Lophiomeryx, 564 Lophiomyinae, 75 Lophiomys, 10 Lophochoerus, 446, 452; himalayensis, 446 Lopholistriodon, 444, 446, 452; kidogosajia, 442, 446; moruoroti, 442, 446 Lophuromys, 5 Loris, 98 Lorisidae, 1, 5, 6, 66, 90-98 Lorisinae, 90, 91, 93-98 Losodok, Kenya. See Moruorot Hill Lothagam Hill, Kenya, 30, 32, 36, 37; Carnivora, 257; Cercopithecidae, 103, 106; Deinotheriidae, 323; Equidae, 391, 395, 396, 398, 400; Giraffidae, 516, 529; Hippopotamidae, 487-489, 491; Hominidae, 158, 159, 163, 172, 224; paleoenvironmental setting, 227; Proboscidea, 350, 351, 352, 360; Rhinocerotidae, 371, 372, 375; Suidae, 455, 456, 457; Tubulidentata, 271 Lothidok Hill, Kenya: Proboscidea, 342, 359 Lower Fossil Wood Zone, Fayum: Hyracoidea, 290, 292-300, 303, 304, 306, 311; Primates, 124, 131; Proboscidea, 341 Loxodonta, 352-354, 357, 358, 360, 362, 363, 364; adaurora, 352-353, 354, 362; africana, 6, 36, 199, 336, 352, 354, 362, 363, 473; africanava, 358; antiqua recki, 355; atlantica, 35, 36, 352, 353, 354, 357, 358, 362; atlantica angammensis, 354; atlantica atlantica, 353354; atlantica zulu, 353, 354; griqua, 353; prima, 354 Lukeino, Baringo Basin, Kenya, 30, 227; Bovidae, 552, 567; Chalicotheriidae, 369; Hominidae, 156-158, 159, 161, 223; paleoenvironmental setting, 227; Pongidae, 127; Rodentia, 81; Suidae, 455, 457 Lukenya Hill, Kenya: Homo sapiens, 213, 215, 217

631

Lunatoceras mirum, 555 Lunene Lava, Rusinga, 28 Lutetian Stage: Cetacea, 585, 592, 595; Rodentia, 84 Lutra, 5, 102, 255, 256; lutra, 256; maculicollis, 256 Lutrinae, 151, 255, 256, 265 Lycaon, 2, 7, 253, 254, 255; pictus, 253 Lycyaena, 260; silberbergi, 260 Lynx, 10, 260. Caracal, see Felis (Lynx) caracal Maboko Island, Lake Victoria, 28; Bovidae, 564; Cercopithecidae, 110; Deinotheriidae, 321; Hippopotamidae, 486; Hyracoidea, 302; Palaeomerycidae, 499; Pongidae, 112, 120, 127, 136, 137, 138, 144; Proboscidea, 342, 343; Ramapithecus, 147, 148; Suidae, 446; Tragulidae, 537 Macaca, 102-103, 108, 109; dentition, 102, 111; flandrini, 102, 108; fuscata, 102; libyca, 102 Macaques. See Macaca. Japanese, see Macaca fuscata Machaeroidinae, 251 Machaeromeryx, 497 Machairodontinae, 182, 224, 260, 261262, 264, 265 Machairodus, 260, 262; aphanistus, 262; transvaalensis, 262 Macroscelidea, 57 Macroscelides, 58, 59, 60; proboscideus, 10, 59; proboscideus, vagans, 57 Macroselididae, 1, 2, 4, 7, 10, 11, 5760, 61-63, 65, 288; dentition, 58 Macroscelidinae, 58, 59, 60, 65; distribution, 59 Macrotarsomys, 80 Madagascar, 11; Carnivora, 256, 257; geological development, 22; Myzopodidae, 608, 612; Primates, 90, 96, 97; Rodentia, 75, 85; Sirenia, 575, 576; Tubulidentata, 270, 275, 277; Viverridae, 256 Madoqua, 8, 10, 558, 559; avifluminis, 559 Maghreb, 23, 27, 35, 226; artifacts, 226; Equidae, 379, 414, 417, 418 (Equus, 409, 412; Hipparion, 388); Hominidae, 198-199 (Homo sapiens, 203, 210); Ibero-Maurusian industry, 211; paleoenvironment, 231; Rodentia, 82-84 Mahemspan, South Africa: Bovidae, 551, 555 Mahmoud Formation, Tunisia, 26 Majiwa, Kenya: Deinotheriidae, 321 Makamby, Madagascar: Sirenia, 576 Makapan Valley, geomorphological setting, 227 Makapania, 562, 568; broomi, 561-562, 567, 569 Makapanian faunal span, 159 Makapansgat Limeworks, South Africa, 37, 38, 224; age, 159; Bovidae, 541, 569 (Antilopinae, 558, 559, 560; Bovinae, 544, 545; Caprinae, 561; Cephalophinae, 549; Hippotraginae, 550); Carnivora, 262; Cercopithecidae, 103, 104, 106, 107, 109, 110; Chalicotheriidae, 368, 369; Chiroptera, 65, 66; Equidae, 399, 405; Giraffidae, 526, 527;

632

Makapansgat Limeworks, (Continued) Hominidae, 167, 171, 172, 183; Hyracoidea, 310, 311; Insectivora, 57, 59, 61, 63, 64; paleoenvironment, 228, 229; Rhinocerotidae, 375; Rodentia, 70; Suidae, 458, 470, 471; Tubulidentata, 273 Malagasy, Republic: Hippopotamidae, 492. See also Madagascar. Malawi: Equidae, 399, 404; Giraffidae, 522, 525 Malembe, Congo, 27; Carnivora, 253; Chalicotheriidae, 369; Sirenia, 576; Suidae, 441, 444 Mali: Equidae, 408 Mammonteus, 357;primigenius, 357 Mammoths, 328, 357 Mammut, 342, 358, 359 Mammuthus, 35, 36, 352, 357-358, 360, 362-363, 364; africanavus, 353, 357, 358, 363; meridionalis, 36, 353, 357, 358, 363, 364;primigenius, 357; scotti, 357; subplanifrons, 37, 351, 357-358, 359, 363 Mammutidae, 26, 109, 334, 336, 338, 349, 358-359, 360, 364, 606; classification, 339; identification, 342 Mammutoidea, 339, 358-360 Man. See Homo Manatees. See Sirenia. River, see Trichechus Manatherium, 576 Manatus, 578; coulombi, 575 Mandrills. See Mandrillus Mandrillus, 5, 104, 109 Mangabeys. See Cercocebus Manidae, 270, 274 Manis, 268, 270; aurita, 269, 270;javanica, 270; lydekkeri, 270;palaeojavanica, 270; pentadactyla, 270, 277; tetradactyla, 269 Mansoura, Algeria: Bovidae (Antilopinae, 560; Bovinae, 545; Hippotraginae, 554); Suidae, 460 Marada Formation, Libya: Cetacea, 584 Maragha, Iran: Bovidae, 563, 565; Rhinocerotidae, 374 Marceau, Algeria, 27; Bovidae, 547, 550; Cercopithecidae, 102, 103, 108; Proboscidea, 359 Marewa, Uganda, 28 Marmoset, common. See Callithrix Marsabit Road, Kenya: Bovidae, 551; Camelidae, 538; Equidae, 405 Marsupiais, 54 Masai, as pastoralists, 13 Mascara, Algeria, 36; Equidae, 394 Masek Beds, Olduvai Gorge, 226, 230; Hominidae, 205 Masrasector aegypticum, 252 Masrisiren, 575 Masritherium, 424, 427-428, 431; aequitoralis, 428, 431; depereti, 424, 427, 428, 431 Massif, Ethiopian, 17 Massospondylus, 53 Mastodon, 27, 341, 342, 359; angustidens, 341, 342; angustidens libyca, 342, 343; angustidens pygmaeus, 343; aruernensis, 345; borsoni, 359; spenceri, 343 Mastodontidae, 340, 358-359 Mastodonts. See Mammutidae

Index

Matjes River, South Africa: Homo sapiens, 213 Matuyama Reversed Epoch, 199, 200 "Mauicetus" 595 Mbagathi, Kenya: Suidae, 445 M'Bodione Daders, Senegal, 25 Mediterranean, 23 Megaceroides, 497 Megachiroptera, 4, 5, 11, 66 Megadermatidae, 65, 66, 67 Megaloceros, 497; algericus, 497 Megaloglossus, 4 Megalohyrax, 142, 297-299, 302, 312; and Bunohyrax compared, 294, 295; championi, 289, 300, 301; classification, 287, 288, 289; eocaenus, 287, 297, 298-299, 303; and Meroehyrax compared, 308; minor, 287, 298, 299, 303; niloticus, 288, 298, 299; and Pachyhyrax compared, 290, 300, 301, 302; palaeotheroides, 287, 303;pygmaeus, 288, 293, 299, 300; suillus, 288, 295, 298, 299; and Titanohyrax compared, 303, 304 Megalotragus, 555, 568; eucornutus, 555; kattwinkeli, 555; priscus, 555 Megalovis, 567; latifrons, 562 Megantereon, 260, 262; eurynodon, 262; praecox, 262; problematicus, 262 Meganthropus, 192; africanus, 163; palaeojavanicus, 192 Megapedetes, 78, 80, 81, 82, 85 Megatheriidae, 370 Megazostrodon, 46-53; brain size, 51; classification, 46; dentition, 48-49; postcranial skeleton, 51; rudnerae, 21, 46-53, 604 Megistotherium, 253, 486 Melanorosaurus, 53 Melka el Ouidane, Morocco, 27; Proboscidea, 345 Melka Kontoure, Ethiopia: artifacts in, 226; Homo erectus, 200 Melkbos, South Africa: Bovidae, 541, 548, 558 Mellalomys, 83; atlasi, 83 Mellivora, 5, 7, 255, 256; capensis, 255, 256 Mellivorinae, 255, 256, 265 Mena House, Egypt, 35 Menalla, Chad: Proboscidea, 350 Menelikia, 551, 568; lyrocera, 551, 569 Menotyphla, 2 Meriones, 83, 84 Meroehyrax, 308, 311, 312; bateae, 289, 308, 310, 312; characteristics, 285; classification, 289, 307, 309, 310, 312; evolution, 290 Merychippus, 389 Merycopotamus, 26, 424, 429-430, 433, 486; anisae, 424, 429, 430, 433; dissimilis, 429, 430, 433;petrocchii, 424, 429, 430, 433 Merycops africanus, 429 Mesak Beds, Olduvai Gorge, 34 Mesembriportax, 568; acrae, 546 Mesocetus, 587; schuieinfurthi, 587 Mesochoerus, 462, 463, 465; heseloni, 463; lategani, 462; limnetes, 474; olduvaiensis, 479 "Mesochoerus paiceae," 457, 462

