Proceedings of 1999 Beijing International Symposium on [PDF]

ISSN 1000-3193 CODEN REXUEM

ACTA ANTHROPOLOGICA SINICA Supplement to Volume 19

Proceedings of 1999 Beijing International Symposium on Paleoanthropology In Commemoration of the 70th Anniversary of the Discovery of the First Skull-cap of the Peking Man

Editorial Advisors WU Xinzhi ZHANG Senshui Editor DONG Wei

Published by Institute of Vertebrate Paleontology and Paleoanthropology Chinese Academy of Sciences

2000

BEIJING

Sponsored by United Nations Educational, Scientific and Cultural Organization National Natural Science Foundation of China

ACTA ANTHROPOLOGICA SINICA

(Quarterly, Started in Aug. 1982) 2000 Supplement to Vol.19

Preface This volume is an outcome of the International Symposium for Commemorating the 70th Anniversary of the Discovery of the First Skull-cap of Peking Man held at the Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, Beijing, China, in October, 1999 and was sponsored by this institute and the UNESCO. This volume was intended to include all of the presentations given at the symposium, but it is pity that some of the speakers on the meeting did not furnish us with their papers. The articles in this Proceedings are arranged in the order as follows: hominids, hominoids, stone artefacts, fire use, fauna, chronology and biography. On the commemoration of the discovery of the first skull-cap of Peking Man it is interesting to review the positions experienced by him in the research history of paleoanthropology. He held the position of the most important earliest human ancestor for about 30 years until 1959 when M. Leakeys found the Zinjanthropus and associated stone artefacts dated 1.75 mya. Not long later L. Leakey suggested that Homo habilis is the direct ancestor of H. sapiens so that Peking Man was excluded from the ancestors of modern humans. Since 1970s more and more paleoanthropologists became to believe that Peking Man represented an aberrant branch in human evolution on the basis of the application of cladistics in researching the phylogenetic position of Peking Man. Since 1987 DNA studies on the modern humans origin have also made people believing that Peking Man is not among our ancestors. The ancestral status of Peking Man was suggested by Weidenreich mainly on the basis of the similarities found between Peking Man and modern Mongolians, many of these similarities are found unreliable by other anthropologists in last decades. So the status of extinct branch of Peking Man has been accepted by many anthropologists currently and for long time. Opposite to these suggestions other anthropologists presented arguments for re-evaluating Peking Man's position in human evolution. After 1949 many human fossils have been unearthed in China, these made possible for Chinese anthropologists to investigate if there were convincing intermediate links between Peking Man and modern Mongolians. The results are invigorating. It has been found that Peking Man has a series of morphological features shared with most of other human fossils of different periods in China such as flat face, suture between frontal, nasal and maxilla bones forming a more or less horizontally curved line, low nasal saddle, quadrangular orbit, rounded shape of the infero-lateral portion of the orbital margin, more forward facing of the front-sphenoidal process of the zygomatic bone, curved lower margin of the cheek bones and the broadest part locating at the middle third of the cranial vault etc. The shovel shaped upper incisor is especially worthy to be mentioned because it presented in both Peking Man and all other Pleistocene human upper incisors found in China without exception. All these common features existed in Pleistocene humans of China with higher or much higher frequencies than those in other regions. The complex including these features occurred with much higher frequencies in China than in other regions. In addition to these there has been found morphological mosaic between Homo erectus and H. sapiens of China. Some so-called derived features of H. erectus such as thick brow ridges, thick cranial bones, strong post-orbital constriction and angular torus etc. are shown in H. sapiens specimens of China (thick brow ridge in Dali H. sapiens, thick cranial bone in Dali and Xujiayao, strong postorbital constriction. in Maba, angular torus in Dali and Ziyang H. sapiens); some features usually found in H. sapiens and not in H. erectus such as weak post-orbital constriction, high temporal squama, curved border of temporal squama and high cranial index could be seen in H. erectus skull from Hexian. The mosaic is also present in Yunxian skulls regardless of whether they should belong to H. erectus or H. sapiens. The common features of H. erectus and H. sapiens in China and the mosaic between these species indicate the continuity between them. According to the Hypothesis of Multiregional Evolution, gene flow could be well used to explain the unity of all populations of modern humans in one same species, H. sapiens. This explanation is much more convincible than the orthogenesis supposed by Weidenreich. Among the human fossils of China, there are evidence of gene flow such as the protruding of nasal saddle of Yunxian skull, surface bulging between pyriforme orifice and the orbit of Dali and Tangshan skulls, spherical orbit of Maba skull, bun-like structure on the occipital bone of Liujiang, Lijiang and Ziyang skulls and the more lateralward facing of the fronto-spherical process of zygomatic bone of Upper Cave skull No. 102. All of these i

features are seldom found in most of the Pleistocene skulls in China but exist more frequently in Europe. Peking Man as a fire user and successful fire keeper has been also challenged by some scholars in recent years. But all of these challenges have been proved as inconvincible. On the contrary, the laboratory data provided by the last challenger, Weiner and others on the percentages of the microfaunal and macrofaunal burned bones collected at this cave and examined by these authors are just good positive evidence because the percentages “are roughly similar to those obtained in much younger caves where the fire was undoubtedly used by humans.” Weiner et al. strongly suggested the laminated deposit containing the burnt bones as the negative evidence for in situ fire use, but the water transportation of the burnt bones within the cave in very short distance could much better explain the formation of this laminated sediment with burnt bones and the maintenance of the ratio of burnt bones of macro- and micro-fauna than remote transport from outside into the cave. Further more, the successive fission track datings carried out with the burnt deposits of this cave are also strong support to the in situ fire using in the cave. In sum, Peking Man remains the earliest most reliable fire user and one of the most important antecessors of H. sapiens in East Asia. But still there are many problems regarding to Peking Man well worth studying. I wish to thank all the participants who presented excellent articles at the symposium and joined in exciting discussions during the meeting. I also thank all the authors who provided their papers for enclosing into this volume. I am grateful to Professors ZHANG Senshui and DONG Wei who acted as academic advisor and editor of this volume respectively. Finally, thanks are due to Professor Qiu Zhuding, the former director and Professor Zhu Min, the director of IVPP for their generous support and wonderful assistance for the symposium and the publication of this volume respectively. Thanks are also due to the UNESCO in supporting the symposium and the publication of this volume.

WU Xinzhi 21 August, 2000, Beijing

ii

Préface C'est un grand honneur pour moi d'avoir été invité à ouvrir, aux côtés d'éminents collègues chinois, les actes du mémorable congrès international réuni en octobre 1999 à Beijing pour célébrer le 70eme anniversaire de la découverte du premier crâne de l'Homme de Pékin. Il convient de se souvenir de l'extravagante histoire qu'a été celle de la découverte de l'Homme de Pékin. Lorsque, dans les années 20, Johan Gunnar Andersson, prospecteur minier suédois, s'intéressa aux vertébrés fossiles chinois et fit venir un jeune post-doc de l'Université d'Uppsala, Otto Zdansky, pour rechercher de manière plus précise d'éventuels restes d'Hommes fossiles, la paléoanthropologie ne connaissait, pour raconter notre histoire, que l'Homme de Néandertal, l'Homme de Cro Magnon, l'Homme de Java et l'Homme de Piltdown qui s'avèrera n'être qu'une supercherie. Ce fut donc une découverte de taille de mettre au jour dès 1921, la première dent d'un autre Hominidé, le cinquième. Otto Zdansky en sera d'ailleurs si conscient qu'il le cachera à Andersson et attendit 1926 pour en faire publiquement état. Quelques autres dents isolées suffirent ensuite au paléontologue canadien Davidson Black pour nommer Sinathropus pekinensis ce nouveau maillon de notre phylogénie, le démarquant des autres, tout en le rapprochant en même temps de l'Homme de Java plus volontiers que de tout autre. Mais l'éclairage le plus brillant vint en 1929 lorsque le paléontologue chinois Pei Wenzhong mit au jour, lors d'une fouille conduite par un nouvel étudiant suédois d'Uppsala, Birger Bohlin, le premier crâne de cet Hominidé. Cette découverte eut évidemment le retentissement qu'elle méritait, puisqu'elle venait confirmer l'existence d'un très ancien Homme fossile en Chine et que pour la première fois elle avait les moyens d'en dessiner les grandes caractéristiques. Bien d'autres restes suivirent ainsi que d'abondants outillages lithiques et restes de faunes probablement consommés par ces Hominidés pour qui le site avait tenu lieu à beaucoup de reprises d'abri et de séjour. Lorsque Pierre Teilhard de Chardin, successeur de Davidson Black, mort prématurément, dut fermer le chantier de fouilles pour des raisons d'insécurité en 1937, le bilan pour un site paléoanthropologique, était impressionnant et sans précédents puisqu'on comptait 14 crânes, 11 mandibules, 147 dents et quelques os postcraniens. Grand site donc, immense bilan scientifique, puissant intérêt mondial (la Fondation Rockefeller sollicitée par Davidson Black, avait généreusement aidé les fouilles), large participation internationale durant les 17 premières années de cette recherche : le site de l'Homme de Pékin, certes préhistorique, était ainsi devenu historique, et ce notamment depuis cette découverte de Pei Wenzhong de 1929. Après une interruption d'un quart de siècle durant lequel la collection d'Hominidés des années 20 et 30 disparut, les recherches ont repris avec succès sous l'autorité de Woo Rukang. Zhoukoudian est ainsi demeuré un des 7 ou 8 sites paléoanthropologiques les plus fameux du monde ; il a été déclaré Site protégé par le State Council of China en 1971, inscrit sur la liste des sites du patrimoine mondial par l'UNESCO en 1987 et choisi, comme lieu de formation de la jeunesse de Beijing par la municipalité de cette très grande métropole en 1992. Au début de 1994, les collègues chinois se sont inquiétés devant la dégradation naturelle de la localité 1 de Zhoukoudian, fréquentée par plus de 200.000 visiteurs par an et ils ont demandé à l'UNESCO son aide pour l'entretien de ce "Monument". L'UNESCO répercuta alors cette demande, pour la collecte des fonds, sur une Société française privée Mondial Assistance qui créa pour la circonstance une Société "antenne" Assistance Ethno. L'UNESCO et son Directeur général d'alors, Federico Mayor, me demandèrent d'être, pour cette question, consultant de l'Institution tandis que Mondial Assistance et Assistance Ethno me prirent pour expert. Le 29 mars 1995, un accord tripartite pour la Rehabilitation, Protection and Conservation of the Peking Man World Heritage Site fut signé à Paris, à l'UNESCO, entre l'Académie chinoise des Sciences, l'United Nations Educationnal, Scientific and Cultural Organization, et l'Association Assistance Ethno - la délégation chinoise était conduite par Madame Hu Qi Heng, Vice-Présidente de l'Académie chinoise des Sciences. Le but de ce projet était en fait plus large qu'il n'avait été proposé à l'origine ; il concernait trois points : la restauration du site et sa conservation, le renouvellement de l'information scientifique du public, la reprise de la recherche scientifique. Un ITC de l'UNESCO (Comité technique international) fut établi ; il se réunit une première fois à Beijing en 1996 sous la présidence de Madame Chang Meeman et traita des 3 points proposés. Une première campagne de recherche eut lieu juste après, menée par une équipe française de iii