Mesonychidae, 251, 592, 593 Mesopithecus, 108, 110 Mesoplodon, 584, 585 Mesozoic mammals, 46-54, 604; associated faunas, 53; biology of, 51-52; Brancatherulum tendaguruense, 54; Megazostrodon and Erythrotherium, 46-53 Metacheiromyidae, 270 Metailurus, 260; africanus, 260 Metaphiomyinae, 79 Metaphiomys, 76, 77, 78-79 Metapterodon, 253; kaiseri, 253 Metarchidiskodon, 353 Metasayimys, 82, 83 Metaschizotherium: bavaricum, 369; hennigi, 368, 369; transvaalensis, 368, 369 Metasinopa, 252; ethiopica, 252; fraasi, 251 Metatherian mammals, 276 Metaxytherium, 576, 577, 578 Metoldobotes, 59; stromeri, 57, 58 Metridiochoerus, 470, 471-475, 476, 477, 479; andrewsi, 471, 472-473, 474, 475, 526Jacksoni, 470, 471, 473-474, 475; meadowsi, 474; nyanzae, 474475;pygmaeus, 474 Mfwanganu Island, Kenya, 28; Chalicotheriidae, 368; Deinotheriidae, 321; Hyracoidea, 302; Primates, 142 (Dendropitheus, 140, 141; Proconsul, 133, 134, 135, 136, 137); Rodentia, 70; Tragulidae, 537; Tubulidentata, 271, 273. See also Mwafanganu Miacidae, 257 Mice. See Rodentia Microchiroptera, 5, 7, 10, 11, 65 Microdyromys, 82, 83 Microfelis, 7, 260 Microgale, 64 Micromeryx, 497 Micropotamogale, 63, 64 Microstonyx, 452 Microvertebrates, 159 Mid-Tertiary surface, 23 Miniopterus schreibersi, 65 Miocene Epoch: Anthracotheriidae, 423, 424, 427-429, 431, 433, 486; Bovidae, 547, 563, 564, 567; Carnivora, 249, 250, 251, 252, 263, 264, 265 (Canidae, 253, 254-255; Creodonta, 253; Felidae, 260, 262; Hyaenidae, 257, 258, 259, 260; Mustelidae, 255, 256; Viverridae, 257); Cercopithecidae, 101-112; Cervidae, 496; Cetacea, 582, 584, 591, 592, 596; Chalicotheriidae, 368, 369; Chiroptera, 65, 66; Deinotheriidae, 315, 316, 321, 322, 323, 324, 325; Desmostylia, 333; East African primates, 124-127, 134; 135, 143, 144; Equidae, 379, 388 {Equus, 414; Hipparion, 389396, 401); extinctions, 612-613; geological developments, 22, 23; Giraffidae, 509, 511, 528, 529, 530 (Giraffinae, 518-522; Palaeotraginae, 513-518; Sivatheriinae, 522-527); Hippopotamidae, 486, 489, 492; Hominidae, 155, 158, 159, 172, 224 {Ramapithecus, 156); Hyracoidea, 284, 288, 289, 294, 299, 300, 307-312; immigration to Africa, 80-82; Insectivora, 56, 57, 60,

Index

Miocene Epoch, (Continued) 61, 63, 64, 65; Lagomorpha, 84; localities of fossil mammals, 26-30; Lorisidae, 91, 92-98; Palaeomerycidae, 498, 500; paleoenvironments of Hominidae, 227; pattern of faunal evolution, 605609, 611, 612, 613; peneplain, 23; Pholidota, 270, 603; Pongidae, 120-121, 128, 142, 143, 144, 151 (Dendropithecus, 140-141; Pliopithecus, 139, 140; Proconsul, 132, 134, 135, 136); Proboscidea, 338, 360, 364 (Gomphotherioidea, 339, 341-345, 349, 350, 351; Mammutoidea, 339, 359); Rhinocerotidae, 371-376; Rodentia, 70, 71, 7483, 85-86; Sirenia, 573-587; Suidae, 441-457, 477; Tragulidae, 536, 537; Tubulidentata, 270, 271, 273, 274 Mioechinus oeningensis, 60 Mioeuoticus, 91-92, 93, 94, 95, 96; bishopi, 91, 93, 96; talapoin, 115, 116, 129 Miosiren, 574, 575 Miosireninae, 574 Miosorex, 61; grivensis, 61 Miotragocerus, 547, 565, 568 Mirocetus, 595 Mirounga leonina, 263 Mixodectidae, 58 Mixohyrax, 287, 288, 298; andrewsi, 287, 298; niloticus, 287, 297, 298; suillus, 287, 298 Mockesdam, South Africa: Bovidae, 551, 555 Modder River, South Africa: Bovidae, 555 Moeripithecus, 114, 129-130; markgrafi, 124, 129 Moeritheriidae, 333, 336, 337, 573 Moeritherioidea, 330, 333-335; affinities, 335; extinction, 605 Moeritherium, 25, 328, 329, 333-335, 604; ancestrale, 334; andrewsi, 334; gracile, 334; lyonsi, 334, 335; pandionis, 334; trigodon, 334-335 Mogadishu, Somalia: Sirenia, 577 Moghara, Egypt, 26; Anthracotheriidae, 424, 427, 428, 430, 431; Cercopithecidae, 111; Cetacea, 584, 591, 592; Deinotheriidae, 321; Proboscidea, 342, 343, 344; Rhinocerotidae, 371, 372, 373, 376 Moinik Formation, Tanzania, 34 Mokattam Formation, Egypt: Cetacea, 585, 587; Marine mammals, 25; Sirenia, 575 Moles. Golden, see Chrysochloridae. Marsupial, see Notoryctes Mollusca, 227; Pliocene-Pleistocene, 30, 34 Molossidae, 66, 67 Monachus, 37, 263; monachus, 262 Mongolia, Bovidae specimens, 563, 564 Mongoose. See Herpestinae Monkeys, 5, 606, 608; African Swamp, 129; catarrhine, 123, 124, 128, 143; colobus, 5; New World, 100; Old World, 100, 102, 113, 116, 120, 121; Oligocene, 100, 112-117; owl, see Aotus trivirgatus; parapithecine, 142; patas, see Cercopithecus patas; platyrrhine, 123; spider, see Ateles; vervet, see Cercopith-

ecus aethiops. See also Cebidae; Cercopithecidae Monotremata, 46, 53, 54 Montane forest, 20 Montpelier: Hyracoidea, 289 Morganucodon, 46, 48-49, 51, 53, 604 Morganucodonta, 46 Morganucodontidae, 21, 48-49, 53; brain size, 51; distribution of, 52; postcranial skeleton of, 51; relationships of, 52-53 Morocco, 27, 231; Bovidae, 541, 556; Carnivora, 249, 255; Cervidae, 496, 497; Chiroptera, 66; Equidae (Equus, 408, 412-Hipparion, 390, 399); geological development, 22; Giraffidae, 516, 522; Hominidae, 159 (Homo sapiens, 203, 205); Proboscidea, 345, 353, 354, 358; Rodentia, 70, 71, 75; Sirenia, 575 Moroto, Uganda: Deinotheriidae specimens from, 321; Miocene fossils from, 28; Primates, 126, 127, 134, 135, 142; Proboscidea specimens from, 342, 359 Morrison Formation, North America, 54 Moruorot Hill, Kenya, 28; Bovidae, 564; Deinotheriidae, 321; Hyracoidea, 301302, 310; Palaeomerycidae, 498, 501, 502; Primates, 134, 135; Proboscidea, 342; Rhinocerotidae, 371, 373; Suidae, 445; Tragulidae, 537 Moschidae, 497 Moschus, 504, 537 Mount Cameroon, 6 Mount Elgon: Palaeomerycidae, 502 Mousterian associations: Cervidae, 497; Equidae, 412 Mousterian industry, 210, 562; Lavellois facies, 210, 211 Mozambique, geological development, 22 Mpesida Beds, Kenya, 30, 37; Bovidae, 546, 561; Hippopotamidae, 486-487, 488; Proboscidea, 350, 360; Rhinocerotidae, 372, 375 Mugharet el'Aliya, Morocco: Homo sapiens, 210, 211, 212 Multituberculata, 46, 53 Mumba Hills, Tanzania: Hominidae, 159 Mumbwa, Zambia, Homo sapiens, 213 Mungos, 257; mungo, 257 Muntiacus, 496 Muriculus, 10 Muridae, 2, 7, 10, 11, 70, 75, 81, 83, 84 Murinae, 10, 69, 81, 82, 83, 86 Muroidea, 69, 70, 75, 86 Mursi Formation, Ethiopia, 32, 37; Bovidae, 545, 558; Equidae, 390, 394; Hippopotamidae, 489, 491; Proboscidea, 353; Rhinocerotidae, 374, 375; Suidae, 456, 458 Muruyur, Kenya: Deinotheriidae, 321 Mus, 84; musculus, 80, 84 Muscardinidae, 5, 10, 11; distribution, 69 Muskox. See Budorcas taxicolor; Ovibovini Mustela, 256; nivalis, 255, 256; numidica, 256; putorius, 256 Mustelidae, 2, 5, 7, 11, 255-256, 607; numbers of, 263 Mustelinae, 255, 256 Mwafanganu Island: Lorisidae, 92, 93. See also Mfwanganu Mylomygale, 59-60; spiersi, 57

633

Myocricetodon, 83; cherifiensis, 83; irhoudi, 81; parvus, 81, 83 Myocricetodontinae, 81, 82, 83, 85, 86 Myohyracidae, 288, 289 Myohyracinae, 58, 59, 60, 65, 288 Myohyrax: doederleini, 57; osborni, 52 (see also Protypotheroides beetzi); oswaldi, 57, 58, 288; skull, 58 Myomorpha, 75; definition, 70 Myonycteris, 4 Myophiomyidae, 71, 77, 78-79, 80 Myophiomys, 78-79 Myorycteropus, 271, 273, 275, 276; africanus, 270, 271 Myosciurus pumilio, 5 Myosorex, 61, 62; cafer, 62\geatus, 62; robinsoni, 61, 62; schalleri, 62; varius, 62 Myotis, 5, 66 Myrmecophagidae, 274, 275-277 Mysticeti, 583, 595, 596, 597 Mystromyinae, 75 Mystromys, 70, 81, 82 Myxomygale, 60 Myxomygalinae, 60 Myzopoda, 65, 66; aurita, 66 Myzopodidae, 65, 66, 67, 608, 613 Nagor, 550 Nagri Formation, Siwaliks: Bovidae, 546, 563; Ramapithecus, 147-148 Nairobi Park, 9 Naisiusu Beds, Olduvai Gorge, 34 Nakali Tuffs, Kenya: Deinotheriidae, 323; Giraffidae, 517; Rodentia, 81 Nakuru-Naivasha Basin, Kenya, 35 Namib Desert, Namibia, 17, 27, 605; Bovidae, 564; Insectivora, 57; Lagomorpha, 84; Rodentia, 70, 71, 74, 75, 79, 80; Suidae, 444, 447; Tragulidae, 536, 537 Nandinia, 256; binotata, 5, 11, 256 Nandiniinae, 5 Napak, Uganda, 28; Carnivora, 257; Cercopithecidae, 111; Chalicotheriidae, 368, 369; Insectivora, 63; Lorisidae, 91, 92, 93; Primates, 126, 127, 134, 135, 138, 142; Proboscidea, 342; Rhinocerotidae, 372, 373, 374, 376; Rodentia, 70, 79; Tubulidentata, 271 Nasilio. See Elephantulus fuscus leakeyi Natal: Rodentia, 75 National parks of Africa, 12, 14 Natodameri, Sudan: Proboscidea, 356 Nautilus Beds, Somalia: Sirenia, 577 Ndutu Beds, Olduvai Gorge, 34 Neanderthals, 210-213 Near East: Equidae, 390, 391, 395 Nearctic: Bovidae, 567 Necrolestes, 64 Necromanis, 270 Negroes, 213, 215, 217 Nelson Bay, South Africa: Bovidae, 549; Pholidota, 268 Neodermus, 578 Neogene Period: Cetacea, 597, 599; Deinotheriidae, 315; Hominidae, 156; Paleoenvironment, 102; Rhinocerotidae, 372, 376; Sirenia, 577; Suidae, 435 Neolithic: Bovidae, 555; Cervidae, 497;