géologues et de géophysiciens d'EDF (Electricité de France). C'est donc un plaisir pour moi de saisir cette tribune pour faire connaître les premiers résultats de cette toute récente recherche : des mesures microgravimétriques, électromagnétiques(EM 38, EM 31, EM 34), magnétiques et électriques ont été réalisées sur l'ensemble du site ; ont été pris 639 points microgravimétriques sur profils, 143 points microgravimétriques EMG, 834 points électromagnétiques EM 31, 985 EM 38 en position horizontale, 985 EM 38 en position verticale, 100 points électromagnétiques EM 34, 176 points magnétiques et faits 8 sondages électriques. Ce travail sous l'autorité de Marc Albouy, Contrôleur général et responsable du Mécénat technologique et scientifique d'EDF, a été réalisé par Pierre Delétie, Jean-Paul Blais, Patrick Allombert, Jean-Pierre Baron, André Cocquart ingénieurs et par Qinqi Xu, Dong Wei et Tong Haowen, universitaires de l'Institut de Paléontologie des Vertébrés et de Paléoanthropologie de l'Académie des Sciences de Chine. Je retiendrai trois importants résultats : la colonne sédimentaire de la localité 1 s'est avérée puissante d'une cinquantaine de mètres au moins alors que quarante seulement ont été fouillées, mais ceci était connu depuis longtemps des collègues chinois ; l'entrée probable de la localité 1 s'est dessinée sous les escaliers d'accès au site et au-delà, dans cette même direction ; les localités 4 et 5, alignées le long de failles parallèles au grand axe de la faille principale de la localité 1, se révèlèrent plus riches qu'elles n'étaient apparues à la fouille ; enfin et surtout une localité nouvelle très développée, repérée pour la première fois, est apparue sur le flanc ouest de la colline de Zhoukoudian avec débouché possible sur le thalweg limitant ce flanc au Sud. Je souhaite évidemment vivement que cette première phase de prospection non invasive s'achève par une campagne de forages, confirmant l'existence (ou non) de ces karsts et informant sur l'existence et la nature de leurs contenus. La phase suivante appartient évidemment aux collègues chinois. Je voudrais saluer pour finir la haute tenue de ce rendez-vous scientifique international de 1999 à Beijing ; il a permis à la Chine de présenter à la communauté paléoanthropologique mondiale la très belle qualité de ses jeunes générations de paléontologues et préhistoirens, la longue liste des nouveaux sites de cet immense pays, l'incroyable ancienneté des plus vieux d'entre eux ; les collègues européens, nord-américains, sud-américains, africains et d'autres pays d'Asie , le Japon, la Vietnam, la Corée, la Thaïlande, l'Inde, sont venus débattre du statut des Hommes fossiles d'Asie de manière comparée, en leur appliquant parfois des méthodes nouvelles d'études telles que la reconstitution assistée par ordinateur en 3 dimensions , débattre aussi de leurs outillages, de leur environnement, de la taphonomie de leurs sites ; des travaux originaux, sans rapport obligé avec le monde asiatique, ont été évidemment présentés à cette occasion, des travaux théoriques - le rôle de la mousson dans le déploiement des Hommes, l'origine bipède des ancêtres communs des Hominidés et des Grands singes -, des travaux généraux - la flexion spheno-occipitale et les canaux semicirculaires des Hominidés successifs des descriptions de pièces fossiles originales (d'Allemagne par exemple, d'Espagne aussi) et des tentatives de synthèse sur le peuplement humain de certains continents, sur la maîtrise du feu etc. Mais ces Actes sont évidemment en eux-mêmes le meilleur témoignage de l'importance de la rencontre. Comment ne pas user de ce privilège de préfacier pour dire pour finir toute ma reconnaissance aux collègues chinois pour la chaleur et l'élégance de leur accueil, et à la République populaire de Chine, qui fêtait d'ailleurs son cinquantième anniversaire en même temps que se réunissait notre Congrès, pour son invitation ; pour dire aussi aux collègues chinois toute mon admiration pour la qualité et l'originalité de leurs travaux dont je suis toujours très curieux de connaître les résultats ; pour leur dire enfin qu'après être venu à Beijing et à Zhoukoudian déjà 4 fois (1995, 1996, 1998, 1999), j'ai grand espoir de leur rendre encore bien d'autres fois visite, par intérêt scientifique, mais aussi par plaisir et par affection.

Yves Coppens 29 September, 2000, Paris

iv

Contents Story of Monsoon — A New Environmental Interpretation of Origination of Hominid⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 1 LIU T-S, WANG Q

Human Evolution in the Last Million Years — The Atapuerca Evidence ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 8 ROSAS A

What Constitutes Homo erectus? ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 18 SCHWARTZ JH, TATTERSALL I

Review of the Phylogenetic Position of Chinese Homo erectus in Light of Midfacial Morphology ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 23 WANG Q, TOBIAS PV

Restoration of the Face of Javanese Homo erectus Sangiran 17 and Re-evaluation of Regional Continuity in Australasia ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 34 BABA H, AZIZ F, NARASAKI S

Two New Human Fossil Remains Discovered in Sangiran (Central Java, Indonesia) ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 41 GRIMAUD-HERVE D, WIDIANTO H, JACOB T

Finding of a Hominid Lower Central Incisor During the 1997 Excavation in Sangiran, Central Java ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 46 BABA H, AZIZ F, NARASAKI S et al.

Thickness Mapping of the Occipital Bone on CT-data — a New Approach Applied on OH 9 ⋅⋅⋅⋅⋅⋅ 52 WEBER GW, KIM J, NEUMAIER A et al.

The Period of Transition between Homo erectus and Homo sapiens in East and Southeast Asia: New Perspectives by the Way of Geometric Morphometrics ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 62 DETROIT F

Neural Tube, Spheno-occipital Flexion and Semi-circular Canals in Modern and Fossil Hominids ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 69 DAMBRICOURT MALASSE A, MARTIN JP, de KERVILER E

Enamel Microstructure of Lufengpithecus lufengensis ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 77 ZHAO L-X, LU Q-W, XU Q-H

Arboreal Primates and Origin of Diagonal gait ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 83 LI Y

Computer-Assisted Paleoanthropology: Methods, Techniques and Applications ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 90 ZOLLIKOFER CPE, PONCE de LE ÓN MS

Variability of Pliocene Lithic Productions in East Africa ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 98 ROCHE H

Greeting Chinese Paleolithic Archaeology in the 21th Century (A Retrospective) ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 104 HUANG W-W

Trends Peculiar to the Chinese Palaeolithic ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 115 OTTE M

A Use-Wear Study of Lithic Artifacts from Xiaochangliang and Hominid Activities in the Nihwean basin ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 119 SHEN C, CHEN C

Early Palaeolithic Occupation of Southwestern China and Adjacent Areas of Vietnam and Thailand ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 126 SCHEPARTZ LA, MILLER-ANTONIO S, BAKKEN DA

A Study of Lower Palaeolithic Cultural Remains from Pallahara in Eastern India ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 132 RAY R

v

Lower Palaeolithic Hunting Weapons from Schöningen, Germany –The Oldest Spears in the World– ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 140 THIEME H

Man in South America ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 148 BELTR ÃO MCMC, PEREZ RAR

Interpretation of Lithic Technology at Zhoukoudian Locality 15 ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 156 GAO X

Appearance of Early Blade Technique in Northeast Asia ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 166 MATSUFUJI K

Early Middle Palaeolithic Blade Technology in Southwestern Asia ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 170 MEIGNEN L

Blade Production During the Middle Paleolithic in Northwestern Europe ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 181 DELAGNES A

The First part of Upper Paleolithic in Western Europe: New Results on the Abri Pataud (Les Eyzies-de-Tayac, Dordogne) ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 189 NESPOULET R

Paleolithic Site Discovered at Dongfang Plaza, Beijing ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 194 LI C-R, FENG X-W, YU J-C

Fire Control by Homo erectus in East Africa and Asia ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 198 ROWLETT RM

Cave Occupation, Fire-making, Hominid/Carnivore Coevolution, and Middle Pleistocene Emergence of Home-base Settlement Systems ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 209 ROLLAND N

Evidence for the Use of Fire at Zhoukoudian ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 218 WEINER S, BAR-YOSEF O, GOLDBERG P et al. Large Mammalian Carnivores as a Taphonomic Factor in the Bone Accumulation at Zhoukoudian ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 224 BOAZ NT, CIOCHON RL, XU Q-Q et al.

A Preliminary Study on the Early Pleistocene Deposits and the Mammalian Fauna from the Renzi Cave, Fanchang, Anhui, China ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 235 JIN C-Z, DONG W, LIU J-Y et al.

A Comparative Analysis on the Mammalian Faunas Associated with Homo erectus in China⋅⋅⋅⋅⋅ 246 DONG W, JIN C-Z, XU Q-Q et al.

Quaternary Rhinoceros of China ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 257 TONG H-W, MOIGNE A-M

Immigration of Mammals into Japan during the Quaternary, with Comments on Land or Ice Bridge Formation Enabled Human Immigration ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 264 KAWAMURA Y, TARUNO H

Revised Paleomagnetic Age of the Nihewan Group at Xujiayao Palaeolithic Site ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 270 L ØVLIE R, SU P, FAN X-Z et al.

Chronological Studies on Chinese Middle-Late Pleistocene Hominid Sites, Actualities and Prospects ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 279 SHEN G-J, WANG J-Q

A Review of the Tephrochronological Studies of Paleolithic Cultures in Japan ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 285 SODA T, SUGIYMA S

The Scientific Influence that Dr. Davidson Black (Bu Dasheng) Had on Chinese Prehistory ⋅⋅⋅⋅⋅⋅ 292 CORMACK JL

Davidson Black and Raymond A. Dart: Asia-African Parallels in Palaeo-Anthropology ⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅⋅ 299 Tobias PV, Wang Q, Cormack JL vi

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

1-7

Story of Monsoon –-- A New Environmental Hypothesis of Origination of Hominid: A preliminary Account LIU Tunsheng1 , WANG Qian2, 3 (1. Institute of Geology, Chinese Academy of Sciences, Beijing 100029,China; 2. Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, Beijing,100044, China; 3. Sterkfontein Research Unit, Department of Anatomical Sciences, University of the Witwatersrand, Parktown 2193, Johannesburg, South Africa)

Abstract A new environmental hypothesis of origination of hominid, “Story of Monsoon”, is preliminarily introduced. The geochronological correlation of the key events during evolutions of the monsoon system and hominid discloses environmental forces behind hominid evolution and implies that the emergence of modern monsoon system in Miocene might have been triggered origination of hominid. The monsoon-dominated provinces are therefore theoretically sites cradling hominid. Miocene China and adjacent regions are important in research into development of monsoon system and origination of hominid.

Key words: Monsoon; Origination of hominid; Miocene China

1 Environmental background of hominid origination The origin of the human lineage is one of the most fascinating issues in the field of paleoanthropology. To date, the fossils of the accepted earliest hominid, the australopithecines, are only recovered in Africa. Thus, it is generally accepted that hominid originated from a kind of African ape. The timing is somewhere around 7 million years before the present on the basis of reconciliation between fossil record and molecular clock. But how did human originate? It has for over a century been a problem since Darwin’s time. It is self-evident that hominid appeared and evolved in the context of changing environment before human culture prevailed. The exact environmental background in which hominid originated has being concerned for long[1-4]. The research in environmental around the origination of hominid has been concentrated in Africa. Several hypotheses of the ecosystem have been put forward to interpret why and how human originated, such as “Savannah Hypothesis” (SH) and “East Side Story” (ESS)[1]. SH has been proposed for decades. It emphasizes the role of the Savannah playing in driving ape-hominid transformation. When forest gave way to the Savannah, apes had to go down to the ground and then became ground dwelling. Some apes began to habitually walk with only two hind limbs that marked the emergence of hominids. ESS then elaborates SH by singling out the exact site where origination of hominid happened. It is said that the Africa’s great barrier, the Rifts, which is responsible for the splitting of ape and hominid. To its west, where apes persisted in the surviving forest, while from the more arid open provinces of its east, hominids emerged. However, while there is an increasing number of discoveries of australopithecines and more detailed analyses of their associated fauna, the above-mentioned hypotheses are put in question. In Eastern and Southern Africa, the environments reconstructed for australopithecines when they lived are not the open Savannah, but woodland and forest[5]. Furthermore, a new species Australopithecus bahrelghazali was found in early 1990s in Chad, about 2500 kilometers west of the Rift. This surprising discovery expands the australopithecine territory from the southern and eastern Africa to middle Africa, westward across the Rift. Here again hominid lived with woodland and forest associated animals[6-7]. The significance of these advances suggest origination of hominid is not necessarily an Eastern African event or an African Savannah story, and negate the popular traditional hypotheses of ecosystem mentioned above. In fact, no regional climate system exists in isolation. It is part of a Biography: LIU Tunsheng, Professor at the Institute of Geology and Member of Chinese Academy of Sciences, specialised in Quaternary research.

ACTA ANTHROPOLOGICA SINICA

2

Supplement to Vol. 19, 2000

global integrated climatic system. Like any other animals and plants, humankind is one of the products of the lowest level within the Earth Ecosystem at a certain period. Therefore, it is better to address this issue from a globe perspective. When we look at the global environmental changes in Miocene, monsoon enters the scene. We have noticed that the Miocene great apes (the candidates of common ancestor of hominids and modern apes) and the early members of the first hominids (australopithecines) have been found exclusively in the modern monsoon-dominated regions (including large parts of Northeastern Africa and Asia, even the coastal areas along the Mediterranean in Europe. During geological periods, monsoons once covered further north than those do today as the results of northward drift of all continents except the Antarctic) (Figure 1), where the monsoon controls the natural environment by transporting vapor and heat This phenomenon triggers our interest in considering monsoon and hominid together.