634

Neolithic, (Continued) Equidae, 412, 418; Giraffidae, 509; Hippopotamidae, 493; Suidae, 469 Neolithic industry, 211 Neoschizotherium, 288; rossignoli, 289 Neosciuromys, 78 Neotragini, 6, 541, 558-559, 567 Neotragus, 6, 8, 13, 558 Nesomyidae, 70, 75 Nesomyinae, 11, 75, 85 Nesotragus, 8, 558 New Zealand: Cetacea, 584, 595 Ngandong, Java: Hominidae, 196, 216 Ngorora Formation, Kenya: Bovidae (Antilopinae, 559; Bovinae, 546, 547; Caprinae, 563; Cephalophinae, 550); Carnivora (Canidae, 254; Hyaenidae, 259; Mustelidae, 255); Cercopithecidae, 108; Deinotheriidae, 321, 322; Equidae, 389, 391; Hippopotamidae, 486-487, 488; Hominidae, 156, 159, 160, 223; paleoenvironment, 30, 227; Primates, 127; Proboscidea, 344, 351, 360; Rhinocerotidae, 372, 373, 374, 376; Rodentia, 75, 81; Tayassuidae, 477; Tragulidae, 537 Nigeria, 13, 22; Cetacea, 584; geological development, 22; Giraffidae, 522 Nile Valley: Cervidae, 497; Equidae, 412, 414; Hippopotamidae, 483; Sirenia, 576 Nilgai, Indian. See Boselaphini Nimravus, 261 Norfolk Forest Bed, England: Equidae, 404 Norkilili Member Tuff, Tanzania, 205 North Africa: Anthracotheriidae, 424, 427, 430, 431, 486; Bovidae, 541, 567, 568 (Antilopinae, 560; Bovinae, 543, 548, 549; Caprinae, 562; Hippotraginae, 550, 554); Camelidae, 538; Carnivora, 249 (Felidae, 262; Mustelidae, 255); Cercopithecidae, 101, 102, 108; Cervidae, 496-497; Cetacea, 595; correlation with East Africa, 40; Deinotheriidae, 315, 321; Equidae, 379, 417, 418 (Equus, 403-404, 408, 409, 412; Hipparion, 388-392, 394-396, 398401); Giraffidae, 509, 517, 527, 528; Hippopotamidae, 487, 489, 492, 493; Hominidae, 161; Hyracoidea, 289, 294; Miocene fossils, 26-27; pattern of faunal evolution, 609 (Miocene, 605, 606; Oligocene, 604-605); PliocenePleistocene fossils, 35-37; Proboscidea, 363, 364 (Gomphotherioidea, 341, 343, 344, 349, 354, 358; Mammutoidea, 358, 359); Rhinocerotidae, 374; Rodentia, 74, 75, 82-84, 86; Sirenia, 573; Suidae, 436, 451, 452, 453, 460, 462, 464, 469 North America, 592; Anthracotheriidae, 423; Bovidae, 548, 561; Carnivora, 251, 255; Cetacea, 584, 591, 593, 595, 596, 599; Chalicotheriidae, 368; Equidae, 415, 416, 417 (Equus, 403, 404, 408; Hipparion, 389); Jurassic Period, 54; Palaeomerycidae, 497; Pholidota, 270; Proboscidea, 360, 364 (Gomphotherioidea, 339, 342, 344, 351; Mammutoidea, 358); rodent migration, 86-87; Rodentia, 75, 76; Sirenia, 574, 576, 577, 578; Tayassuidae, 435, 476; Triassic Period, 53; Tubulidentata, 270

Index

North Nyabrogo, Uganda: Suidae, 464 North Pacific: Sirenia, 574 Notharctus, 143 Notochoerus, 36, 37, 453, 457-460, 462, 477; broomi, 470; capensis, 457, 458, 459, 460; compactus, 474, 475; dietrichi, 473; euilus, 458, 459-460, 465, 479; hopwoodi, 473; meadowsi, 473; and Metridiochoerus compared, 472, 474; scotti, 458, 460; serengetensis, 476; teeth of, 479 Notocricetodon, 78, 80 Notohipparion, 388; namaquense, 389, 390, 392, 393-394, 402 Notoryctes, 64 "Nubian Sandstone," Egypt, 22 Numidocapra, 563, 568; crassicornis, 562-563 Nyala. See Tragelaphus. Mountain, see Tragelaphus buxtoni Nyamavi Beds, Lake Albert, 28 Nyanzachoerus, 30, 102, 441, 451, 453457; devauxi, 454-455;jaegeri, 36, 37, 455, 457, 459-460; kanamensis, 453, 454; North African, 452; and Notochoerus compared, 459; origins, 477; pattersoni, 37, 456 (characteristics, 479; Hadar Formation, 465; and Notochoerus compared, 458; and Nyanzachoerus devauxi compared, 455; and Nyanzachoerus jaegeri compared, 457; and Nyanzachoerus kanamensis compared, 454);plicatus, 457; syrticus, 455; tulotos, 455-456 Nyawiega, Uganda: Bovidae, 551 Nycteridae, 66, 67 Nycticebus, 94, 98 Nycticeius (Scoteinus) schlieffeni, 65

Ochotonidae, 83, 84 Odontoceti, 583, 591-592, 595, 596, 597 Oegoceros, 550 Oenomys, 5 Oioceros, 500, 563, 567\grangeri, 563; noverca, 563; tanyceras, 563 Okapi. See Giraffidae Okapia, 514-515, 517-518, 528, 530; and Canthumeryx compared, 501; characteristics, 505, 511; and Climacoceras compared, 500; and Giraffa compared, 520, 521, 522; Ituri Forests, Zaire, 509; johnstoni, 6, 518, 520-521, 528, 530; andPalaeotragus compared, 515; stillei, 517, 518, 520-521, 522, 530; and Zarafa compared, 514, 515, 516 Okote Tuff Complex, 225 Oldowan industry, 176, 224-227; Chemoigut Formation, 225; Olduvai, 224225, 230 Olduvai Gorge, Tanzania, 30, 32, 34, 35, 36; artifacts, 224-226, 230; Bovidae, 541, 568, 569 (Alcelaphinae, 555, 556, 557, 558; Antilopinae, 559, 560; Bovinae, 543-544, 545, 548, 549; Caprinae, 562; Cephalophinae, 559; Hippotraginae, 550, 551-552, 553, 554); Camelidae, 538; Carnivora, 249 (Canidae, 254; Felidae, 262; Mustelidae, 255; Viverridae, 257); Cercopithecidae, 104, 107; Chalicotheriidae, 368, 369; coexistence of hominid taxa, 222-223;

correlation with South Africa, 38, 40; Chiroptera, 65, 66; Deinotheriidae, 323, 324; Equidae (Equus, 404, 407; Hipparion, 389-390, 399, 400); extinctions, 613; Giraffidae, 530 (Giraffinae, 519, 521; Sivatheriinae, 525, 526); Hippopotamidae, 491; Hominidae, 156, 158, 165, 172, 173, 191 (Australopithecus, 176, 177, 178, 185, 229; Homo, 186-187, 193, 222, 229, 230; Homo erectus, 192, 195, 196, 199-200); Insectivora, 57, 59, 60, 61, 62-63, 65; Lorisidae, 91; paleoenvironment, 229-230; Proboscidea, 355-356; Rhinocerotidae, 375; Rodentia, 70, 75, 81-82, 84; Suidae, 464-466, 469, 470, 473-476 Oligocene Epoch: Anthracotheriidae, 423, 427, 430, 431, 433; Bovidae, 564; Carnivora, 249, 250, 252, 253, 263, 264 (Creodonta, 253; Felidae, 260-261); Cercopithecidae, 100, 101, 112-117; Cervidae, 496; Cetacea, 583-584, 595, 599; Chalicotheriidae, 368; Chiroptera, 65; Embrithropoda, 280, 281, 603; fauna of Fayum, 84; geological developments, 22, 27; Giraffidae, 528; Hyracoidea, 284, 287-294, 297, 299, 302, 306; Insectivora, 56, 57, 65; localities of fossil mammals, 25-26; Moeritherioidea, 334, 335; Palaeomerycidae, 498; pattern of faunal evolution, 604-607, 609, 612, 613; Pecora, 564; Pholidota, 270, 603; Primates, 120, 127, 128, 143, 144 (Aegyptopithecus, 130, 131; Aeolopithecus, 139;Dendropithecus, 140; Dryopithecus, 132; from Egypt, 121-214); Proboscidea, 338, 364 (Gomphotherioidea, 340, 341, 342; Mammutoidea, 358); Rhinocerotidae, 371, 372, 373, 374, 376; Rodentia, 70, 71, 74, 75, 76-77; Sirenia, 573, 575, 576, 577-578; Suidae, 435, 442, 443, 444, 448, 452; Tayassuidae, 476; Tragulidae, 536 Oligokyphus, size of, 52 Oligopithecus, 124, 128-129, 131, 143; and Aeolopithecus compared, 139; savagei, 124, 128-129, 143 Olorgesailie, Kenya, 40; Cercopithecidae, 107, 108; Giraffidae, 524; hand axe remains, 35; Proboscidea, 356; Suidae, 465, 473 Ombo, Kenya, 28; Anthracotheriidae, 429; Cercopithecidae, 111; Deinotheriidae, 321; Primates, 138; Proboscidea, 342; Rhinocerotidae, 374; Suidae, 447 Omnivores, savanna zone, 7 Omo, Ethiopia, 30, 32-33; artifacts, 224; Bovidae, 560; Carnivora, 249, 254 (Felidae, 262; Hyaenidae, 257, 259; Mustelidae, 255; Viverridae, 257); Cercopithecidae, 102-105, 107, 108, 110; Chalicotheriidae, 368, 369; Deinotheriidae, 323, 324, 326; Equidae, 388, 391, 394, 395, 402, 405; Giraffidae, 519, 520, 521; Hippopotamidae, 489, 491-492; Hominidae, 156, 158, 172, 179, 186, 224; Insectivora, 63; paleoenvironment, 228; Proboscidea, 353; Suidae, 458, 459, 463, 464, 465, 474 Omochoerus, 462, 463, 464; maroccanus, 464; pachygnathus, 463

Index

Omoloxodon, 354 Onagers, 402 Ongoliba, Zaire: Cercopithecidae, 103 Orange Free State, South Africa: Giraffidae, 522; Suidae, 469 Orangiatherium, 523, 524, 525; vanrhyni, 525, 526, 527 Orangutan. See Pongo pygmaeus Orasius, 518 Oreonagor tournoueri, 556 Oreopithecus, 114; relationships, 116 Oreotragus, 558, 559; major, 559; oreotragus, 8, 11, 559 Oriental region: faunal relationship with Africa, 1 - 2 ; species in, 1 Ornithodelphia, 46 Orthostonyx, 475; brachyops, 475 Orycteropodidae, 2, 227, 608 Orycteropodinae, 275 Orycteropus, 2, 271-274, 275, 276, 277; afer, 270, 271, 273, 275, 276, 277; browni, 273; chemeldoi, 273, 274; crassidens, 273; depereti, 273; gaudryi, 270, 273, 274, 275; mauritanicus, 273; minutus, 273, 274; Miocene, 27;pilgrimi, 273; pottieri, 273; in Southwest Cape, 11 Oryx. Arabian, see Oryx leucoryx. Scimitar, see Oryx dammah Oryx, 8, 10, 547, 552, 553, 554, 561, 565; dammah, 552; domestication of, 12; gazella, 8, 10, 552; leucoryx, 552; sivalensis, 554 Oryzorictes, 11 Oryzorictinae, 63 Osbornictis, 5, 256; piscivora, 256 Ostrich, 109; Miocene fossils of, 27 Otocyon, 10, 253, 255; megalotis, 253, 254; recki, 254 Otomyinae, 7, 75 Otomys, 81, 82 Otters. See Aonyx; Lutra; Lutrinae Ouadi Derdemy, chad, 36 Oued Akrech, Morocco, 35 Oued Bou Sellam, Algeria: Bovidae, 548 Oued Constantine, Algeria: Proboscidea, 354 Oued el Atteuch, Algeria: Bovidae, 560 Oued el Hammam, Algeria, 26-27, 102; Bovidae, 563, 568 (Antilopinae, 561; Caprinae, 562); Carnivora, 258; Equidae, 389, 390, 391, 396, 398, 401; Rhinocerotidae, 374; Suidae, 454-455 Oued Fouarat, Morocco, 35 Oued Zra, Morocco: Rodentia, 70 Ouljian transgression, Homo sapiens from, 213 Ounianga Kebir, Chad, 36 Ourayia, 129 Ourebia, 558, 559 Ovibos, 567; moschatus, 561 Ovibovini, 541, 561-562, 565, 567 Oxyclaeninae, 251 Oxystomus, 578 Pachycetus robustus, 596 Pachyhyrax, 299-302, 308, 312; and Bunohyrax compared, 295; championi, 299, 300, 301-302; characteristics, 285; classification, 287, 289, 290, 293, 294; crassidentatus, 287, 299, 300, 301, 302; and Megalohyrax compared, 298; pygmaeus, 293, 299, 300-301, 302, 305

Pachyportax, 547, 565, 566 Pachytragus, 26, 563, 568; crassicornis, 563; solignaci, 547, 563 Pakistan: Anthracotheriidae, 429; Deinotheriidae, 315; Ramapithecus, 147, 149 Palaeanodonta, 270, 277 Palaearctic region, 85; Bovidae, 565, 566, 567; evolution of African rodents, 8 2 84 Palaeochoerus, 443, 444, 448 Palaeoerinaceus, 60 Palaeohypsodontus asiaticus, 564 Palaeoloxodon, 354; antiquus, 356; kuhni, 356, 357 Palaeomastodon, 25, 337, 338, 339, 340341, 342, 358; barroisi, 341; beadnelli, 340, 341; intermedius, 341; minor, 341; minus, 341; Palaeomastodon (Palaeomastodon), 339, 340, 341; Palaeomastodon (Phiomia), 339, 340, 341; parvus, 341; relationships, 328, 360; serridens, 341; wintoni, 341 Palaeomastodontidae, 340 Palaeomerycidae, 497-506, 513, 528, 564, 606, 607; relationships, 503-506; terminology, 498-499 Palaeomerycinae, 497 Palaeomerycini, 497 Palaeomeryx, 497, 498, 502, 506, 511, 515, 527, 528; africanus, 498, 563; bojani, 497; eminens, 497; and Palaeotragus compared, 516; and Zarafa compared, 513-514 Palaeoreas, 567 Palaeoryctes, 63 Palaeoryx, 562, 565, 567 Palaeothentoides, 59; africanus, 57 Palaeotherium, 286, 288, 368 Palaeotraginae, 511-518, 521, 528, 606 Palaeotragiscus longiceps, 559 Palaeotragus, 509, 513-517, 520, 527530; and Climacoceras compared, 500; expectans, 527; germaini, 509, 516-517, 523, 528, 529; primaevus, 509, 514517, 529 (and Canthumeryx compared, 501; and Giraffa compared, 520; relationships, 505, 516; and Samotherium compared, 517; and Zarafa compared, 514); quadricornis, 512, 527, 531; relationships, 505, 521, 523; rouenii, 501, 505, 516; tungurensis, 505, 528; and Zarafa compared, 513-515, 527 Paleocene Epoch: Equidae, 388; geological developments, 22; pattern of faunal evolution in, 604, 609; Primates, 113, 116 Paleoenvironments, 30, 102, 143, 227232, 452-453 Paleolithic: artifacts, 154, 182; Cervidae, 497; Equidae, 404, 412, 418 Paleozoic, geological development, 20, 22 Palestine: Bovidae, 560, 565, 566; Proboscidea, 355 Palikao. See Ternifine Pan, 125, 151, 168; paniscus, 5, 151, 152; and Proconsul compared, 133; troglodytes, 5 Panda, 151 Pangolins. See Pholidota Pannonian Basin: Ramapithecus, 218

635

Panthera, 260, 262, 264; crassidens, 262; leo, 260, 262; leo shawi, 262; leo spelaeus, 262, pardus, 5, 7, 11, 182, 260, 262 (and Hominidae, 225) Pantholops, 567 Pantolestes, 251 Pantolestidae, 56, 251 Pantotheria, 54 Papio, 5, 7, 11, 103, 104, 105, 106, 107, 108, 110; angusticeps, 104; baringensis, 104; cynocephalus, 104; cynocephalus "kindae" 104; cynocephalus ursinus, 105; dentition, 102; gelada, 10; hamadryas, 10, 104, 107; izodi, 104; robinsoni, 104; uiellsi, 104 Papionini, 102-104 Pappocetus, 585-587, 592, 595; lugardi, 584, 585-587, 590, 593, 598 Parabos, 549, 566; boodon, 549; cordieri, 547 Paracolobus, 109-110; chemeroni, 109110 Paracrocidura, 63 Paracryptomys, 77, 79 Paracynohyaenodon, 252 Paradiceros, 30, 375-376; mukirii, 371, 375 Paradoxurinae, 256 Paraechinus, 61, 67 Paraethomys, 83, 84 Paranomalurus, 78, 80 Paranthropus, 169 Paraonyx, 256; congica, 256; microdon, 256; philippsi, 256 Parapapio, 103-104, 105, 221, 222; antiquus, 103; broomi, 103; coronatus, 109; jonesi, 103, 104; major, 105; whitei, 103 Parapedetes, 80; namaquensis, 77 Paraphiomys, 76, 77, 78, 79, 81, 85; pigotti, 78 Parapithecidae, 112-113, 124, 605; Fayum, 112-113; relationships, 115117 Parapithecus, 100, 101, 112, 113, 114115, 120, 123; fraasi, 114, 115, 123; grangeri, 115, 116; and Oligopithecus compared, 128; relationships, 115-117, 124 Paratarsomys, 78, 80, 85 Paraulacodus, 79 Paraxerus, 7 Parelephas, 357; armeniacus, 357; columbi, 357; trogontherii, 357 Parestigorgen gadjingeri, 558 Parmularius, 557, 558, 561, 568, 569; altidens, 557, 558, 569; angusticornis, 557, 569; rugosus, 557, 569 Parurmiatherium, 567 Patriomanis, 270 Patterns of faunal evolution, 603-614; endemism, 609; extinction, 612-614; turnover, 609-612 Paurodontidae, 54 Pecarichoerus: africanus, 477; orientalis, 477 Peccaries. See Tayassuidae Pecora, 503-504, 506, 522, 536, 564 Pedetes, 2, 7, 10, 74, 75, 77, 78, 81, 82 Pedetidae, 7, 74-75; distribution, 6 9 - 7 0 Pedetoidea, 74, 80, 84, 85, 86; distribution, 69 Pelea, 11, 558; capreolus, 558

636

Pelomys, 84 Pelorocerus: broomi, 555; elegans, 555; helmei, 555 Pelorovis, 548-549, 566, 568; antiquus, 207, 217, 548, 549, 568, 569; oldowayensis, 548, 549, 555, 568, 569 Pelycodus, 129 Penguins, 37 Peninj, Tanzania, 34; Bovidae (Alcelaphinae, 557, 558; Antilopinae, 560; Bovinae, 544; Hippotraginae, 553); Suidae, 465 Pentacodontidae, 251 Pentacodontinae, 251 Pentalophodon, 345 Penultimate Glacial Complex: Homo sapiens, 203 Peramus, 54 Percrocuta, 37, 259; algeriensis, 259; australis, 259; tobieni, 259 Perissodactyla, 7, 26, 270, 286, 405, 605606; Chalicotheriidae, 368-370; Equidae, 379-418; and Hyracoidea compared, 312; Rhinocerotidae, 371-378 Permian Period, 20, 22 Perning, Java: Hominidae, 192 Perodicticus, 94, 95, 98; potto, 93, 97 Petralona, Macedonia: Homo sapiens, 203 Petrodromus, 59; tetradactylus, 4 Petromus, 78, 79; classification, 69, 70; distribution, 69; typicus, 10 Petromyidae, 10 Petromyscus, 83 Phacochoerinae, 436 Phacochoerus, 7, 458, 459, 467-476, 479; aethiopicus, 467, 469, 475; africanus, 436, 437, 440, 441, 468-469, 470; africanus helmei, 469; altidens, 476; antiquus, 470; barbarus, 469; characteristics, 437-438, 450-451; classification, 448; complectidens, 476; congolensis, 469; dreyeri, 469; helmei, 469; a n d H i p pohyus compared, 453; kabuae, 469; laticolumnatus, 469; mauritanicus, 469; meiringi, 469; modestus, 469-470; North African Pleistocene, 436; and Notochoerus compared, 460; and Nyanzachoerus compared, 456; origins, 452; robustus, 476; stenobunus, 469; venteri, 469 Phanerozoic sediments, 20 Phataginus: gigantus, 269, 270; longicaudata, 270; temmincki, 269, 270; tricuspis, 269, 270 Phenacotragus, 560 Philantomba monticola, 550 Phiocricetomys, 77 Phiomia, 25, 26, 327, 328, 340, 341, 345; osborni, 341;pygmaeus, 343; serridens, 341 Phiomorpha, 71, 74, 77-81, 84, 85; relation with Caviomorpha, 86-87 Phiomyidae, 71, 79 Phiomyoides, 78 Phiomys, 76-77, 78, 79; North African, 82 Phocidae, 262-263 Pholidota, 1, 5, 6, 268-270, 277, 368, 603, 609 Pholidota, North American. See Epoiocatherium Phosphorites du Quercy, 261

Index

Phyllostomatidae, 66 Phyllotillon, 368, 369 Physical setting in Africa, 17-40; fossil mammal localities, 23-40; geological development, 20-23; present environment, 17-20 Pigeon Cave, Homo sapiens, 211 Pigs. See Suidae; Bush, see Potamochoerus Pikermi, Greece: Bovidae, 562, 563, 565; Carnivora, 260; Giraffidae, 529; Hyracoidea, 286, 289 Pilgrimia, 354; antiqua recki, 355; archidiskodontoides, 356, 357; subantiqua, 356; wilmani, 356, 357; yorki, 356, 357 Pinjor Formation, Siwaliks: Bovidae, 552 554, 557, 561, 565, 566, 567 Pinniped, 576 Pipistrellus (Scotozous) rueppelli, 65 Pithecanthropus, 123, 192, 193 Platanista, 597 Platanistidae, 596, 597 Platybelodon, 339, 342, 343, 344; danovi, 344; filholi, 342; grangeri, 344 Platyosphys, 587, 594 Platyrrhini, 100 Platystoma, 578 Platystomus, 578 Pleisiorycteropus, 274 Pleistocene Epoch: Anthracotheriidae, 423; Aterian industry, 211; Bovidae, 541, 566, 567, 568, 569 (Alcelaphinae, 555, 556, 558; Antilopinae, 558, 560; Bovinae, 543, 548, 549; Hippotraginae, 550-551, 552, 553); Camelidae, 538; Carnivora, 249, 250, 263, 264, 265 (Canidae, 254-255; Felidae, 261-262; Hyaenidae, 260; Mustelidae, 255, 256; Ursidae, 255; Viverridae, 257); Cercopithecidae, 101-112; Cervidae, 496, 497; Cetacea, 584, 599; Chalicotheriidae, 368, 369; Chiroptera, 65, 66; Equidae, 379, 414-415, 416, 417, 418 CEquus, 403, 404, 406-410, 412, 413; Hipparion, 389-395, 397-402); extinctions, 612, 613; Giraffidae, 509, 529, 530 (Giraffinae, 518, 520, 521, 522; Sivatheriinae, 525, 526); Hippopotamidae, 483, 489, 491, 492; Hominidae, 158, 178, 185-186, 189, 224 (Homo erectus, 200; Homo sapiens, 201, 203207, 212-217); Hyracoidea, 284, 285, 308; Insectivora, 57, 60, 61, 64, 65; Lagomorpha, 84; localities of fossil mammals, 23; Maghreb, 198; pattern of faunal evolution, 608-609, 612, 613; Pholidota, 268, 270, 603; Proboscidea, 338, 360, 362, 363, 364 (Gomophotherioidea, 339, 345, 349, 351-358; Mammutoidea, 339, 359); Rhinocerotidae, 374, 375, 376; Rodentia, 70, 75, 76, 82, 83, 84, 86; Sirenia, 578; Suidae, 435, 436, 448, 451-479; Tayassuidae, 476; Tubulidentata, 273 Plesiadapis, 113 Plesiaddax, 567 Plesiorycteropodinae, 275 Plesiorycteropus, 274, 275, 276, 277; madagascariensis, 270, 274 Plesiosoricidae, 63 Plesippus, 403, 404, 416, 417; simplicidens, 403, 404, 407, 408, 416