Figure 1

Modern Monsoon dominated regions and distribution of Early Homo and Miocene Hominoid

2 Correlation of Key events in Time between Monsoon and Hominid From current knowledge of origins and evolutions of human and his ancestors, the following five splitting points are very necessary to lead to modern humans.

LIU et al.: Story of Monsoon -- A New Environmental Hypothesis of Origination of Hominid: A preliminary Account

1. 2.

3.

4. 5.

3

Emergence of high primates, or their splitting from the reminder of low primates. Current Chinese fossils evidence points to ~45 Ma B.P. in the late Middle Eocene [8]. Splitting of hominoids from one another or from an orangoid lineage. This event led to the emergence of the latest common ancestor of humans and modern great apes. The fossil evidence in Africa and Asia is over ~20 Ma B.P., but the molecular evidence supports a latter date. The timing of reconciliation is about 15 Ma B.P. Splitting of an (?African) hominoid lineage into the Hominidae and the lines leading respectively to Pan and Gorilla. This is without date the most critical event relating to humankind. From molecular date, supported tolerably by fossil evidence, this cladogenesis is set at somewhere between 9 to 5 million years, but the earliest fossil evidence (Australopithecus or australopithecine) observed in Africa so far in no more than 4.2 million years ago. Again, a compromising date, 7 Ma B.P. is accepted for this event. Emergence of Homo, splitting of Homo habilis from the rest of australopithecines. On present fossil and paleolithic evidence in Africa, this event seems to lie about 2.5 Ma B.P.. The emergence of modern humans (However it is a cladogenetic or non-cladogenetic event is not clear). Again, the date the molecular date, about 200 ka is earlier than that of fossil record observed, about 125 ka B.P. in Africa and Western Asia. The accepted date is 150ka.

A good way to detect the internal link between hominid and his environment is to correlate the events in evolutions of humankind and climate respectively in a temporal frame, which has been conducted by many scholars[3,9-10]. Following these examples, we correlate the key events in time of hominid and monsoon and surprisingly we find that some key events of both sides are almost contemporaneous. Based on the current temporal frame of hominid events, we are able to compare it with that of monsoon events. Time & Epoch (in Million years B.P.)

Key Events in Hominid Evolution

Major Events in Monsoon Evolution

0.15 - l .(M.Pleistocene)

Modern human originates

Winter Monsoon intensifies

2.5

Homo habilis emerges

Winter Monsoon intensifies

Hominid originates

Modern Monsoon configuration appears with emergence of Winter Monsoon

The latest common ancestor of modern apes and hominids appears

Monsoon system reorganized with disappearance of planetary circulation and intensification of Summer Monsoon

Higher Primates originate

Planetary circulation dominates

7 15

(L. Miocene) (m.M. Miocene)

Oligocene 45 (l.M. Eocene) ( Abbreviations:

Emergence of Summer Monsoons

l: late; m: middle; M: Middle )

Such a comparison discloses striking temporal coincidence in key events between hominid and monsoon. If both temporal frames based on current knowledge of the dating methods (both geochronological and molecular) and fossil records, are correct and the correlation in time is not pure coincidence, then based on the rough synchronicity, we are able to conclude that the process of hominidization might have been influenced or triggered by the monsoon. The hominid events since origination of hominid are exclusively linked to the intensification events of the Winter Monsoon, which may suggest the tempos of hominid evolution is more controlled by the worsening of environmental conditions. Besides, the Miocene events are worth of specially considering: the emergence of the latest common ancestor of modern apes and hominids, and the splitting of hominid lineage from ape’s, are contemporaneous with the beginning of the elements of modern monsoon system and the emergence of the configuration of modern monsoon system. On considering this phenomenon, we may be able

4

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

to conclude that emergence of modern monsoon system might have been triggered origination of hominid. Hominid is a natural result of the coupling of the environment dominated by monsoon and a Miocene great ape living in monsoon-dominated provinces. We thus build a new hypothesis of origination of hominid, “Story of Monsoon”. The formation of modern monsoon system has been very well studied[11-14]. The changes of land-sea distribution in the Miocene led to the transformation of atmospheric circulation and the emergence of modern monsoon system; monsoon then changed the environment with the heat and water it transported; and as a result, environmental forces triggered ape transformed to hominid. Origination of hominid thus is basically a result of global changes. However, as a middle-level production of the Earth Ecosystem (interaction among lithosphere, hydrosphere, biosphere and extraterrestrial energy), monsoon forms a gook link between humankind and global changes, therefore only in this sense, we state that hominid phenomenon in the Miocene is a monsoon event. Figure 2 shows the relations between humankind and monsoon.

Earth Ecosystem Monsoon Ecosystem within Monsoon Dominated Region Humankind Figure 2

Relation between Monsoon and Humankind

But now we are not able to make a statement that which kind of niche first hominid possessed. The concrete painting of this key event needs more detailed work and more discoveries. However, “Story of Monsoon” provides an alternative and probably a better macro-geographical foundation for advancing the study of Late Miocene environmental variation and origination of hominid. It encourages a much wider investigation of the ape-hominid transition, which has focused on Eastern Africa during all the second half of the recently past 20th Century. The forest and dense woodlands once formed a wide area connecting Africa, Europe and Asia[15-17], where Miocene hominoids or great apes widely lived, and were under the influence of the same monsoon system[18]. It would not be rash to suggest that each and every one of the monsoon-dominated provinces once wandered by great apes could cradle hominid. The present-day China has bearing on both monsoon and hominid evolutions, and deserves serious attention.

3 Miocene China’s role in research into evolutions of Monsoon and Hominid Various studies in geology, paleontology, paleoanthropology and paleoenvironment in Miocene China have together disclosed its important role in research into development of monsoon and origination of hominid. First, Miocene tectonic movements in China and adjacent regions played a critic role in shaping the modern monsoon system. The Himalayan and Tibetan Plateau uplift disturbed the heat transportation from the Equator to the Pole, so the Equator-Pole temperature gradient increased, the ice sheet then developed in the Arctic[11-14,19] , the later then directly invigorated the atmospheric circulation and initiated the Asian and Indian monsoon systems, and also influenced the African monsoon though changing the water temperature of the Atlantic Ocean[10]. Thus the tectonic movement of China and adjacent regions shaped the configuration of modern monsoon system. A comprehensive study of the uplift Qinghai-Tibetan Plateau can enable people to restore the scenario of onset and development of the modern monsoon system in a macro way. Secondly, the consecutive eolian deposits of the red clay and the Loess in China locked the valuable information of the past climatic development linked to the development of the monsoon.

LIU et al.: Story of Monsoon -- A New Environmental Hypothesis of Origination of Hominid: A preliminary Account

5

The study of the eolian deposits in China could reveal the development of the monsoon and concrete environments in a micro scale. Thirdly, there are many discoveries of fossils Miocene hominoids or great apes in China (in monsoon regions of course, Figure 3 ) and especially, Lufengpithecus in Yunnan Province, southern China. It has been proposed to be a possible common ancestor of modern apes and hominid by certain scholars[20]. Fourthly, present-day China underwent the same trend of environmental changes and showed deforestation conditions in ape habitats as Africa at that time. It may be one of the candidate provinces for ape-hominid transition.

Figure 3

The Miocene of China

4 The implication of “Story of Monsoon”: From Origination to Extinction Monsoons are of immerse importance to modern mankind too. It is estimated that over half of the earth’s population lives in the monsoon zones today. If there exist internal links between monsoon and humankind, even the mechanism behind this nexus is not very clear now, we should be alert at the human-induced alternation on modern climates. The knowledge of the past is our useful guide to what may befall us in the future. In view of the “Story of Monsoon”, a better understanding of present variations of monsoons and environments is imperative. If monsoon could

6

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

trigger hominid origins on the one hand; it may be able to exterminate hominids in certain cases on the other hand. The recent strong and devastating El Nino, has been triggered by human-induced greenhouse[21-22]. When we disturb nature enough, as a result the monsoon configuration may be transformed. What kind of evolutionary scenario shall we be confronting then?

5 Conclusions How to vindicate and then elaborate “Story of Monsoon” still awaits further studies and discussions. We can draw some preliminary conclusions at the present: 1. The origination of hominid is probably a monsoon event, triggered by the emergence of the modern monsoon system in the Late Miocene. 2. Hominid might have originated in any of monsoon-dominated provinces where great apes once lived. 3. Miocene China has important bearings on research into development of monsoon and origination of hominid. We hope that this preliminary discussion on the environmental forces behind origination of hominid could help enhance our awareness of the nexus between humankind and nature, and to strengthen the multidisciplinary research in the past, present and future of humankind. Acknowledgements: We are grateful to Professor WU Xin-zhi for reviewing this article, and also to Mr. SHEN Wen-long for preparing the figures. The second author (QW) sincerely thanks the Fund of Special Subject of the Natural Scientific Foundation of China and the National Research Foundation (NRF) of South Africa for financial support. References: [1] COPPENS Y. Le singe,l’Afrique et l’Homme [M]. Paris: Rayard Press, 1983. [2] COPPENS Y, HOWELL FC, ISSAC GL et al. eds. Earliest Man and Environments in the Lake Rudolf Basin [M]. University of Chicago Press, 1984. [3] VRBA ES. Early hominids in southern Africa: updated observations on chronological back ground [A]. Hominid Evolution: Past, Present and Future. Alan R. Liss. 1985, 195-200. [4] TOBIAS PV. The environmental background of homonid emergence and the appearance of the genus Homo [J]. Hum Evol, 1991, 6:129-142. [5] WOLDEGABRIEL G, WHITE TD, SUWA G et al. Ecological and temporal placement of early Pliocene hominid at Aramis, Ethiopia [J]. Nature, 1994, 371:330-333. [6] BRUNET M, BEAUVILLAIN A, COPPENS Y et al. The first australopithecine 2,500 kilometres west of the Rift valley (Chad) [J]. Nature, 1995, 378:273-274. [7] BRUNET M, BEAUVILAIN A, COPPENS Y et al. Australopithecus bahrelghazali, une nouvelle espece d’Hominide anciene de la region de Koro Toro (Tchad) [J]. C.R. Acad. Sci. Paris, 1996, 322:907-913. [8] BEARD KC, TONG Y, DAWSON MR et al. Earliest complete dentition of an Anthropoid primate from the late Middle Eocene of Shanxi Province, China [J]. Science, 1996, 272:82-85. [9] TOBIAS PV. Ten climatic events in hominid evolution [J]. S Afr J Sci, 1985, 81:271-272. [10] DEMENOCAL P. Plio-Pleistocene Africa climate [J]. Science, 1995, 270:53-59. [11] LIU T, ZHENG M, GUO Z. The origin and development of Asian monsoon system and its temporal coupling with the ice sheet in two Poles and regional tectonic movements [J]. Quat Res, 1998, 3:194-204. [12] LU Y, DING G. A few Asian paleo-monsoon associated problems of Cenozoic tectonic development in China and adjacent regions [J]. Quat Res, 1998, 3:205-212 [13] WANG P. Asian deformation and global cooling – a research into the relation between climate and tectonics [J]. Quat Res, 1998, 3:213-221. [14] SHI Y, TANG M, MA Y. Discussion of relation between the second-step Qinghai-Tibetan uplift and conception of Asian monsoons [J]. Chinese Science Series D, 1998, 28:263-271. [15] KURTEN B. The Chinese Hipparion fauna [J]. Soc Sci Fennica Comment Biol, 1951, 8:1-34.