Pliocene Epoch: Anthracotheriidae, 423, 424, 429, 487; Bovidae, 541, 564, 566 (Bovinae, 546, 547; Hippotraginae, 552); Camelidae, 538; Carnivora, 249, 250, 265 (Canidae, 254; Felidae, 262; Hyaenidae, 257, 259, 260; Mustelidae, 255; Phocidae, 263; Viverridae, 257); Cervidae, 496, 497; Cetacea, 584; Chalicotheriidae, 369; Chiroptera, 65, 66; dating technique, 185; Desmostylia, 333; Equidae, 379, 415, 416, 417 (Equus, 402, 403, 404, 406, 408; Hipparion, 389-399, 401-402); extinctions, 612-613; Giraffidae, 509, 528, 529, 530 (Giraffinae, 520, 521, 522; Palaeotraginae, 517; Sivatheriinae, 523, 524, 525, 526); Hippopotamidae, 486, 487; Hominidae, 155, 156, 163, 171, 224 (Homo, 222); Hyracoidea, 284, 286, 288, 289, 307, 308, 312; Insectivora, 57, 60, 61, 64, 65; pattern of faunal evolution, 607-608, 612, 613; Pholidota, 268; Proboscidea, 360, 362, 363, 364 (Gomphotherioidea, 345, 349-358; Mammutoidea, 358, 359); Primates, 101, 132, 143; Rhinocerotidae, 371, 375, 376; Rodentia, 70, 74, 75, 76, 81, 83, 84; Sirenia, 574, 575, 577; Suidae, 435, 441, 442, 444, 448, 451-455, 462, 477, 479; Tayassuidae, 476, 477; Tubulidentata, 270, 273 Pliocene-Pleistocene: Carnivora, 249 (Mustelidae, 255, 256; Viverridae, 257); Cercopithecidae, 101, 103, 106, 110; coexistence of hominid taxa, 222-223; Deinotheriidae, 315, 323; Hippopotamidae, 492; Hominidae, 147, 151, 156, 158, 159, 169, 172, 176; Hyracoidea, 310, 311; localities of East African fossils, 30-35, 36, 37; localities of North African fossils, 35-37; localities of South African fossils, 37-40; Lorisidae, 91 Pliohyracidae, 286, 288, 289-308, 312, 606, 607 Pliohyracinae, 289, 290, 307-308, 312 Pliohyrax, 285, 286, 289, 290, 302, 307, 308, 309, 312; antiqua, 289; capensis, 289; championi, 288, 289; graecus, 289; kruppii, 286; obermeyerae, 289; occidentalis, 289; robertsi, 289; rossignoli, 289; transvaalensis, 289 Pliopithecus, 129, 130, 144; and Aegyptopithecus compared, 130-133; and Aeolopithecus compared, 139; and Dendropithecus compared, 140-141; (Limnopithecus) legetet, 138; (Limnopithecus) macinnesi, 140 Podocarpus, 230 Poecilictis, 256; libyca, 256 Poecilogale, 7, 256; albinucha, 256 Poiana, 5, 256; richardsonii, 256 Pointe-Pescade cave: Suidae, 460 Polecat. See Ictonyx striatus Pomatodelphis, 591 Pomonomys, 78, 79 Pondaungia, 112; cotteri, 120 Pongidae, 2, 120-144, 165-167, 218, 605, 606, 608; classification, 100, 152; dentition, 149, 150-151, 169; distribution, 141-143; East African, 121, 124-126, 144; habitat, 5, 227; Miocene, 121,

Index

Pongidae, (Continued) 124-126, 144, 147, 159; origins, 123 Ponginae, 128 Pongo, 141; pygmaeus, 144 Pongola Reserve, Transvaal, 12 Pontian Stage: Carnivora, 262; Cervidae, 497; Chalicotheriidae, 369; Deinotheriidae, 315, 324; Equidae, 396; Rhinocerotidae, 372, 373, 374, 375, 376; Rodentia, 74; Suidae, 443, 445; Tubulidentata, 270 Pontogeneus brachyspondylus, 591 Pontotherium, 577 Porcupine. Old World, see Hystricidae. Tree, see Atherurus Porcupine Cave, Ethiopia: Homo sapiens, 216, 217 Porpoises, See Cetacea Port de Mastaganem, Algeria: Proboscidea, 356 Postpalerinaceus vireti, 60 Postschizotherium, 288, 289, 307, 308, 312; chardini, 288; intermedia, 289; licenti, 289 Potamochoeroides, 470-471, 472, 477; hypsodon, 470; shawi, 470-471, 479 Potamochoerops, 469 Potamochoerus, 6, 11, 436, 455, 461-462, 477, 479; classification, 448; dentition, 449, 462, 463; habitat, 452; and Hylochoerus compared, 466; intermedins, 464; and Kolpochoerus compared, 463, 465; larvatus, 436; and Lopholistriodon compared, 446; major, 466; majus, 464; and metridiochoerus compared, 474; and Nyanzachoerus compared, 453, 456; and Phacochoerus compared, 467, 468, 470;porcus, 436-440, 453, 461462, 477; and Potamochoeroides compared, 471; and Sus compared, 437439, 450, 460 Potamogale, 63, 64 Potamogalidae, 2, 4 Potamogalinae, 63, 64, 65 Pottos, 5, 60 Power's Site, Vaal River: Bovidae, 559 Praedamalis deturi, 554 Praemadoqua, 559 Praesorex, 63 Praomys, 7, 84; morio, 5; pomeli, 84; verreauxi, 11 Precambrian geological development, 20 Prechoeropsis ,483 Present-day mammals of Africa, 1 - 1 4 Presoltian: Hominidae, 159, 197, 212 Priai de Morrungusu, Mozambique: Deinotheriidae, 323 Primates, 2, 5, 7, 11, 100, 605, 606, 607; biomass, 6; Cercopithecidae, 100-112; East African Miocene, 124-127; Egyptian Oligocene, 121-124; Gebel Qatrani Formation, 25; Hominidae, 147152, 154-232; Lorisidae, 90-98; Madagascar, 11; Parapithecidae, 100-101, 112-117; Pongidae, 120-144; See also under individual taxa Primelephas, 351-352, 360\ gomphotheroides, 351, 352; korotorensis, 36, 351, 352 Prionodelphis, 263; capensis, 263; rovereti, 263

637

Pristiphoca occitana, 263 Proamblysomus antiquus, 64 Proamphibos, 548, 565, 566 Probainognathus, brain size, 49 Proboscidea, 1, 5 - 6 , 7, 9, 11, 21, 25, 26, 32, 35, 37, 270, 336-364, 604, 606, 607, 608; diet, 370; domestication, 12; Elephantidae, 349-358, 360-364; extinctions, 364; fossils in Africa, 339360; Gomphotheriidae, 339-349, 363364; and Hippopotamidae compared, 483; and Hyracoidea compared, 284, 285, 287, 312; Mammutidae, 358-360; phylogeny and evolution, 360-363; and Sirenia compared, 573, 574, 575; structure, 338-339; zoogeography, 363-364 Procamptoceras brivatense, 563 Procapra, 567

Pronotochoerus, 470, 473; nyanzae, 474 "Pronotochoerus": jacksoni, 470, 474; shawi, 470 Propalaeochoerus, 441, 443, 444, 447, 448-449, 452; leptodon, 444 Propalaeomeryx, 497; sivalensis, 497 Propalaeoryx, 498, 502-503, 506, 564; austroafricanus, 502-503, 506, 564; nyanzae, 503, 506, 564 Propliopithecus, 123, 124, 128, 129-130, 143; and Aegyptopithecus compared, 130-132; and Aeolopithecus compared, 139; classification, 125, 127; and Dendropithecus compared, 140; haeckeli, 123, 129-130; markgrafi, 129-130; and Oligopithecus compared, 128-129; and Parapithecus compared, 114; and Prohylobates compared, 111, 112 Propotamochoerus, 455, 456, 462, 477; de-

Procavia, 286, 288, 289, 308, 309, 310311, 312; antiqua, 288, 310, 311; antiquus, 288; capensis, 2, 11, 310; habessinica, 310; johnstoni, 310; obermeyerae, 288, 311; robertsi, 288, 311; ruficeps, 310; transvaalensis, 288, 310, 311 Procaviidae, 2, 286, 288, 289, 290, 308311, 312, 606, 607 Procervulus, 497, 498 Prochrysochloris, 64, 142; miocaenicus, 64 Proconsul, 132-137, 140, 144; and Aegyptopithecus compared, 124, 130-131; africanus, 124-125, 127, 132-134, 142, 143, 144 (and Limnopithecus compared, 138; and Proconsul gordoni compared, 136; and Proconsul major compared, 135); classification, 126, 127; evolution, 125, 138; gordoni, 132, 135-136, 137, 138, 142, 143, 144; major, 125, 126, 127, 134-135, 142, 143, 144; nyanzae, 125, 127, 134, 135, 142, 143, 144; vancouveringi, 127, 136-137, 138, 142, 143, 144 Prodeinotherium, 317, 320-326, 327, 328, 329; bavaricum, 320, 321, 322, 323; hobleyi, 320-323, 324, 328; orlouii, 320; pentapotamiae, 320, 321, 322 Prodissopsalis, 252 Prodremotherium, 564 Profelis, 260 Progalago, 91, 93, 94, 95, 98; dorae, 91, 92, 93, 94, 95, 96; minor, 91; robustus, 91; songhorensis, 91, 93, 95, 96 Progenetta, 260 Progiraffa, 498 Progonomys, 83, 84; anomalus, 84; cathalai, 83; miocaenicus, 83 Proheliophobius, 78, 79, 80, 81 Prohylobates, 26, 111-112; tandyi, 111112 Prohyrax, 308-310, 311, 312; tertiarius, 288, 309, 310, 312 Prolagus sardus, 84 Prolibytherium, 500, 503, 522; and Canthumeryx compared, 502; and Climacoceras compared, 499; and Giraffidae compared, 509, 511, 528, 529; magnieri, 506, 522, 527, 528, 530 Promesochoerus, 462; mukiri, 464 Pronghorn, North American. See Antilocapridae Pronolagus, 84

vauxi, 455 Propotto, 66; leakeyi, 65, 66, 91 Prorastomidae, 574 Prorastomus, 574, 575 Prosauropoda, 53 Prosimian primates, 5, 90-98, 100, 101, 605; characteristics, 90; collections, 91; evolutionary relationships, 97-98; Miocene, 94-97; fossil sites, 92-93; Galaginae, 93-97; history of nomenclature, 91-92; living species, 90; Lorisinae, 94-97; taxonomy, 93 Prosinotragus, 563 Protalattaga, 83; grabaui, 83 Protanancus macinnesi, 342, 343 Protatera, 83 Protechinus, 61; salis, 60 Proteles, 2, 7, 10, 11; cristatus, 258 Protenrec, 63-64, 142; tricuspis, 63 Protocetidae, 585-587, 591, 592, 595, 596, 597, 598 Protocetus, 585, 587, 588, 592, 593, 595, 598; atavus, 585, 587, 590, 591, 592, 593, 596, 598; isis, 590; zitteli, 589 Protohominids, habitat adjustments of, 227 Protosiren, 574, 575; dolloi, 575; dubia, 575-Jraasi, 575, 585 Protosirenidae, 574, 575 Prototheria, 46, 53 Prototherium, 576; veronense, 575 Protoxerus, 5 Protragocerus, 500, 547, 563, 606; gluten, 565; labidotus, 547, 563 Protypotheroides, 60, 288; beetzi, 57, 58, 288; dentition, 58 Provampyrus orientalis, 65 Proviverrinae, 251-252, 253 Prox, 497 Prozeuglodon, 587, 588, 590-591, 594595, 599; atrox, 590; isis, 587, 588, 590-591, 595, 598; stromeri, 590 Prozeuglodontidae, 587-588 Pseudaelurus, 260 Pseudoquagga, 404 Pseudotragus, 567; potwaricus, 563 Pterodon, 253; biincisivus, 253; dasyurroides, 253 Pteromys, 2 Pteropodidae, 65, 66, 67 Ptolemaia, 251; lyonsi, 56; relationships, 56, 57 Ptolemaiidae, 56-57, 65, 251

638

Pucangan Formation, Java: Hominidae, 192 Pugmeodon, 576 Pultiphagonid.es, 556; africanus, 551 Pygathrix, 110 Pygmies, hunting people, 13 Pythons, 126 Qarunavus: dentition, 57; meyeri, 56 Qasr el Sagha Formation, Fayum, 25, 287; Anthracotheriidae, 424; Barytheriidae, 329; Carnivora, 252; Cetacea, 588, 589, 590, 591, 593, 598; Hyracoidea, 287; Moeritherioidea, 334, 335; Proboscidea, 339, 340, 341; Sirenia, 576 Quagga, 403, 404, 406, 409, 417 Quaternary Period, Acheulian industry, 225; Carnivora, 250; coexistence of hominid taxa, 223; dating techniques, 185; Deinotheriidae, 315; Equidae, 379, 403; extinctions, 612; Hippopotamidae, 489; Hominidae, 155-156, 158, 159, 162, 213, 222, 227, 229; paleoenvironment, 231-232; faunal evolution in, 608, 609, 612, 613; Proboscidea, 336; Sirenia, 574 Queen Elizabeth National Park, 9 Quercy, France, 253 Quercytherium, 251, 252 Rabaticeras, 257, 556; arambourgi, 556, 568; porrocornutus, 556 Rafting: monkeys to South America, 101; rodents to South America, 101 Rainfall in Africa, 4, 17-20 Ramapithecus, 126, 129, 147-152; adaptive grade and structural-functional zone, 217-218; and Australopithecus compared, 219; brevirostris, 147, 148, 149; description, 149-151; evolutionary relationships and classification, 152; hungaricus, 218;punjabicus, 127, 147, 218; wickeri, 127, 144, 148, 149, 151, 152, 218 Rangifer, 506 Rangwa, Kenya: Primates, 142; Dendropithecus, 140; Proconsul, 132, 134, 135, 136 Rangwapithecus, 127, 132, 137 Raphicerus, 545, 558-559, 568; melanotis, 11 Rat. See Rodentia. Cane, see Thryonomyidae; Thryonomys. Maned, see Lophiomys Ratel. See Mellivora Rawe Beds, Kenya, 35; Giraffidae, 519, 520; Hippopotamidae, 489; Suidae, 473, 474, 475 Recent Epoch, 609, 611, 613; Cetacea, 583, 592, 596, 597; Equidae, 388, 402, 407, 408, 409, 410, 412, 415; Giraffidae, 522, 529; Hyracoidea, 299, 308, 310, 312; Mammal fauna, 1-14; Proboscidea, 339, 354, 362; Sirenia, 573, 574; Suidae, 435-441; Tayassuidae, 476 Red Beds, Stromberg Series, Lesotho: Mesozoic mammals, 47; vertebrate fauna, 53 Red Sea: Equidae, 417, 418; geological development, 22

Index

Red Sea Rift, 17 Redunca, 550, 551, 569; arundinum, 550; darti, 550, 551, 569; fulvorufula, 550; redunca, 550 Reduncini, 8, 540, 550-552, 558, 565, 566-567, 568, 569 Reedbucks. See Reduncini. Bohor, see Redunca redunca Reptilia, 126, 227, 595, 604; PliocenePleistocene, 30; Triassic, 53 Rhagatherium, 427; aegypticum, 427 Rhinoceros, 286, 328, 371; pachygnathus, 286 "Rhinoceros" aurelianensis, 372 Rhinoceroses. See Rhinocerotidae. Black, see Diceros bicornis. White, see Ceratotherium simum Rhinocerotidae, 1, 2, 7, 26, 27, 30, 32, 35, 371-376, 402, 483, 606, 607; evolution, 376; and Hyracoidea compared, 285; Lothagam, 227; Lukeino, 227; Ngorora, 372; Rusinga, 142; Songhor, 142 Rhinolophidae, 65, 66, 67 Rhinolophus: capensis, 65; ferrumequinum millali, 65 Rhinopomatidae, 66, 67 Rhizomyidae, 81 Rhizopodidae, 608 Rhodesia: Carnivora, 254; Equidae, 406, 408; game farming, 14; Pliocene-Pleistocene, 40; wildlife, 12 Rhynchocyon, 4, 57, 58, 59, 60, 65; clarki, 57, 58; rusingae, 57, 58 Rhynchocyoninae, 58, 59, 65 Rhynchogale, 257; caniceps, 257; melleri, 257 Rhynchotherium, 343 Rhynchotragus, 558, 559; semiticus, 556 Ribbok. See Pelea Ribodon, 575 Rift Valley System, 17, 27; Hominidae, 154-155 Riojasaurus, 53 Riverton Formation, South Africa, 39 Rodentia, 1, 5, 6, 7, 10, 11, 27, 69-87, 126, 227, 277, 285, 286, 604, 609; distribution of recent faunas, 69-70; evolution, 76-87; Gebel Qatrani Formation, 25; number of genera and species, 2; rafting to South America, 101; sites and collections, 70; systematics, 70-76 Romans, and elephants, 12, 336 Rudolf. See Lake Turkana; East Turkana Ruminantia, 26, 536 Rupicapra, 541; rupicapra, 563 Rupricaprini, 541, 567 Ruscinomys, 83, 84 Rusinga, Kenya, 28; Anthracotheriidae, 428, 429, 431; Bovidae, 563; Carnivora, 253; Cercopithecidae, 110; Chalicotheriidae, 368; Chiroptera, 65, 66; Deinotheriidae, 321, 322, 326; Gelocidae, 536, 537; Hippopotamidae, 486; Hyracoidea, 289, 297, 299, 302, 308; Insectivora, 56, 57, 58, 60, 61, 63; Lorisidae, 91, 92, 93, 95; Palaeomerycidae, 498, 501; paleoenvironment, 453; Pongidae, 125, 126, 142, 143 (Dendropithecus, 140, 141; Limnopithecus, 138; Proconsul, 132, 134, 135, 136, 137); Proboscidea, 342, 343; Rhinocerotidae, 372, 373, 374, 376; Rodentia, 70, 79; and

Songhor compared, 142; Suidae, 442, 444, 445, 453; Tubulidentata, 271, 273 Russia: Bovidae, 567; Hyracoidea, 289, 307; Proboscidea, 342; Tubulidentata, 273 Ruwenzori Mountains, Uganda, 5, 6 Rytiodus 574, 577, 578; capgrandi, 577 Sabertooths. See Machairodontinae Sabi Game Reserve, South Africa, 12 Saccolaimus incognita, 65 Saghatheriidae, 287, 289, 290 Saghatheriinae, 290, 294-307, 308, 309, 310, 311, 312 Saghatherium, 294, 304-306, 311, 312; annectens, 288, 306; antiquum, 287, 293, 304, 305, 306; antiquus, 293; characteristics, 285; classification, 287, 289, 290; euryodon, 288, 306; and Geniohyus compared, 293; macrodon, 288, 306; magnum, 287, 293, 300, 305, 306; majus, 287, 291, 293, 295, 296, 305, 306; and Meroehyrax compared, 308; minus, 287, 305, 306; andPachyhyrax compared, 300; and Prohyrax compared, 309; sobrina, 288, 304, 306; and Thyrohyrax compared, 307 Saguinus, 128 Sahabi, Libya, 26; Anthracotheriidae, 424, 433, 486; Bovidae, 567, 568 (Bovinae, 547, 549; Hippotraginae, 552, 554); Cetacea, 584; Deinotheriidae, 323; Hippopotamidae, 486, 487; Proboscidea, 350; Rhinocerotidae, 371, 372; Suidae, 441, 455, 457 Sahara Desert, 9, 10, 17; barrier against north-south dispersion, 609, 611, 613; climate during Pleistocene, 37 Saiga, 567 Saint Arnaud, Algeria, 35; Equidae, 394; Proboscidea, 359 Saldanha Bay, South Africa: Cetacea, 585; Equidae, 406; Hominidae, 159 Sale, Morocco: Hominidae, 194, 197, 203 Sallamys, 78 Salonica, Greece: Giraffidae, 529; Tragulidae, 537 Samiri, 113 Samos, Iran: Bovidae, 562, 563, 565, 567; Hyracoidea, 268, 289; Rhinocerotidae, 374; Tubulidentata, 270, 273, 274 Samotherium, 517, 526, 528, 529; africanum, 517, 524; characteristics, 500, 505, 514, 516; classification, 509; and Giraffa compared, 520; and Helladotherium compared, 523; and Palaeotragus compared, 514, 516; sinense, 516 Sandelzhausen: Suidae, 453 Sanitheriinae, 442, 446-447, 449, 452,453 Sanitherium, 442, 446, 447, 449-450; cingulatum, 447; nadirum, 442, 447, 449, 452; schlagintweitii, 447 Sansanosmilus, 262 Saragata Deare, Ethiopia: Rhinocerotidae, 374 Saurischia, 54 Sauropoda, 54 Savanna, 4, 6 - 9 , 226, 227; biomass, 9; ecological separation, 8 - 9 ; game farming, 14; vegetation, 20, 231 Sbietla, Tunisia: Deinotheriidae, 321; Proboscidea, 345

Index

Schizochoerus, 445; vallesensis, 445 Schizodelphis, 591-592, 596; sulcatus, 591-592, 596 Schizotherium, 368, 369 Schurveberg, South Africa: Cercopithecidae, 104; Insectivora, 57 Sciuridae, 2, 5, 6, 7, 11, 78, 80, 81, 85; distribution, 69; North American, 82 Sciurognathy, definition, 70 Sciuroidea, 76; distribution, 69 Sciuromorph, definition, 70 Sciuromorpha, 76 "Sclater line," 8 Scoteinus, 65 Scotozous, 65 Scutisorex, 63 Sea cow. See Sirenia Seal. Elephant, see Mirounga leonina. Leopard, see Hydrurga. Mediterranean, see Monachus monachus. Monk, see Monachus pristiphoca Selenoportax, 547, 565; vexillarius, 546 Semliki, Congo, 34; Deinotheriidae, 321; Proboscidea, 342 Senegal, 22, 25, 334, 584, 589, 604 Serengeti Park, 9 Serra da Neve, Angola: Equidae, 403 Serridentinus, 341 Serval. See Felis Shanidar, Iraq: Homo sapiens, 210 Shansitherium, 517 Sharks, 37 Sheep. See Bovidae; Caprini Shrews. Elephant, see Macroscelididae. Otter, see Potamogalidae Shungura Formation, Ethiopia, 32-33, 34, 35, 36, 37; artifacts, 224; Bovidae, 541, 565, 569 (Alcelaphinae, 555, 557, 558; Antilopinae, 560, 561; Bovinae, 545, 548, 549; Hippotragus, 551, 552, 553, 554); Camelidae, 538; Equidae (Equus, 404, 405;Hipparion, 391, 394, 395, 399, 402); Giraffidae, 521; Hippopotamidae, 487, 489, 491-492; Hominidae, 172, 179, 186, 193, 223, 224 (Australopithecus, 219;Homo erectus, 200); paleoenvironment, 228, 230; Proboscidea, 353, 354, 355; Rhinocerotidae, 375; Rodentia, 81; Suidae, 456, 460, 463-465, 469, 472, 473, 475, 476 Siamang, 132, 133, 136, 140 Sicily: Hippopotamidae, 489 Sidi Abdallah: Rodentia, 82 Sidi Abderrahman, Morocco, 35; Hominidae, 197, 203, 205; Proboscidea, 353, 354, 356, 357 Sidi Bou Kouffa, Tunisia, 36; Bovidae, 549 Sidi Hakoma Member, Hadar Formation, Ethiopia: Hominidae, 164, 173 Simatherium, 566; kohllarseni, 549, 569 Simopithecus, 101, 102, 104-109, 112, 151, 219; jonathoni, 107 Sinai Peninsula: geological development, 22 Sinda, Congo: Carnivora, 260; Rhinocerotidae, 372 Singa, Sudan: Bovidae, 548; Homo sapiens, 215, 217 Sinoconodontidae, 46 Sinosaurus, 53