LIU et al.: Story of Monsoon -- A New Environmental Hypothesis of Origination of Hominid: A preliminary Account

7

[16] BERNOR RL. Geochronology and zoogeographic relationships of Miocene Hominidae [A]. New Interpretation of Ape and Human Ancestry. New York: Plenum Press, 1983, 21-66. [17] QIU Z. Tongur Micromammal Fauna in Middle Miocene in Innermongolia [M]. Beijing: Science Press, 1996. [18] LIU T, DING Z. The similarities of environmental changes in monsoon-dominated provinces and their bearings on hominid evolution [J]. Quat Res, 1999, 3:289-298. [19] SAVIN SM, DOUGLAS RG, STEHLI FG. Tertiary marine paleotemperature [J]. Geol Soc Am Bull, 1975, 86:14991510. [20] WU R, XU Q, LU Q. The relationship between Lufeng Sivapithecus and Ramapithecus and their phylogenetic positions [J]. Acta Anthropol Sin, 1986, 5:1-30. [21] MEEHL GA, WASHINGTON WM. El Nino-like climatic change in a model with increased atmospheric CO2 concentration [J]. Nature, 1996, 382:56-60. [22] TIMMERMANN A, OBERHUBER J, BACHER M et al. Increased El Nino frequency in a climate model forced by future greenhouse warming [J]. Nature, 1999, 398:694-697.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

8-17

Human Evolution in the Last Million Years — The Atapuerca Evidence Antonio ROSAS (Departamento de Paleobiología, Museo Nacional de Ciencias Naturales, c/ José G. Abascal 2, 28006 Madrid, Spain)

Abstract The contribution of the human fossil remains recovered at the Sierra de Atapuerca sites (Spain) to the model of human evolution during the last million years is explored. The Atapuerca Research Team (ART) have been studying these fossils over 20 years and their results have helped to discern a clearer picture of the processes and events in human evolution. One of the most relevant contributions of the ART to paleoanthropology has been the discovery of human remains from the Lower Pleistocene, and the proposal of the new species Homo antecessor as the last common ancestor to Neanderthals and modern humans (H. sapiens). We consider that the species H. antecessor emerged in Africa as a descendent of H. ergaster populations, about 1 my ago. According to our model, once this new species became differentiated, its most significant feature being the acquisition of a modern human middle face, some of its populations emigrated out of Africa and reached Europe. Other Homo antecessor populations remained living in Africa where, following evolutionary processes still badly known, they gave rise to the species H. sapiens; perhaps by means of an intermediate species. In Europe, however, populations of H. antecessor suffered divergent evolutionary processes giving rise to the human populations inhabiting Europe during the Middle Pleistocene; a species known under the name H. heidelbergensis. These populations maintained their specialisation processes, and gave rise to the Neanderthals. Therefore, Homo antecessor is located at a key point of the hominid evolution, at the divergence of H. sapiens and Neanderthals. There is still a long way in the study and comparison of the Atapuerca remains with other fossils from Africa and Asia in order to contrast the proposed evolutionary scenario and to define the processes of evolution that, from H antecessor, originated the modern human forms.

Key words:

Atapuerca; Homo antecessor; Human evolution; Lower Pleistocene; Middle Pleistocene

1

Introduction

The Atapuerca Hills (Burgos, Spain) contain a system of karstic cavities with rich archaeological and paleontological record. The sedimentary infillings range in time from Lower Pleistocene (older than 1.2 my) to present time. The South bank of the Atapuerca Hills was sectioned during the engineering of a railway trench, and several cavities and sedimentary sequences become exposed. Three of the different sites of the Atapuerca karst system have given Pleistocene human remains: Gran Dolina (TD-6 level), Sima de los Huesos and Galería. The age of the human fossils ranges from the Lower Pleistocene 0.8 my, in the case of the hypodigm coming from Gran Dolina, to Middle Pleistocene 0.3 my, in the case of the fossils recovered at the Sima de los Huesos site. In this work the way the Atapuerca human remains contribute to the model of human evolution in the last million years is discussed. One of the most relevant results has been the proposal of a new species, Homo antecessor, to accommodate the variability detected in the Gran Dolina fossils [1]. The discovery of human remains in the level 6 (TD-6) in the mentioned Gran Dolina site [2] has conducted to a research program for the re-evaluation of the late Lower and Middle Pleistocene paleoanthropological record [3-4]. In addition, the unique sample of human remains recovered from the Sima de los Huesos (SH) site[5-7] offers a solid basis of comparison for exploring the complex biological variation detected in the European Pleistocene human populations.

2 Evolutionary context and the ongoing debate One of the most enriching debates in paleoanthropology during the last two decades has been the discussion of the origin of the modern human species. There are a number of implications in this debate because the elucidation of the tempo and mode of the H. sapiens origins, even though being of great

ROSAS: Human Evolution in the Last Million Years — The Atapuerca Evidence

9

interest, goes beyond the study of a single species. Its resolution depends on the understanding of the evolution of the genus Homo in, at least, the last million years. In this sense, the human fossils from Europe, and especially those from Atapuerca, represent a key element in the clarification of the problem. The above mentioned debate has been focused on two models: the multiregional model and the “out of Africa” model, widely discussed in the literature [8-12]. The multiregional model supports the evolutionary continuity of human populations during the last million years, which is resolved with geographic differentiation of the living human groups during the Middle Pleistocene [8]. According to this interpretation of the fossil record, the H. sapiens species emerged as an anagenetic differentiation of the ancestor species H. erectus, considered of a wide geographic distribution. On the contrary, the model of the single origin maintains that the H. sapiens was differentiated in Africa, by means of an event of cladogenesis in a relatively recent time (no more than 300.000 years ago). From somewhere in Africa they would have colonised the remaining areas of the planet. A direct consequence of this hypothesis is that hominids from the Middle Pleistocene from Asia and Europe would become extinct without descendants. In this sense, the meaning of the Neanderthals from Europe has been and still is vital for clarifying in what way and under what processes human evolution is resolved (Figure 1). Therefore, the questions about the phylogenetic relationships between H. sapiens and Neanderthals, their degree of kinship and what has been its last common ancestor are important aspects for the explanation of the human evolutionary model.

Figure 1

The outstanding role that the European hominids play at the time of clarifying the origin of our species is recorded

ACTA ANTHROPOLOGICA SINICA

10

3

Supplement to Vol. 19, 2000

The European lineage. The fossils from the Sima de los Huesos site

From a paleontological point of view, defenders of a single and recent origin of H. sapiens have developed an evolutionary scenario in which Neanderthals and modern humans share a common ancestor, represented by the species H. heidelbergensis [12-13]. These authors, after the analysis of the Lower and Middle Pleistocene human remains, have concluded that at some time, at least 0.6 my ago, a speciation event that modified the primitive H. erectus and gave rise to a new species intermediate between H. erectus and H. sapiens took place. For some years, these intermediate populations were called archaic H. sapiens [14], but because of the exigencies of the International code of zoological nomenclature this intermediate human group was denominated H. heidelbergensis. Without going into technical details, the oldest linnean name given to a fossil remain attributed to this new species corresponds to the mandible from Mauer (Heidelberg, Germany), that was named H. heidelbergensis [15]. The most conspicuous change appreciated in the transit H. erectus - H. heidelbergensis concerns the increment of the cephalic volume as well as a series of morphological details that, in one way or another, are related to this cephalic increment. According to this model, H. heidelbergensis would have been originated in Africa by means of a genetic bottle neck [13]. From this hypothetical origin, the new species dispersed and colonised Europe, reaching an Afro-European distribution. The remains from Bodo (Ethiopia) and Kabwe (Broken Hill) (Zambia) are the most representatives from Africa, while the skull from Petralona has represented the H. heidelbergensis from Europe. There exist some remains in Asia, very specially the fossil skull from Dali, found in the Shaanxi province (China), with an estimated age of 0.3 my, whose advanced features have provoked discussion [16]. Authors like Rightmire [13] have proposed that Dali could be an Asian member of H. heidelbergensis. Nevertheless, other more detailed studies have shown that this cranium presents a morphology close to H. erectus, in spite of its size increment [17]. The scenario above summarised presents, however, some unsolved questions. A major problem to the model of H. heidelbergensis is the presence of Neanderthal features in, at least, some of the European Middle Pleistocene human remains, between 0.5 and 0.2 my. The European fossils of this period exhibit an unusual morphological diversity, in such a way that some of them show a large similarity with the Neanderthals whereas others display a more undifferentiated morphology. This pattern of variation has given rise to a variety of interpretations, among which the pre-Neanderthals and pre-sapiens theory should be noted [18]. This state of the art has begun to be solved in a large degree due to the discovery of a large collection of fossil remains in the Sima de los Huesos site at Atapuerca.

Figure 2

Three mandibular specimens from the Sima de los Huesos site, at Atapuerca, with an age of 0.3my. The Atapuerca-SH sample has allowed to ascertain the pattern of variability of the European Middle Pleistocene populations

The unique sample of human remains recovered at the Sima de los Huesos site is clarifying the pattern of variation of the populations of this period, revealing that some individuals express the Neanderthal features more clearly than others (Figure 2). On this basis, a more detailed study shows that all the specimens from the European Middle Pleistocene have already developed derived Neanderthal characteristics. Our conclusion is that the ancestors of Neanderthals were living in Europe at least 0.5

ROSAS: Human Evolution in the Last Million Years — The Atapuerca Evidence

11

my ago. Whatever the process of change giving rise to the Neanderthals this had already begun at least half a million years ago, being well represented in the Mauer mandible [3]. Therefore, the name H. heidelbergensis should be used only for defining the direct ancestors of the Neanderthals, and should not be applied for the Middle Pleistocene human populations from Africa and Asia.

4

The Gran Dolina site and the human fossils from the TD-6 level

In the context of the Atapuerca karst system, the Gran Dolina site is outstanding because of its chronological amplitude, with a long sedimentary sequence of 20 m, rich in archaeological and paleontological record. The upper part of the sequence is of a Middle Pleistocene age [2], while the inferior half dates from the late Lower Pleistocene [19]. The human remains of the species H. antecessor come from the lower half of the sequence[1-2, 20]. The lower levels of Gran Dolina (TD4 and TD5) were excavated in the nineties in a small area, and large mammals were recovered, e.g. equids, rhinoceros and cervids. The TD4 and TD5 levels also had abundant rodent remains, including the species Mimomys savini, of special interest in biochronology. Its presence allows dating the lower part of the Gran Dolina sequence as older than 0.5 my. In addition, as a result of this limited excavation, four stone tools made of quartzite were recovered[21], indicating human occupation in the Atapuerca Hills, and therefore in Europe, in an age older than previously thought. The oldest European hominid, until the discovery of human remains in the Gran Dolina site, had been represented by the Mauer mandible (Germany), whose antiquity was estimated in 0.5 my old. Based on the data known at that date, Roebroeks and kolfschoten [22] published a theory supporting the idea that Europe was depopulated in the lower part of the Middle Pleistocene. In their opinion, the archaeological record did not have enough information to support the hypothesis of human occupation in Europe before 0.5 my. In order to define better their theory, these authors established a biochronological reference. They maintained the genus Homo never coexisted in Europe with the vole Mimomys savini. The earliest human populations would have arrived to Europe after Mimomys savini evolved into Arvicola cantiana. These conclusions were not coincident with the recently published study by Carbonell y Rodríguez [21] regarding the stone tools discovered in the TD-4 level at Gran Dolina. These stone tools, though of a rudimentary appearance, were clearly of human manufacture, and they were associated to a faunal assemblage including Mimomys savini. The theory of a “young Europe”, that is, a first colonisation younger than 0.5 my, did not recognise this evidence. Motivated by this circumstance, a stratigraphic survey of approximately 6 m2 was undertaken in 1994 in the Gran Dolina site in order to look for more evidence supporting an “older Europe” model. To be right in our hypothesis, the results of this survey should offer more definite proof of human occupation in Europe in sediments of the beginning of Middle Pleistocene. In addition, this survey would provide a preliminary documentation of the site before starting extensive digging. Just before the beginning of the survey, experts in paleomagnetism had sampled the stratigraphic sequence of Gran Dolina. Paleomagnetism is a dating technique that relates the age of the sediments with changes of the magnetic polarity occurring in the Earth’s history. A first surprise occurred. At the level 7 of Gran Dolina, a change in the magnetic orientation of the sediments was identified. According to all the indications, change of polarity should correspond to the Matuyama/Brunhes reversal. This meant that the Gran Dolina stratigraphic levels from 1 to 7 are older than 780.000 years [19]. Under these promising antecedents we stated the survey in Gran Dolina. On the sixth of July 1996, the team found, approximately 1 meter below the Matuyama/Brunhes reversal, the first human remains associated to mammal fossils and stone tools. Therefore, the hominids were living in Europe more than 780.000 years ago [2], and furthermore, their archaeological assemblage had the extra value of being located within a long and fertile stratigraphic sequence. This discovery was going to change our picture of the human evolution. For some years, the survey in Gran Dolina has been going on, and has produced results well above the expectancies. It has been possible to delineate which of the levels in the sequence preserve M. savini

ACTA ANTHROPOLOGICA SINICA

12

Supplement to Vol. 19, 2000

and where the evolutionary transition to the derived species A. cantiana is located[23]. The mentioned transition is located in the TD-8 level, meaning that below this layer the age is older than 0.5 my. All these data ratify the evidence of human occupation in the TD-4 level, whose age is nowadays beyond all doubt.

5 The new species: Homo antecessor At present, 85 specimens compose the sample of human remains found at the TD-6 level. This hypodigm includes several fragments of the neurocranium, among which a nearly complete frontal bone is included, 14 isolated dental pieces, a fragment of mandible with M1 and M3 in situ (Figure 3), as well as several post-craneal remains. In 1995 an important fragment of facial skeleton was found. The specimen preserves part of the malar bones and several dental pieces in different stages of development that infer that the individual died at an age of 11 years. The complete set of fossils belongs to a minimum number of six individuals: two adults, possibly females, two individuals that died near to an adolescent age, and two others that died at an immature age of 3-4 years [1].

Figure 3

Essay of composition of the Homo antecessor face, made by a fragment of frontal bone and a reasonably well preserved portion of the middle face (specimen ATD6-69)

The human remains recovered at the Gran Dolina site (Atapuerca) present, among hominids, a unique combination of anatomical characteristics (Figure 4). Based on the human remains recovered at the TD-6 of Gran Dolina site, a new species has been proposed to accommodate the singular variability detected in the sample. The pattern of cranial, dental and mandibular features suggest these hominids belonged to a new species of Homo that was denominated Homo antecessor [1, 24]. The name «antecessor» comes from the Latin word meaning explorer, the pioneer, alluding that the human populations whose remains have been found in Atapuerca represent the first human populations arriving into Europe. The study of these remains has revealed that the anatomy of the first settlers of Europe had a peculiar combination of features. The cranial capacity of these hominids has been estimated over 1000cc, and their skeleton was of a gracile complexion, conversely to the great robusticity of the later

ROSAS: Human Evolution in the Last Million Years — The Atapuerca Evidence

13

hominids that inhabited Europe during the Middle Pleistocene (between 780.000 and 120.000 years). A double-arched browbridge in the Gran Dolina hominids -clearly different from the continuos torus of H. erectus- is to be noted. Furthermore, the dentition displays a primitive condition in H. antecessor, similar in several details to that of Homo ergaster, a hominid species living in East Africa between 1.8 and 1.4 my ago. At the same time, the anterior mandibular teeth are slightly enlarged, like in the European Middle Pleistocene hominids. The mandible is also similar in some features to that of the European populations. But, perhaps, the most surprising feature of H. antecessor is its facial architecture, with a similar configuration to that of Homo sapiens. In our species, the surface of the infraorbital plate has a backward inclination, giving rise to a depression called canine fossa; easily identifiable as a concavity below the cheekbone. The fragment of a malar bone, ATD6-58, belonging to an adult individual also presents a sketch of canine fossa. This circumstance excludes the possibility that the typically modern human middle face of H. antecessor is only an immature feature [25].

Figure 4

The combination of features detected in the human remains from the Lower Pleistocene of the TD-6 level, Gran Dolina site, is shown. H. antecessor, in the middle, shares primitive features with H. ergaster, from the Lower Pleistocene of Africa, as well as typical features of H. sapiens; most especially, the topology of the middle face

This singular combination of features is not found in any other hominid species previously known. In our work, the first step was to test whether the TD-6 hominids could be classified with the European Middle Pleistocene (Mauer, Arago or Sima de los Huesos, this last site also in Atapuerca). We confirmed that these European fossils already show features typical of the Neanderthals, while the TD-6 hominids do not display these sort of features. In our analysis, we also considered the possibility of classifying the TD-6 hominids in H. erectus. However, this latter species has a set of derived traits [26-27] that were not present in the Atapuerca hominids. Could the TD-6 hominids be primitive members of our own species Homo sapiens? The middle face seemed to show a close similarity; but many other features were too primitive. In view of these data, the adopted solution was to define a new species: Homo antecessor.

6

The middle face

The morphology of the middle face is perhaps the most surprising feature found in H. antecessor. The oldest remain showing this morphology before the discovery of the Gran Dolina hominids were the fossils from Djebel Irouhd (Morocco) [28], and Skuhl y Qafzeh, in the Near East, all of them of an age inferior to 0.2 my (late Middle Pleistocene/ early Upper Pleistocene).

14

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

The more singular features defining the architecture of the modern human middle face are the maxillary flexion and the presence of a canine fossa. In order to understand these features we should consider the configuration of the coronal planes of the face. These planes are defined as follows. On the one hand, the infraorbital plane is defined by the surface of the infraorbital plates; that is, the portion of the maxillary and malar bones located below the inferior margin of the orbit. On the other hand, the pyriform plane is defined by the virtual surface of the nasal aperture [25]. In the primitive forms of the genus Homo, both coronal planes of the face are located approximately at the same level, determining a flat facial topography. On the contrary, in the derived forms, such as Neanderthals and H. sapiens, a separation of the coronal planes is found, in such a way that the pyriform plane is located more anteriorly than the infraorbital plane, producing the development of a true nose. H. antecessor is the first species in the hominid evolution that shows a face with a marked relief and a real nose. According to our hypothesis, a divergent evolution takes place from the derived morphology detected in H. antecessor. In the Neanderthals, with the development of a marked mid-facial prognatism, the infraorbital plane undertakes a secondary bulking, as a consequence of the increment of the maxillary sinus located behind the bony wall forming the infraorbital plane. The result is a face where relieves become smoothed. In the evolution of H. sapiens the morphology detected in H. antecessor is maintained. That is, a canine fossa is maintained or increased, representing a depression located on the wall forming the infraorbital plate[29].

7 A new model and pending questions The association of features found in the human fossils from TD-6 lead us to pose the hypothesis that H. antecessor is the last common ancestor to H. sapiens and Neanderthals (Figure 5). In our opinion, H. antecessor originated in Africa 1 my ago, as a descendant of H. ergaster. The presence of a great number of primitive traits in H. antecessor supports this postulate. Once the new species was differentiated in Africa, populations of H. antecessor would have left this continent towards other regions of the planet. At present, we know the presence of H. antecessor in Southern Europe, though it is possible that this species reached some regions of Asia. The Nanjing skulls, dated to 0.4 my, in which a canine fossa can be appreciated, could be potential Asian members of H. antecessor. Recently, Wang[16] and Aguirre (in preparation) have proposed the hypothesis that the forms with a canine fossa would be the ancestors of H. sapiens. Because H. heidelbergensis (sensu Rightmire) does not have a canine fossa, neither the African nor the European members could be ancestors of H. sapiens. In our opinion, the first members of H. antecessor arrived to Europe and during the Middle Pleistocene originated the Neanderthals. Simultaneously, the African populations of H. antecessor evolved during the Middle Pleistocene to give rise to another intermediate species, in this case ancestral to H. sapiens, whose name should be either H. rhodesiensis or H. helmei [12]. This postulate poses the problem that the facial morphology should experience a short of reversion because a canine fossa has not been identified in the hypothetical direct ancestor of modern humans. At the time of understanding this phenomenon it should be considered that the specimen from Gran Dolina belongs to an immature individual of 11 years old; that is, the growth was not completed. One of the hypotheses for consideration is that adults of H. antecessor would have a smooth facial relief, meaning that the canine fossa would disappear or decrease along the growth. Conversely, H. sapiens would retain the immature morphology in the adults. It is open to the possibility that neoteny, a process of heterochrony, could from part of the evolutionary origin of H. sapiens; a classic view of human evolution [30]. All of these aspects raise a number of questions and offer a new scenario for research in paleoanthropology. There are several aspects to be studied, among which the possible relationship between North- and East African populations, on the one hand, and those from South Europe during the late Lower Pleistocene is outstanding.

ROSAS: Human Evolution in the Last Million Years — The Atapuerca Evidence

Figure 5

Phylogenetic scheme of the genus Homo after including the new species H. antecessor, recovered at the Lower Pleistocene of the Gran Dolina site, in Atapuerca. It can be seen that H. antecessor is located in a central position, as the last common ancestor of the Neandertals and H. sapiens. The species H. erectus evolved independently in Asia, and is identified as a lateral branch that is detached early from the basal stem defined by H. ergaster

15

ACTA ANTHROPOLOGICA SINICA

16

Supplement to Vol. 19, 2000

Acknowledgements: I am grateful to the people working every year in the Atapuerca sites. I am thankful to Markus Bastir and Cayetana Martínez for his comments and suggestions. I am especially grateful to Clubs de Rotarios for its financial support during my visit to China. This work is part of the project PB96-1026-C03 of the DGESIC (Spanish government), Programa de Unidades Asociadas, CSIC, and Junta de Castilla y León. References: [1] BERM ÚDEZ DE CASTRO JM, ARSUAGA JL, CARBONELL E et al. A hominid from the Lower Pleistocene of Atapuerca, Spain: Possible ancestor to neandertals and modern humans [J]. Science, 1997, 276:1392-1395. [2] CARBONELL E, BERM ÚDEZ DE CASTRO JM, ARSUAGA JL et al. Lower Pleistocene hominids and artifacts from Atapuerca-TD6 (Spain) [J]. Science, 1995, 269:826-829. [3] ROSAS A, BERM ÚDEZ DE CASTRO JM. The Mauer mandible and the evolutionary significance of Homo heidelbergensis [J]. Geobios, 1998, 31:687-697. [4] ROSAS A, BERM ÚDEZ DE CASTRO JM. On the taxonomic affinities of the Dmanisi mandible (Georgia) [J]. Am J Phys Anthropol, 1998, 107:145-162. [5] AGUIRRE E, ARSUAGA JL, BERM ÚDEZ DE CASTRO JM et al. The Atapuerca Sites and the Ibeas Hominids [J]. Hum Evol, 1990, 5:55-73. [6] ARSUAGA JL, CARRETERO JM, GRACIA A et al. Taphonomical analysis of the human sample from the Sima de los Huesos Middle Pleistocene site (Atapuerca/Ibeas, Spain) [J]. Hum Evol, 1990, 5:505-513. [7] ARSUAGA JL, MARTÍNEZ I, GRACIA A et al. Sima de los Huesos (Sierra de Atapuerca, Spain) [J]. The site. J Hum Evol, 1997, 33:109-127. [8] WOLPOFF MH, WU X, THORNE AG. Modern Homo sapiens origins: A general theory of hominid evolution involving the fossil evidence from East Asia [A]. In: The Origins of Modern Humans, Spencer FH ed. New York: Alan R. Liss, 1984, 411-483. [9] STRINGER CB, ANDREWS P. Genetic and fossil evidence for the origin of modern humans [J]. Science, 1988, 239:1263-1268. [10] LAHR MM. The multiregional model of modern human origins: A reassessment of its morphological basis [J]. J Hum Evol, 1994, 26:23-56. [11] LIEBERMAN DE. Testing hypotheses about recent human evolution from skulls [J]. Curr Anthropol, 1995, 36:159197. [12] STRINGER CB. Current issues in modern human origins [A]. Contemporary Issues in Human Evolution, 1996, 115134. [13] RIGHTMIRE GP. The human cranium from Bodo: Evidence for speciation in the Middle Pleistocene [J]? J Hum Evol, 1996, 31:21-39. [14] STRINGER CB, HOWELL FC, MELENTIS JK. The significance of the fossil hominid skull from Petralona, Greece [J]. J Archaeol Sci, 1997, 6:235-253. [15]

SCHOETENSACK O. Der unterkiefer des Homo heildergensisaus den Sanden von Mauer bei Heidelberg. W. Engelmann, 1908.