Sirenavus, 574 Sirenia, 25, 26, 27, 328, 333, 335, 337, 339, 573-578, 584, 604, 605; distribution, 573; fossil history, 573-578; origins, 312, 339, 573-574; phylogeny and evolution, 578; structure, 574-575 Sirte Basin, Libya, 25, 26 Sitatunga. See Tragelaphus Sivachoerus, 102, 441, 451, 456; giganteus, 453, 455, 456, 457; syrticus, 453, 455 Siuahyus, 447, 449-450;punjabicus, 450 Swapithecus, 127, 218; africanus, 125, 126, 134, 148; and Dryopithecus compared, 132; indicus, 135, 144; and Proconsul compared, 132 Sivatheriinae, 36, 182, 503, 522-527, 606 Sivatherium, 500, 505, 509, 523-529, 530; giganteum, 524, 525; hendeyi, 509, 526; maurusium, 509, 524-527; olduvaiense, 524 Sivatragus, 554; bohlini, 554; brevicornis, 554 Sivoreas eremita, 565 Sivoryx cautleyi, 554 Siwa, Egypt, 26: Cetacea, 591; Proboscidea, 342 Siwalik Series, India: Anthracotheriidae, 423, 430, 431, 433; Bovidae, 552, 565, 566, 567, 568; Camelidae, 538; Carnivora (Felidae, 262; Mustelidae, 256); Deinotheriidae, 315-316, 322; Giraffidae, 509, 517, 524, 529; Hippopotamidae, 489; Proboscidea, 355; Primates, 91, Ramapithecus, 147-148, 218; Rodentia, 71, 74, 79, 82, 83; Suidae, 446, 447, 449, 452, 456; Tayassuidae, 477; Tragulidae, 537; Tubulidentata, 273 Skates, 37 Skoonspruit, South Africa, 40 Smendou, Algeria: Giraffidae, 523; Proboscidea, 345, 359 Smilodectes, 129 Smutsia, 269, 270 Soblay: Hyracoidea, 288 Solenodontidae, 63 Soltanian: Hominidae, 159 Somali Arid Zone, 9, 10 Somali Coastal Zone: Equidae, 417, 418 Somalia: Sirenia, 25, 575 Songhor, Kenya, 28; Bovidae, 563; Chalicotheriidae, 368; Chiroptera, 65, 66; Gelocidae, 536; Hyracoidea, 297, 302; Insectivora, 57, 60, 63, 64; Lorisidae, 92, 93; Palaeomerycidae, 498; Primates, 127, 142 (Pendropithecus, 140, 141; Limnopithecus, 138; Proconsul, 133, 134, 135, 136, 137); Proboscidea, 342; Rhinocerotidae, 374; Rodentia, 70, 79; and Rusinga compared, 142; Tubulidentata, 273 "Sorex": dehmi, 61, 62; dehmi africanus, 61, 62; pusilliformis, 61 Soricella, 61; discrepans, 61 Soricidae, 4, 61-63, 64, 65, 67; Rusinga, 61-62; and Tenrecidae compared, 63 Soricinae, 61 South Africa, 23, 37-40; Bovidae, 541, 569 (Alcelaphinae, 555, 556, 557; Bovinae, 543; Hippotraginae, 551, 552); Carnivora, 249 (Canidae, 254; Felidae,

639

262; Hyaenidae, 260; Mustelidae, 255; Phocidae, 263; Viverridae, 257); Cercopithecidae, 101, 103, 104, 109; Cetacea, 584-585; correlations with East Africa, 40; Equidae (Equus, 403-410, 412; Hipparion, 388-390, 392, 393, 399, 401); game farming, 14; Giraffidae, 509 (Giraffinae, 518, 520, 522; Sivatheriinae, 524, 525); Hippopotamidae, 491; Hominidae, 159, 166-167, 169, 185, 196, 227; Hyracoidea, 284, 289, 310; Proboscidea, 353, 354, 356-357, 358; Rhinocerotidae, 375, 376; Rodentia, 75, 81-82; Suidae, 453, 462, 463, 471; Tayassuidae, 476-477 South America: Ateles, 141; Ceboidea, 101; early geological development, 22; Hyracoidea, 287; Pleistocene extinctions, 11; Proboscidea, 339; Rodentia, 86-87; Sirenia, 574, 578; Tayassuidae, 435, 476; Tubulidentata, 270 Southern Savanna Zone, 10 Southwest Africa: Carnivora, 255, 256, 257; Equidae, 403, 406, 408; Hyracoidea, 284, 288, 308, 309, 310, 312; Rodentia, 70, 77; Suidae, 441, 447, 448, 469 Southwest Arid Zone, 9, 10 Southwest Cape, 10-11; extinctions, 11; flora and fauna, 10-11; rainfall, 18 Spain: Deinotheriidae, 315; Equidae, 390, 394-395; Hippopotamidae, 486, 487, 489; Hyracoidea, 289; Primates, 144; Suidae, 445 Spirocerus, 567 Springbok. See Antidorcas Squalodontidae, 596 Squirrels. See Sciuridae. Flying, see Anomaluridae. Ground, see Xerus. Lemur, see Daubentonia Steenbok. See Raphicerus Stegodibelodon, 350, 360; schneideri, 350-351 Stegodon, 350, 352, 359, 364; kaisensis, 359; korotorensis, 352; trigonocephaly, 359 Stegodontidae, 336, 339, 349, 359-360, 364 Stegolophodon, 349, 357, 359, 364; sahabianus, 350, 360 Stegorhinidae, 587 Stegosaurs, 54 Stegotetrabelodon, 26, 349-350, 351, 360; lybicus, 350; orbus, 30, 349-350, 351; syrticus, 349, 350, 351, 360, 424 Stegotetrabelodontinae, 349, 350, 364 Stenella, 594 Stenoplesictinae, 257 Sterkfontein, South Africa, 37, 38, 224; Bovidae, 541, 553, 562; Carnivora, 260; Cercopithecidae, 102, 104; Hominidae, 166, 171, 172, 183, 184, 189, 191 (Australopithecus, 221-222, 228); Hyracoidea, 288, 311; Insectivora, 57, 60, 61, 64; paleoenvironment, 228, 229, 231; Rodentia, 70, 82 Sterkfontein Faunal Span, 38; Bovidae, 569 Sterrohippus robustus, 405 Steynspruit, South Africa: Bovidae, 555 Stochomys, 5 Stone Age: Late (Bovidae, 559; Homo

640

Stone Age, (Continued) sapiens, 214, 215, 216; artifacts, 39); Middle, 469 (Homo sapiens, 206, 214, 215; artifacts, 37, 39, 40) Stone artifacts, 32, 33, 222, 224-227; Kromdraai, 185; Sterkfontein, 38, 172, 191; Swartkrans, 189 Stormberg Series, Lesotho, 21; mammals, 47 Strepsiceros, 543 Strongulognathus, 503; sansaniensis, 502, 503, 564 Stylochoerus, 472, 475-476, 479; compactes, 469, 475-476 Stylohipparion, 37, 388, 390, 392, 393, 397; albertense, 399; hipkini, 399; libycum, 389, 390, 392-394, 397-402, 415; steytleri, 389-390, 399, 402 Sub-Saharan Africa, Miocene, 27-30 Subterranean mammals, fossil, 268, 270 Sudan: Camelidae, 538; Cercopithecidae, 108; Equidae, 404, 412; Giraffidae, 525; wildlife, 12 Sudanese Arid Zone, 9, 10 Suez Isthmus: Sirenia, 575 Suidae, 2, 7, 11, 26, 27, 30, 32, 36, 37, 159, 183, 227, 287, 290, 483, 592, 606; center of origin and migration, 452; environment and habitat, 477-479; geologic range and phylogeny, 477; living species, 435-441; lower-middle Miocene, 441-453; paleoecology, 452-453; phylogeny and classification, 448-452; upper Miocene-Pleistocene, 451, 453476 Suinae, 442, 447, 460-476, 607, 608; origin, 451-452 Suncus, 11, 61, 63, 65; chriseos, 62; distribution, 62; etruscus, 62, 63; fellowesgordoni, 62; infinitesimus, 61, 62; varilla, 61, 62, 63 Surdisorex, 62 Suregei Tuff Complex, East Turkana, 33 Suricata, 257; suricatta, 257 Sus, 436-440, 460, 477, 479; africanus, 467; algeriensis, 460; barbarus, 460; classification, 448, 449, 450; and Hylochoerus compared, 466; and Hyotherium compared, 443; andKolpochoerus compared, 462, 463, 465; limnetes, 463; and Nyanzachoerus compared, 455, 456, 457; origins, 451-452; phacochoeroides, 464; and Phacochoerus compared, 467; and Potamochoeroides compared, 471; and Potamochoerus compared, 461, 462; scrofa, 436-437, 440, 443, 460-463, 465, 471, 477; verrucosus, 443, 456, 471, 477; waylandi, 464 Swartklip, South Africa: Bovidae, 541, 552, 558, 559 Swartkrans, South Africa: artifacts, 225, 226; Bovidae, 541 (Alcelaphinae, 556; Antilopinae, 558, 559); Carnivora, 260; Cercopithecidae, 103, 104, 109; Equidae, 412; fossil accumulations, 37, 38, 181-182; Hominidae, 167, 181-182, 183, 186, 189 (Australopithecus, 219220, 222; Homo erectus, 200); Hyracoidea, 311; Insectivora, 61; paleoenvironment, 229, 230; Rodentia, 70; Suidae, 470, 473

Index

Swartkrans Faunal Span, 38 Swartlintjesfarm, South Africa: Rhinocerotidae, 375 Sylvicapra, 8; grimmia, 549, 550 Sylvisorex, 61, 62\granti, 61, 62; johnstoni, 62; megalura, 62 Synaptochoerus hieroglyphicus, 475 Syncerus, 6, 8, 9, 11, 548, 549, 566; caffer, 548; caffer caffer, 548, 556 Tabün, Israel: Homo sapiens, 210 Tachyoryctes, 80 Tachyoryctinae, 75 Tadla Beni Amir, Morocco, 27 Takin. See Ovibovini Talapoin, 129 Tanzania: Bovidae, 560; geological development, 22; Equidae, 389, 392, 399, 404, 408; Giraffidae, 518, 520, 521, 522, 525; Hominidae, 155; PliocenePleistocene, 30; Rhinocerotidae, 371; Rodentia, 75; wildlife, 12 Taphozous incognita, 65 Tapinochoerus, 467, 469, 473; meadowsi, 473; minutus, 470 Tapir, 328, 453 Tarsius, 114 Tatal Gol, Mongolia: Bovidae, 564 Tatera, 81; afra, 11 Tatrot Formation, Siwaliks: Bovidae, 548, 552, 566 Taung, South Africa: Bovidae, 549, 559; Cercopithecidae, 103, 104; geology, 3 7 38; Hominidae, 171, 172, 221; Hyracoidea, 288, 311; Insectivora, 57, 59, 61, 63; paleoenvironment, 227-228; Rodentia, 82 Taurotragus, 8, 9, 11, 543, 568; arkelli, 543, 568, 569; oryx, 543, 569 Tayassuidae, 433, 442, 448, 476-477 Taza, Morocco: Cervidae, 497 Tchadanthropus uxoris, 36 Tebessa, Algeria: Deinotheriidae, 321 Telanthropus capensis, 182, 189 Teleoceras, 372; aurelianense, 372; medicornutum, 372; medicornutus, 372; snowi, 371, 372 Temara, Morocco: Bovidae, 556; Hominidae, 210, 212-213 Tendaguru, Tanzania, 54, 604 Tenrecidae, 4, 11, 63-64, 65; Miocene, 61 Tenrecinae, 63 Tenrecoidea, 63 Tensiftian: Hominidae, 159, 197, 203 Teratodon, 251, 252 Teratodontidae, 251 Termites and ants, 275, 276, 277 Ternifine, Algeria, 36; Acheulian artifacts, 226; Bovidae, 543, 556; Camelidae, 538; Hominidae, 198-199; Proboscidea, 353; Rodentia, 70, 84; Suidae, 460, 476 Tertiary Period, 23, 609; Anthracotheriidae, 486; Bovidae, 547, 565, 566, 567, 569; Cercopithecidae, 101; Equidae, 379, 394; geological development, 23; Hippopotamidae, 487, 489, 491, 492; Hominidae, 121, 150; numbers of fossil genera, 603; Pongidae, 144, 227; Proboscidea, 336; Sirenia, 574, 577, 578; Suidae, 450 Testouromys, 83