[16] WANG Q. An analysis of facial topography and its implication for Homo erectus. Phylogenetic position [A]. In: Abstracts International Symposium on Palaeoanthropology, Beijing 1999, 2. [17] CAPARROS M. Dali: archaic Homo sapiens or evolved Homo erectus?[A] In: Abstracts International Symposium on Palaeoanthropology, Beijing 1999, 21. [18] VALLOIS HV. Neandertals and Presapiens [J]. J R Anthropol Inst, 1954, 84:111-130. [19] PARÉS JM, PÉREZ-GONZ ÁLEZ A. Paleomagnetic age for hominid fossils at Atapuerca archaelogical site, Spain [J]. Science, 1995, 269:830-832. [20] CARBONELL E, BERM ÚDEZ DE CASTRO JM, ARSUAGA JL. Preface [J]. J Hum Evol, 1999, 37:309-311. [21] CARBONELL E, RODRÍGUEZ XP. Early Middle Pleistocene deposits and artifacts in the Gran Dolina site (TD4) of the "Sierra de Atapuerca" (Burgos, Spain) [J]. J Hum Evol, 1994, 26:291-311. [22] ROEBROEKS W, VAN KOLFSCHOTEN T. The earliest occupation of Europe. A short chronology [J]. Antiquity, 1994, 68:489-523. [23] CUENCA-BESC ÓS G, LAPLANA C, CANUDO JI. Biochronological implications of the Arvicolidae (Rodentia, Mammalia) from the Lower Pleistocene hominid-bearing level of Trinchera Dolina 6 (TD6, Atapuerca, Spain) [J]. J Hum Evol, 1999, 37:353-373.

ROSAS: Human Evolution in the Last Million Years — The Atapuerca Evidence

17

[24] ROSAS A, BERM ÚDEZ DE CASTRO JM. Human remains from the Gran Dolina (TD6 level, Sierra de Atapuerca, Spain), and the questionn of the common ancestor of Modern Humans and Neanderthals [A]. In: Abstracts International Symposium on Palaeoanthropology, Beijing, 1999, 37. [25] ARSUAGA JL, MARTÍNEZ I, LORENZO C et al. The human cranial remains from Gran Dolina Lower Pleistocene site (Sierra de Atapuerca, Spain) [J]. J Hum Evol, 1999, 37:431-457. [26] ANDREWS P. An alternative interpretation of characters used to define Homo erectus [J]. Cour Forsch Inst Senckenberg, 1984, 69:167-175. [27] WOOD BA. The origin of Homo erectus [J]. Cour Forsch Inst Senckenberg, 1984, 69:99-111. [28] HUBLIN JJ, TILLIER AM. Les enfants mousteriens de Jebel Irhoud (Maroc) comparaison avec les néandertaliens juvéniles d'Europe [J]. Bull. Mém. Soc. d'Anthrop. de Paris, 1988, 5 :237-246. [29] MAUREILLE B. La face chez Homo erectus et Homo sapiens: recherche sur la variabilité morphologique et métrique [D]. Thèse de l'Université Bordeaux I, nº 1157, Université Bordeaux, Bordeaux, 1994. [30] GOULD SJ. Ontogeny and Phylogeny [M]. Cambridge, Massachusetts: Harvard University Press, 1977.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

18-22

What Constitutes Homo erectus? Jeffrey H. SCHWARTZ1, 2, Ian TATTERSALL2 (1. Departments of Anthropology and History and Philosophy of Science, University of Pittsburgh, Pittsburgh, PA 15260, USA; 2. Department of Anthropology, American Museum of Natural History, New York, NY 10024, USA)

Abstract Although some paleoanthropologists maintain that the taxon Homo erectus not only subsumes Asian earlymiddle Pleistocene fossil hominids, but also contemporaneous specimens from Africa, morphology clearly sets off the Asian material from all other hominids. In recognition of notable morphological differences, the species H. ergaster has been proposed for the non-erectus African specimens. But even with this nod toward recognizing taxic diversity in the early-middle Pleistocene hominid fossil record, our studies indicate that neither “species” constitutes a unified taxon. Instead, within “erectus” as well as within “ergaster,” different cranial and dental morphs can be distinguished. In other groups of primates (e.g. lorisids), similar levels of difference serve to delineate genera. Although we are not necessarily advocating this degree of taxonomic representation across the board with regard to the morphs within “ergaster” and “erectus,” we do find this situation compelling in terms of calling for a greater recognition of species diversity within Homo. In this regard, we note that the more complete Sangiran crania, which are in external details similar to the type specimen of Homo erectus from Trinil, and thus also distinguished from the Ngandong and Zhoukoudian “Homo erectus,” are further distinguished from these as well as all other known hominids (and primates in general) in having a very derived pattern of intracranial sinus drainage. As such, the species erectus, which should be restricted to the Trinil and Sangiran hominid specimens, cannot be ancestral to any other known hominid.

Key words: Trinil; Sangiran; Ngandong; Zhoukoudian; Koobi Fora; Homo ergaster

1

Introduction

The history of Homo erectus as a recognized hominid species parallels that of hominids in general. Although early on in hominid studies many different taxa—species as well as genera— were recognized, the tendency during the latter half of the twentieth century was to collapse this acknowledgement of potential taxic diversity to conform to notions of a single evolving lineage, notions that were especially promulgated by two of the “founders” of the Evolutionary Synthesis, the geneticist Dobzhansky [1] and the ornithologist Mayr [2]. To be sure, separating Chinese from Indonesian H. erectus at the level of the genus (“Sinanthropus” and “Pithecanthropus,” respectively) is not systematically warranted at present. But the lumping together of hominid specimens into the same taxon largely because of the general time period they represent (indeed, the lumping together of any specimens on the basis of their chronostratigraphy) obscures not only a taxic diversity that typically characterizes other vertebrate clades, but also the morphologies that would serve to demonstrate this diversity by demoting their significance to the realm of mere “ individual variation.” Consequently, the differences that clearly exist not only among specimens of Asian “H. erectus,” but also between Asian and African “H. erectus,” are seen as resulting from “geographic variation” of the order that one expects only among “races” or “subspecies” of other vertebrate species. Wood’s [3] adoption of Groves and Mazak’s [4] species H. ergaster to recognize differences between African and Asian “erectus,” as bold as this was at the time, has not met with much support, as is evidenced by Walker and Leakey’s [5] monograph entitled The Nariokotome Homo erectus Skeleton and Rightmire’s [6] argument that neither morphological nor metric comparisons serve to distinguish significant patterns within the combined early-middle Pleistocene sample. Since tradition also has it that “Homo erectus” is ancestral to “Homo sapiens” (itself long-held to subsume widely differing “varieties” of human, even as uniquely configured as Neanderthals), it would be worthwhile to reopen the question for consideration: “What constitutes Homo erectus?”

Biography: The authors are currently engaged in a systematic study of the entire human fossil record to which they bring years of comparative experience from studies on the systematics and phylogenetic relationships of most other groups of fossil and living primates.

SCHWARTZ et al.: What Constitutes Homo erectus?

2

19

Dissecting the problem

From our perspective as mammalian systematists with particular interests in non-human primates, there is nothing biological that would preclude hominids from being approached methodologically in the same way as other taxa have been. Therefore, if we look at other primates, for instance lorisids, we find that at least four genera can be distinguished easily on the basis of various features of the skull (such as differences in orbital frontation, lipping, and orientation superiorly, elevation of the frontal, development of a postorbital sulcus, swelling of the snout by robust canine roots, anterior projection of the nasal bones beyond the lateral margins of the nasal aperture, rostral elongation of the premaxilla beyond the alveolar margin, robusticity of the zygoma and its outward flare relative to the lateral orbital margin, extension of the posterior root of the zygomatic arch relative to the superior margin of the acoustic meatus, completeness of petrosal expansion laterally, inflation of the mastoid region), mandible (relative height of corpus and roundedness of the gonial angle) architecture) as well as of the teeth (such as differences in relative sizes of upper incisors, canines, and anterior premolars, upper and lower last premolar morphology, cheek tooth cusp height and distinctiveness, and molar cusp patterns) [7-9]. In addition, lorisids are an interesting parallel to “Homo erectus” because they have African and Asian representatives, although, in the prosimians, an African genus pairs phylogenetically with an Asian genus, rather than with its geographic neighbor [7-9]. Turning to Homo erectus, and given the array of morphologically distinct specimens that have been allocated to the species, it would seem appropriate to begin discussion with the type specimen—“Pithecanthropus I” from Trinil, Indonesia. This is a thin-, not thick-boned calotte that is small in size, low in height, and relatively long. It sports a narrow, shelflike, and laterally flaring supraorbital region the flows without interruption into the gently sloped, long, and straight frontal plane. This latter bears an almost imperceptible midline keel, and is bounded low-down by extremely faint temporal lines that serve as the “borders” of the somewhat laterally swollen braincase. In posterior view, the specimen is very low and yet quite broad, with slightly inwardly tilted, short lateral vault walls and a severely posteriorly distended nuchal angle that presents itself as a horizontal torus-like structure lying between anteriorly angled occipital and nuchal planes. In comparison not only with other hominids, but also with hominoids and anthropoid primates more broadly, these features emerge by virtue of their singularity as potential apomorphies. These, in and of themselves, would certainly exclude Homo erectus from the ancestry of Homo sapiens. Sangiran 2, a virtually complete calvaria, represents a more robust individual than “Pithecanthropus I”, but the two are recognizably similar in their shared derived cranial features. It is thus reasonable to consider these specimens as belonging to the same morph, which, in this case, would be unambiguously identified as Homo erectus. Since Sangiran 2 is more complete toward its base than the Trinil calotte, additional morphological data can be gleaned from this specimen that expands our knowledge of H. erectus cranial morphology. During study of this specimen we found that the sigmoid sinus is unusual in that it bifurcates along the backside of the petrosal, with one branch descending inferiorly behind, and the other along the horizontal midline body of the petrosal, which is marked by an anteriorly opening groove or fissure that comes to engulf the region of the internal acoustic meatus. The inferior sigmoid branch may have exited the braincase via the jugular foramen, but it appears to have been coursing directly to the foramen magnum. It is not possible to follow the anteriormost extent of the superior branch because of damage to the specimen. Other Sangiran specimens, such as Sangiran 4, which is represented by the posterior half of the cranium, are of interest because they are more robust and thicker-boned than the Trinil and Sangiran2 specimens. In posterior profile, Sangiran 4 is not quite as low as the latter two H. erectus specimens, and it bears a distinct sagittal keel. Nevertheless, it is reasonable to include this specimen in erectus inasmuch as, accounting for variability, the occipital region does display the relevant derived features of this species. In addition, Sangiran 4 preserves internally a somewhat thin sigmoid sinus that bifurcates—actually arborizes—giving rise to two major branches: a superior one that courses along the fissured body of the petrosal and an inferior branch that

20

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

descends from the former just below the posterior extent of the thin superior petrosal sinus. As in Sangiran 2, the site of exit of the superior branch from the braincase cannot be identified. The inferior branch is coursing in the direction of the foramen magnum. In contrast to the condition of the Indonesian Homo erectus specimens, the typical configuration among mammals, including anthropoids, is for the single-channeled transverse sinus to descend uninterruptedly anteroinferiorly to the jugular foramen [10]. In such taxa as H. sapiens and H. neanderthalensis (and, see below, other “erectus”), the sigmoid sinus is strongly curved, while in many hominids (e.g. “australopiths”) as well as all other catarrhine primates, it is not. Nevertheless, since bifurcation of the sigmoid sinus is clearly not the common pattern, we can reasonably suggest that its presence in the Sangiran specimens would be a derived feature for them, and, given their clear association with the type specimen of H. erectus, derived for this species. In addition, since a fissured medial petrosal wall is not the rule in primates, this, too, emerges as a uniquely derived feature of H. erectus. Consequently, these features, in conjunction with the other derived cranial features, would preclude an ancestor-descendent relationship between H. erectus and any known hominid. But while the Sangiran specimens can reasonably be grouped with “Pithecanthropus” I, the Ngandong specimens are more enigmatic: Although bearing shelflike supraorbital tori, their crania are tall and relatively narrow [11], and thus totally unlike the Trinil and Sangiran crania in shape and proportions. Furthermore, as we discovered preserved in Ngandong 1, 7, and 13, these specimens also differ from those from Sangiran in not having an arborizing sigmoid sinus. Indeed, the Ngandong transverse sinus is single-channeled and, as the sigmoid sinus, arcs simply and primitively around the posterior margin of the unfissured petrosal. This is also the pattern that Weidenreich [12] described for “Sinanthropus” III, V, and XI, for which the temporal region was relatively intact; (Weidenreich illustrated only the temporal bone of “Sinanthropus” V). And, as the Ngandong crania differ markedly not only in size, but also in details of shape, from the Trinil and Sangiran specimens, so, too, as evidenced in the reconstruction by Tattersall and Sawyer [13], did Zhoukoudian “erectus.” While the systematic relations of the Ngandong and Zhoukoudian specimens awaits resolution, it is clear that by not sharing with the Sangiran and Trinil specimens their cranial and especially petrosal and sigmoid sinus apomorphies, the former should not be included in the taxon Homo erectus.

3

The African specimens

Compared to the type specimen of Homo erectus—with its long and low frontal flowing into a thin, laterally flaring shelflike brow, very acute nuchal angle, and, as it is wider than it is tall, its somewhat rectangular posterior outline—the African “ergaster/erectus” cranial specimens are clearly different. For instance, in KNM ER 3733 the brows arc over each orbit and project both up and forward, forming a posttoral sulcus behind; the frontal rises steeply from this sulcus and achieves the highest point of the cranial vault well forward in the profile; the temporal lines, which are distinctly crestlike and raised from the vault’s surface, arise behind the midpoints of the orbits; and, posteriorly, the skull is relatively taller than it is wide and the sidewall of the vault is arced. KNM WT 15000 differs from both Indonesian Homo erectus and KNM ER 3733. The braincase is short and well-rounded in profile, and the brows are merely thickenings of the superior orbital margins, which do not project superiorly, anteriorly, or laterally. Furthermore, the entire facial architecture below nasion is completely different from KNM ER 3733: for example, the lower face is narrower, longer and with greater alveolar prognathism, the nasal aperture is taller and narrower, and the long and relatively wide nasal bones are gently concave in profile and would not have projected above the nasal aperture. The specimen is subadult, and these features would, if anything, have become more exaggerated with growth, rather than transformed into KNM ER 3733 or the lower face of the cranially H. erectus-like Sangiran 17. This latter, although deformed upward, apparently broadened to the alveolar region and bore a narrowly triangular nasal aperture with short, superiorly tapering, and non-projecting nasal bones.

SCHWARTZ et al.: What Constitutes Homo erectus?

21

KNM ER 3883 differs yet again from Indonesian Homo erectus as well as from KNM ER 3733 and KNM WT 15000. The thickened supraorbital margins protrude out and slightly down, overhanging nasion and the face below; the flat-across nasal bones were apparently quite vertical below nasion and probably did not curve outward very much thereafter; and the frontal slopes strongly up and back, with the profile reaching its highest point well back along the sagittal suture. In notable contrast to KNM ER 3733 and KNM WT 15000, this specimen has a large and protrusive mastoid process and the zygoma flares out from top to bottom, whereas it is more vertical in KNM ER 3733 and KNM WT 15000. The distinctively different morphologies of each of the three “ergaster” skulls is highlighted by a second but more gracile specimen, KNM ER 3732, which is clearly similar to 3883 in preserved regions [14]. The unnatural association of specimens in a single species (erectus), or even two (ergaster and erectus), is further demonstrated by comparing the preserved M2 of KNM ER 3733 (with distinct trigon cusps, of which the paracone is much larger than the metacone; distolingually offset hypocone; well-marked cristae; and a thick precingulum, excavated trigon and talon basins, and smooth enamel) with the comparable tooth in KNM WT 15000 (which, like the other cheek teeth, is high crowned and almost flat occlusally with filled-in basins and subequal trigon cusps; it has a mesially arcing and thick preprotocrista, a thick postcingulum that broadens into a hypocone that extends lingually beyond and mesially part way around the protocone, and wrinkled enamel). If we were not dealing with hominids, these blatant differences in dental morphology would provoke any mammalian systematist to regard the two as representing different taxa. Comparison of these specimens with the much larger and almost ovoid M2 of Sangiran 4, with its subequal and low trigon cusps, massive and distally rounded hypocone, and accentuated postprotocrista, makes differences between the three even more clear-cut. If Sangiran 4 is Homo erectus, then, in terms of dental morphology the African forms are not. But is any of them H. ergaster? The only way in which one can try to answer that question is by comparing the type specimen of H. ergaster, KNM ER 992, which is a lower jaw with teeth, with the lower jaw of KNM WT 15000, which is the only skull with an associated mandible. In KNM ER 992 the lower canines are somewhat tall and pointed and compressed buccolingually; the anterior premolar is dominated by the protoconid and it bears small mesial and distal foveae; the posterior premolar and molars would not have had very pronounced basins; the posterior premolar protoconid and metaconid are subequal in size; all molars are elongate with rounded and protrusive hypoconulids and somewhat wrinkled enamel. In KNM WT 15000, the lower canine is short crowned with deep and relatively large mesial and distal fovea on either side of a stout lingual pillar that descends from the apex of the cups to the swell out the base; the premolars had deep mesial and distal basins, being even larger in the posterior of the pair; the anterior premolar is distended mesially; the two erupted molars bear deep though restricted talonid basins ringed by distinct, somewhat bulbous cusps; and the hypconulid on both molars is large and offset lingually. Clearly, since KNM ER 992 is the type specimen of H. ergaster, the morphological differences between it and KNM WT 15000 would preclude the latter from being included in the species. And since the much larger, and morphologically differing, preserved lower premolars and molars of Sangiran 9, a presumed representative of H. erectus, are not like either KNM ER 992 or KNM WT 15000, then neither of the African forms can be regarded as belonging to that taxon either. Even the morphology of the mandible of Sangiran 9, being much thicker boned with a long postincisal slope, differs markedly from KNM ER 992 as well as KNM WT 15000. KNM ER 992 is the type specimen of H. ergaster. And since it is distinguished from KNM WT 15000 on the basis of morphology of the lower dentition, and the latter is distinguished from KNM ER 3733 on the basis of M2 as well as cranial morphology, then the latter cannot be regarded as H. ergaster. With neither KNM ER 3733 and KNM WT 15000 allocable to H. ergaster, and each probably representing a distinct taxon, that leaves KNM ER 3883 and 3732 as the only cranial candidates of this species. But until specimens that have lower teeth associated with cranial remains are discovered, this part of the puzzle will remain unsolved.

ACTA ANTHROPOLOGICA SINICA

22

4

Supplement to Vol. 19, 2000

Conclusion

We hope to have demonstrated that even this small portion of the hominid fossil record provides a window onto a picture taxonomic diversity. Further, this diversity resembles that which systematists have long known characterizes the histories of virtually all other vertebrate groups. Using lorisids as an example, the degrees of difference one finds among these four prosimian genera are not appreciably different from those distinguishing the hominid specimens we have discussed. Clearly, there is nothing biological that precludes hominids from being analyzed by the standards that prevail in mammalian systematics generally. Acknowledgements: We thank Dr. Dong Wei and the organizing committee for their hospitality during the conference, and to the many curators and colleagues who graciously granted us permission to study the specimens in their charge that were crucial to this study. We are also grateful to Drs. H. McHenry and D. Pilbeam for reviewing this contribution. References: [1] DOBZHANSKY T. Evolution, Genetics, and Man [M]. New York: John Wiley, 1955. [2] MAYR E. Taxonomic categories in fossil hominids [J]. Cold Spring Harbor Symposium on Quantitative Biology, 1950, 5: 109-118. [3] WOOD B. Koobi Fora Research Project, Volume 4, Hominid Cranial Remains [M]. Oxford: Oxford University Press, 1991. [4] GROVES C, MAZAK V. An approach to the taxonomy of the Hominidae: gracile Villafranchian hominids of Africa [J]. Casopis pro mineralogii a geologii, 1975, 20: 225-246. [5] WALKER AC, LEAKEY R. The Nariokotome Homo erectus Skeleton [M]. Cambridge: Harvard University Press. 1993. [6] RIGHTMIRE GP. Evidence from facial morphology for similarity of Asian and African representatives of Homo erectus [J]. Am J Phys Anthropol, 1998, 106: 61-85. [7]

SCHWARTZ JH. Primate systematics and a classification of the order [A]. Comparative Primate Biology, Vol. 1: Systematics, Evolution, and Anatomy. New York: Alan R Liss, 1985. 1-41.

[8] SCHWARTZ JH. Issues in prosimian phylogeny and systematics [A]. Topics in Primatology, Vol. 3, Evolutionary Biology, Reproductive Endocrinology, and Virology. Tokyo: University of Tokyo Press, 1992, 23-36. [9] SCHWARTZ JH, TATTERSALL I. Evolutionary relationships of living lemurs and lorises (Mammalia, Primates) and their potential affinities with European Eocene Adapidae [J]. Anthropol Pap Am Mus Nat Hist, 1985, 60: 1-100. [10] SCHWARTZ JH, TATTERSALL, I. Variability in Hominid Evolution: putting the cart before the horse? [A]. Ú ltimos Neandertais em Portugal (Last Neandertals in Portugal): Odontologic and Other Evidence. Lisboa: Academia das Ciéncias de Lisboa, 2000. 367-401. [11] SANTA LUCA AP. The Ngandong Fossil Hominids: A Comparative Study of a Far Eastern Homo erectus Group [M]. Yale University Publication in Anthropology, No. 78, 1980. [12] WEIDENRIECH F. The Skull of Sinanthropus pekinensis: A Comparative Study on a Primitive Hominid Skull [M]. Palaeontol Sin, new series D, No. 10. 1943. [13] TATTERSALL I, SAWYER G. The skull of “Sinanthropus” from Zhoukoudian, China: a new reconstruction [J]. J Hum Evol, 1996, 31: 311-314. [14] SCHWARTZ JH, TATTERSALL I. The Human Fossil Record [M]. New York: John Wiley & Sons, in prep.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

23-33

Review of the Phylogenetic Position of Chinese Homo erectus in Light of Midfacial Morphology WANG Qian1,2, Phillip V. TOBIAS2 (1. Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, Beijing,100044, China; 2. Sterkfontein Research Unit, Department of Anatomical Sciences, University of the Witwatersrand, Parktown 2193, Johannesburg, South Africa)

Abstract Facial morphology is perhaps more informative in disclosing evolutionary trends than neurocranial morphology. Five midfacial morphological categories and two topographical forms can be recognized in Pleistocene hominids based on comparisons between Chinese Homo erectus and Afro-European H. heidelbergensis. Similarities between the low and flat faces with flexion of Chinese H. erectus and modern humans suggest their ancestor-descendant link, while a similarity in central puffiness reflects a link between H. heidelbergensis and H. neanderthalensis. But only the European branch of H. heidelbergensis evolved into the Neandertals, while the African one seems to have disappeared from the scene without issue. The facial evidence also discloses that H. erectus was once widespread in the Old World. The result appears to support the ancestral status of Chinese H. erectus and the “Multi-regional Evolution Hypothesis” for the origin of modern humans.

Key words:

Midface; Morphology; Topography; Chinese Homo erectus; Phylogenetic position

1

Introduction

Excavation at Zhoukoudian (Peking Man site) in the 1920s catalyzed the serious recognition of Homo erectus and also began a controversy on its destiny. The issue whether Asian H. erectus became extinct by replacement, or evolved into or is conspecific with H. sapiens is the topic of a vigorous debate, which is connected closely to models of the origin of modern humans [1,2]. Weidenreich noted a couple of anatomical traits in skeletons jointly shared by Zhoukoudian H. erectus (under the old name of Sinanthropus pekinensis) and modern “Mongoloids”, and he therefore proposed an ancestor-descendant relationship of H. erectus and modern Mongoloids[3]. This kind of regional evolutionary continuity was supported by later scholars on the basis of abundant fossil and cultural materials found in China and in Australia, and the “Multi-regional Evolution Hypothesis” for the origin of modern humans was formulated [1, 4-11]; With the recognition of genetic and cultural elements coming from outside, a “Continuity with Hybridization Model” was proposed by Wu Xinzhi to embrace both the local scenario of human evolution and the model of origins of modern humans in China [6, 12]. However, the robust nature of the Asian H. erectus neurocranium with its thick wall and superstructures is perceived by many scholars as a specialization and as a barrier against further evolution towards modern humans. Hence it is asserted that the Asian H. erectus is a side branch of human evolution, becoming extinct without issue in the late Middle Pleistocene [13-16, 18]. The denial of ancestral status to Asian H. erectus’s accompanied the birth of the “Replacement Model” or “Out Of Africa” hypothesis as an alternative to the “Continuity Model” in the interpretation of how modern humans originated. Although currently the denial of Asian H. erectus as an ancestor has resulted from research in genetics [19-20], paleolithic archeology [21-22], etc., the decisive reason still rests on anatomy [13-16]. The alternative claimant for the transition between early Homo and modern humans is believed to be the Afro-European Middle Pleistocene hominids, including African Bodo and Kabwe (Broken Hill), European Arago and Petralona etc., now often grouped in H. heidelbergensis, while the latter is also asserted to be the ancestor of H. neanderthalensis [16].