Tethys Sea: Cetacea 582, 595, 596, 597598; geological development, 22-23 Tetracerus quadricornis, 546 Tetraconodon, 448, 451 Tetraconodontinae, 442, 452, 453; origin, 451 Tetralophodon, 339, 342, 344-345, 360, 364; longirostris, 344-345 Teutomanis, 270 Thalattosiren, 577, 578 Thaleroceros, 568; radiciformis, 551, 552 Thallomys, 7 Thamnomys, 7 Thecodontosaurus, 53 Thecodonts, 53 Themeda, 20 Therailurus, 260, 261-262; diastemata, 261-262

Theria, 54; relationships, 53 Theridomorpha, 84, 85 Theridomyidae, 76 Theridomyoidea, 80 Theridomys, 76 Theropithecus, 101, 102, 104-109, 112, 151, 219; brumpti, 107; darti, 106, 107; gelada, 106, 107, 108; oswaldi, 106, 107, 207 Thomas Quarry, Casablanca: Homo sapiens, 203-204 Thryonomyidae, 2, 71, 81, 85, 86; distribution, 69 Thryonomyoidea, 71, 76, 77, 78, 79-80 Thryonomys, 2, 13, 71, 77, 78, 79; classification, 69, 70; distribution, 69 Thyrohyrax, 306-307, 312; and Bunohyrax compared, 296; characteristics, 285, 286; classification, 289, 290, 294; domorictus, 289, 306, 307; and Meroehyrax compared, 308; and Ρachyhyrax compared, 302; and Prohyrax compared, 309; and Saghatherium compared, 305 Tiang. See Damaliscus lunatus Tierfontein, South Africa: Giraffidae, 526, 527 Tinderet, Uganda, 27, 28; Primates, 135, 136, 140, 142 Tit, Morocco: Cervidae, 497 Titanohyracidae, 288, 289, 290 Titanohyrax, 302-304, 312; andrewsi, 288, 302, 303-304; characteristics, 285; classification, 288, 289, 290, 294; and Megalohyrax compared, 298, 299; and Pachyhyrax compared, 300; palaeotherioides, 288, 303; schlössen, 288, 303; ultimus, 288, 302, 303, 304 Titanotheriidae, 289, 368 Titanotherium, 286, 288 Todenyang, Kenya: Giraffidae, 519 Topi. See Damaliscus lunatus Tortoises, 37 Tortonian: Suidae, 443, 452 Torynobelodon, 344 Tomolabis, 404 Tourism, 14 Toxodonta, 286 Tozeur, Tunisia: Proboscidea, 345 Tragelaphini, 8, 540, 543-546, 565, 566, 567, 569 Tragelaphus, 8, 543, 545, 559, 568, 569; angasi, 543, 545, 546; buxtoni, 10, 543,

Index

Tragelaphus, (Continued) 545; eurycerus, 543; gaudryi, 545; imberbis, 543-545, 565; nakuae, 545-546, 568, 569; pricei, 569; scriptus, 6, 543, 545, 569; spekei, 6, 8, 543, 545, 546; spekei stromeri, 545; strepsiceros, 8, 543-545; strepsiceros grandis, 543; strepsiceros maryanus, 544, 545 Tragocerus, 547 Tragoportax, 547, 565 Tragulidae, 1, 6, 536-537, 606 Tragulina, 536 Traguloidea, 564 Tragulus, 536, 537 Transvaal, South Africa, 37-38; Bovidae, 558; Carnivora, 249; cave breccias, 3738; Equidae, 406, 412; Giraffidae, 522; Hominidae, 159, 166, 171; Insectivora, 57, 59, 62, 65; Rodentia, 75; Suidae, 470 Triassic Period, 21, 22, 53, 604; Erythrotherium parringtoni, 46-53; Megazostrodon rudnerae, 46-53 Triceromeryx, 498 Trichechidae, 573, 578 Trichechus, 573, 574, 575, 578; dugon, 578; inunguis, 573; manatus, 573, 578; senegalensis, 573, 578 Triconodonta, 46; brain volume, 49-51; Morganucodontidae, 46-53; relationships, 52-53 Trilophodon, 341, 342; angustidens kisumuensis, 342; olisponensis pygmaeus, 343; pandionis, 334 Tritylodon, 53 Tritylodontidae, 51, 53 Tropical Forest Zone, 20 Tsaidamotherium, 567 Tsessebe. See Damaliscus Tubulidentata, 2, 57, 227, 270-277, 605 Tugen Hills, Kenya, 30; Tubulidentata, 273 Tuinplaas, Transvaal: Homo sapiens, 213 Tung Gur, Mongolia: Carnivora, 260 Tungurictis, 257 Tunisia, 22, 23, 25, 26, 36; Anthracotheriidae, 424, 429, 433; Bovidae, 541, 550; Carnivora, 249, 255; Cervidae, 496, 497; Equidae, 390, 399, 404, 408; Giraffidae, 516, 517, 522, 523, 525; Proboscidea, 358; Sirenia, 575 Tupaia, skeletal structure, 51 Tupaioidea, 56 Turicius, 359 Turkana, Kenya. See East Turkana; Ekora; Kanapoi; Lake Turkana; Lothagam Turkanatherium acutirostratus, 371, 373 Turkey: Deinotheriidae, 316; Giraffidae, 529; Hyracoidea, 289; Primates, 144; Ramapithecus, 149, 150-151; Tubulidentata, 273 Turolian: protohominids, 218; fauna, 158; Hominidae, 159 Turtles, 595; Miocene, 26, 27; Oligocene, 26 Tylopoda, 537 Typotheria, 286, 287; and Hyracoidea compared, 312 Uganda: Equidae, 388, 390, 392, 394, 399, 404; game farming, 14; Giraffidae,

518, 520; Hyracoidea, 308; Insectivora, 65; Lorisidae, 91; Miocene, 28, 112; Primates, 120, 127, 135, 142; Proboscidea, 351; Rodentia, 77, 79; wildlife, 12 Ugandax, 548, 565, 566; gautieri, 548 Uitkomst, South Africa: Hyracoidea, 288 Ukraine: Deinotherium, 324 . Uluguru Mountains, Tanzania: Insectivora, 62 Ungulates, 2, 8, 9, 11, 226, 277, 368, 376, 592; biomass, 6; game farming, 14; rinderpest panzootic, 12; suiform, 423 United States: Cetacea, 595, 596 Upper Congo, 6 Upper Fossil Wood Zone, Fayum, 294; Hyracoidea, 289, 294-298, 300-302, 304, 306, 307, 311; Primates, 124, 131, 132, 139; Proboscidea, 341 Uraha, Malawi: Hippopotamidae, 491 Urmiatherium, 567 Ursidae, 255, 263, 608, 614 Ursus, 255; arctos, 255 Usno Formation, Ethiopia, 32; Equidae, 390, 394; Hippopotamidae, 489, 491; Hominidae, 172, 224; Suidae, 460 Utilization of wild mammals, 13-14 Vaal-Cornelia Faunal Span, 38, 40 Vaal River, South Africa, 37, 38-40, 207; Bovidae, 550; Equidae, 409, 411; Proboscidea, 353, 356, 357; Suidae, 458, 462, 470, 473, 476 Val d'Arno, Italy: Equidae, 404, 405 Vallesian: Hominidae, 159 Vampyravus, 66, 67; orientalis, 65, 66 Varswater Formation, South Africa, 37; Cetacea, 584; Pholidota, 268; Tayassuidae, 477 Vegetation of Africa, 4, 17-20; communities, 231-232 Vespertilionidae, 1-2, 11, 65, 66, 67 Victoria Nyanza, Kenya: Hyracoidea, 288 Victoriapithecus, 110-111, 112, 113; leakeyi, 111; macinnesi, 110, 111, 112 Villafranchian: Bovidae, 560, 562, 563, 566, 569; Equidae, 389, 404, 405, 409; faunal assemblages, 35, 36, 199; Giraffidae, 524; Hippopotamidae, 487; Proboscidea, 364; Rhinocerotidae, 374; Suidae, 457, 460 Vindobonian: Bovidae, 563; Deinotheriidae, 315, 323; Rhinocerotidae, 372, 374; Suidae, 445; Tragulidae, 536 Vishnutherium, 529 Viverridae, 1, 2, 5, 6, 7, 10, 11, 256-258, 260, 263, 606 Viverrinae, 5, 11, 256, 257, 265 Vlakkraal, South Africa, 40; Bovidae, 551, 556, 557, 559 Vogel River Series, Tanzania: Proboscidea, 355-356 Vulpes, 7, 253, 254; chama, 10, 253;pallida, 253; ruppellii, 253; vulpes, 253 Wadi el Hammam. See Oued el Hammam Wadi Faregh, Egypt, 26; Cetacea, 591 Wadi Haifa, Sudan: Equidae, 404; Homo sapiens, 215, 217 Wadi Natrun, Egypt, 35; Bovidae, 567, 568 (Alcelaphinae, 558; Bovinae, 547; Hippotraginae, 552, 554); Carnivora,

641

255, 262; Cercopithecidae, 102, 108, 109; Giraffidae, 517, 525; Hippopotamidae, 487; Proboscidea, 345; Suidae, 441, 457 Wakondu, Kenya: Primates, 140 Walangania, 564, 606; africanus, 563; gracilis, 498, 563 Warthog. See Phacochoerus. Cape, see Phacochoerus aethiopicus Waterbuck. See Kobus Weasel. See Mustela West Africa: Bovidae, 548; Hippopotamidae, 492; Sirenia, 578; Suidae, 466 West Stephanie, Kenya: Deinotheriidae, 321 Whales. See Cetacea. Baleen, see Mysticeti. Beaked, see Mesoplodon. Toothed, see Odontoceti White Nile, Sudan: Suidae, 465 Wildebeest. See Connochaetes. Black, see Connochaetes gnou. Blue, see Connochaetes taurinus Williams Flat, South Africa: Primates, 138 Wolves. See Canis Wonderwerk Cave, Cape Province; Bovidae, 557; Equidae, 411 Xenarthra, 270, 276, 277 Xenochoerus, 442, 444, 446-450, 452, 453; africanus, 442, 447, 448, 449; leobensis, 447; robustus, 555 Xenogale, 257; microdon, 257 Xenopithecus koruensis, 124, 125, 132, 133, 134, 141 Xenorophus, 595 Xerus, 7, 10, 85 Yaleosaurus, 53 Yayo, Chad, 36; Hominidae, 199; Proboscidea, 354 Zambia: Equidae, 406, 408; game farming, 14; Giraffidae, 522, 524; wildlife, 12 Zarafa, 509, 512-515, 522, 528, 529; and Palaeotragus compared, 516; zelteni, 501, 513-515, 516, 527, 528 Zarhachis, 591, 596 Zebra. See Equus. Burchell's, see Equus (Hippotigris) burchelli. Grevy's, see Equus (Dolichohippus) grevyi. Mountain, see Equus (Hippotigris) zebra Zella Oasis, Libya, 26; Proboscidea, 340, 341 Zenkerella, 80 Zeuglodon, 587, 588, 591, 598; brachyspondylus, 591, 599; elliotsmithii, 590; intermedins, 588; isis, 590; 596; osiris, 588, 589, 590, 595; sensitivus, 590, zitteli, 589 Zeuglodontidae, 587, 588, 590 Zeuglodontinae, 588, 590 Zidania, Morocco: Proboscidea, 345 Ziphiidae, 584, 592, 596 Ziphius, 585 Zouerate, Mauritania: Proboscidea, 356 Zramys, 83, 84 Zululand: Bovidae, 548 Zygolophodon, 358, 359, 364; borsoni, 359; tapiroides, 359 Zygorhiza, 585, 587, 593, 595, 596; kochii, 591, 599