Biography: WANG Q, current post-doctoral fellow at the University of the Witwatersrand in Johannesburg, South Africa, specialized in the research of Homo erectus.

ACTA ANTHROPOLOGICA SINICA

24

2

Supplement to Vol. 19, 2000

The basis of effective phylogenetic analysis

If the current estimate of the oldest date for H. heidelbergensis, over 600ka (e.g. Bodo [16]) is correct, this western assemblage is roughly contemporary with the Middle Pleistocene Eastern Asian classic H. erectus, among which Chinese H. erectus represents an important part. It is true that the Afro-European H. heidelbergensis has more progressive features than the Chinese H. erectus represented by Peking Man at Zhoukoudian and comparable to Chinese Early H. sapiens (e.g. Dali, Jinniushan), in terms of overall cranial dimensions, cranial capacity and weak development of ectocranial superstructures[16, 23-25]. However, given that there are two assemblages at about the same geochronological level, but seemingly at different evolutionary grades, and that they probably represent different evolutionary directions or clades, a logical question arises: Does the evolutionary grade reflect the phylogenetic position? Morphologically, Chinese H. erectus and Early H. sapiens share many common characteristics, and from the former to the latter, there are similar morphological trends to those in other parts of the world [6, 12]. This hypothesized progressive transformation from Chinese H. erectus to Early H. sapiens would suggest that the former is still evolutionarily active and would tend to invalidate the view that it is an evolutionary cul-de-sac. When we take into consideration the fact that from the earliest hominids, the australopithecines, via H. habilis to modern humans, the neurocranium becomes increasingly bigger and rounder (or brachycephalized), it is quite probable that neurocranial morphology as a whole serves to indicate the evolutionary grade rather than the evolutionary clade , and we are led to conclude that neurocranial morphology alone cannot constitute a convincing foundation for rigorous phylogenetic analysis, and is not of complete phyletic valence. Tobias once warned against such partial characterization and developed an effective way of phylogenetic analysis that emphasized the overall comparison among different hominid taxa on the basis of Le Gros Clark’s concepts of “total morphological pattern” and “taxonomic relevance” and of Robinson’s “phyletic valence”[26-28]. To determine the phylogenetic position of Chinese H. erectus, a broader comparison based on more anatomical materials from a broader geographical range and longer geochronological framework is needed. By virtue of incomplete preservation of most hominid fossils, the analysis of the whole head-end would probably be adequately effective for the understanding of phylogenetic position. Therefore, extension of the analysis from the neurocranium alone, so as to include facial parts seems to be a good solution. This extension results in a smaller sample size of Eastern Asian including Chinese H. erectus, but the current accumulation of facial skeletons is sufficient to reveal a general facial pattern in Chinese H. erectus. Even in the face of neurocranial and dental analyses, facial parts still attract the attention of some scholars, such as Tobias [27, 29], Rak [30-31], Wu Xin-zhi [6], Pope [7, 32], Kramer [33], Rightmire [34] and Lockwood [35-36]. Among them, Rak’s detailed work on the faces of the australopithecines shows that the facial morphological and topographical patterns have important implications for our understanding of the diversity and relationships of early hominids [30]; and when Wu Xinzhi sums the common characters shared among Chinese fossils and modern humans, facial characters occupy a greater percentage of all assorted common features (for example, 8 of 11), while all the remarkable diachronic changing features are concentrated almost exclusively in the neurocranial part[6]. This implies that facial morphology can better reveal the real evolutionary scenario in China than neurocranial morphology. Their works as a whole reflect the phylogenetic value of facial morphology and also indicate that facial studies are no longer subordinate to cranial and dental analyses in the assessment of the taxonomic and phylogenetic status of certain early hominids. The analysis of facial morphology among fossils and extant hominids may be expected to disclose phylogenetic relationships more reliably. The aim of this article is to review the status of Chinese H. erectus in human evolution in the light of facial analysis, on the basis of the evolutionary stages disclosed by neurocranial characters. This analysis concentrates on the general mid-facial morphological pattern. The definition of

WANG et al.: Review of the Phylogenetic Position of Chinese Homo erectus in Light of Midfacial Morphology

25

midface adopted in this article is Pope’s [32], which is modified from Rak [30]. Pope defines the middle face as that portion of the anterior cranium that is visible in norma frontalis, and is bounded by the projective distance from the alveolar plane of the maxilla to nasion and transversely from zygion to zygion. The materials are the original fossils or casts of AfroEuropean and Chinese hominids housed in the Institute of Vertebrate Paleontology and Paleoanthropology, Beijing, China, and the Department of Anatomical Sciences of the University of the Witwatersrand, Johannesburg, South Africa.

3

The general midfacial morphology of Chinese H. erectus

An entire intact facial skeleton of H. erectus in China has not yet been found. However, the accumulation of incomplete facial skeletons of H. erectus in China enables us to draw an outline midfacial configuration by viewing the midface as a whole in morphological terms, rather than by emphasizing isolated characters. The facial fragments of the Zhoukoudian hominids permitted Weidenreich to reconstruct a Peking Man skull with a face (female) [3]. Although the face is compounded of facial bones belonging to different individuals of both genders, it provides a general image of Peking Man’s midface. The midface is low and flat. There are two well-defined angulations or flexions. Medially, the portion of the maxilla lateral to the pyriform aperture faces laterally, while the infraorbital part faces anteriorly; thus they jointly constitute a concave flexion. Laterally, a convex flexion is formed between the maxillary process which faces anteriorly and the temporal process of the zygomatic which faces laterally. A bony tubercle is present here, which can be construed as a torus accompanying the bony angulation, like the supraorbital torus and the occipital torus. From the norma basilaris, this lateral flexion is well shown by the angulation between the lower margins of the zygomatic process of the maxilla and the maxillary process of the zygomatic bone. The angle is about 105 in both Zhoukoudian No. II and Tangshan I. A partial skeleton found in 1993 corroborates Weidenreich’s reconstruction. It is attached to the fossil remains labeled Tangshan skull I found in a karstic cave in Tangshan hill, 30 kilometers east of Nanjing. The morphology of the cranium is very comparable to that of Zhoukoudian specimens.[37] The partial left facial skeleton is in perfect condition without deformation. It contains the almost complete nasal bones of both sides, the almost complete left zygomatic bone and the major portion of the left maxilla. The part lateral to the pyriform aperture and the part near the alveolar margin are missing. It is the first time that we have a tell-tale facial skeleton of classic H. erectus found in China to the present. The midface as a whole is low and flat with two well-defined flexions, just like those of the reconstructed Peking Man. A more important contribution that the Tangshan face makes towards the understanding of the Chinese H. erectus is that it factually and materially demonstrates that the low, flat face of modern Mongoloids has ancient roots extending through Early H. sapiens stage as far back as H. erectus. The specimens of Chinese Early H. sapiens likewise display this kind of flexed face, as in the relatively complete Dali skull and Jinniushan skulls. Even before the discovery of the informative Tangshan face, Wu Xin-zhi had proposed that the low flat face had a long history in China and that Dali was not the earliest example [24]. In the Maba skull, despite poor preservation of the facial skeleton, it is possible to observe that the frontal process of the zygomatic bone shows a plane facing forwards just like other Chinese hominids, an indirect sign of a flexed face. It is true for the specimens of Chinese Late H. sapiens such as the Liujiang skull, which has been claimed to be Proto-Mongoloid.[38] Meanwhile, it should be kept in mind that a close examination of the mid-facial region reveals discernible differences between Chinese H. erectus and modern Mongoloids, for example, in the nature of the medial flexion: in the former there is a furrow as Rightmire [34] points out and in the latter there is a real canine fossa. However, these differences can be read as stages in a morphogenetic process (see below). Though displaying variations of dimension and of the degree of robusticity in the facial skeleton, the faces of Chinese ancient hominids and extant humans all follow the same pattern:

ACTA ANTHROPOLOGICA SINICA

26

Supplement to Vol. 19, 2000

low and flat with well-defined flexion. The fact that a general common facial morphology or topography exists among Chinese fossil and extant humans supports the claimed genetic link between Chinese H. erectus and modern Mongoloids including modern Chinese. In brief, the similarity among Chinese H. erectus, Early H. sapiens, and modern humans in general facial morphology and topography is strong supporting evidence for evolutionary continuity in China and also for the ancestral status of H. erectus.

4 The difference between Chinese H. erectus and Afro-European H. heidelbergensis and its implications With the dismissal by many workers of the view that H. erectus was ancestral to H. sapiens, alternative assemblage, the Afro-European Middle Pleistocene hominids, now frequently grouped in H. heidelbergensis, embracing Bodo of Ethiopia, Kabwe (Broken Hill) I of Zambia, Arago XXI of France, Petralona of Greece, have been proposed to be the intermediate link between early African small-brained hominids and Homo sapiens [18]. As mentioned above, the overall neurocranial morphology and endocranial capacities of these Afro-European hominids make them comparable to Chinese Early H. sapiens. However, their faces are distinctly different from those of Chinese H. erectus and Early H. sapiens, in possessing relatively high faces and, moreover, emerging inflation or puffiness in the central part of the face. H. heidelbergensis lacks the medial flexion of the midface by its puffiness in the portion lateral to the pyriform aperture. As a result, the infraorbital portion faces more laterally than in the faces of Chinese H. erectus. Therefore, even the anterolateral surface of the frontal process of the zygomatic bone faces more laterally in H. heidelbergensis than in Chinese H. erectus. This orientation should have made the lateral flexion more acute; however, the lateral orientation of the puffy infraorbital portion in H. heidelbergensis not only offsets this shift, but also weakens the degree of angulation between them. The angle between the lower margins of the zygomatic process of the maxilla and the maxillary process of the zygomatic bone is 125-130 in Petralona, Kabwe (Broken Hill) I and Arago XXI (left side), which is more obtuse than the 105 in Chinese H. erectus. Generally speaking, the face of H. heidelbergensis is weaker in the degree of flexion but shows signs of emerging puffiness in the central facial region. Its high and puffy face contrasts sharply with the low, flat face possessed by Chinese H. erectus. It is a significant difference. Although the degrees of central puffiness are different in different specimens, the face of H. heidelbergensis is readily topographically linked to that of the Neandertals. In short, Chinese H.erectus and Afro-European H. heidelbergensis differ from one another not only in neurocranial development, but also in facial morphology and topography. On the basis of the evolutionary stages reflected by the neurocranial parts, the significant topographic differences in the faces of Chinese H. erectus and Afro-European H. heidelbergensis suggest that Chinese H. erectus acquired the low, flat face and then gave birth to modern humans via Early H. sapiens, whilst H. heidelbergensis with a moderately puffy face was ancestral to H. neanderthalensis. H. neanderthalensis finally attained an evolutionary end-point and became extinct without issue [39], the high and puffy face being not present among modern humans. Thus facial comparison between Chinese H. erectus and H. heidelbergensis is informative. A wider range of general facial morphology and topography in the Pleistocene may help us to see more clearly the evolution of the hominid face and its phylogenetic value.

5

General facial patterns in the Pleistocene and their relationships

Five facial morphological patterns and two topographical variants may be recognized in the Pleistocene after early H. erectus, according to the basic superficial morphology including angle and size in the middle facial region. Facial Pattern I : Moderately high, flat or slightly centrally puffy face.

WANG et al.: Review of the Phylogenetic Position of Chinese Homo erectus in Light of Midfacial Morphology

27

Africa: SK 847, ER3733, ER3883 Facial Pattern II: Low and flat, with two well-defined flexions, well-developed canine jugum, no real canine fossa. well defined sub-infraorbital foramen groove. The infraorbital plane is flat and faces forward. Low middle face. Nasal bones pinched. Low orbit. Eastern Asia: Zhoukoudian, Tangshan, Gongwangling, Yunxian I, Sangiran 17, Dali, Jinniushan Africa: Ndutu Europe: Gran Dolina, Steinheim Facial Pattern III: Moderately or rather high face with moderate central puffiness. Weak first flexion, well defined sub-infraorbital foramen groove. The infraorbital plane is convex and faces a little upward. Asia: None Africa: WT15000, Kabwe (Broken Hill) 1, Bodo Europe: Arago XXI, Petralona, Atapuerca (Sima de los Huesos) Facial Pattern IV: Modern face. Well defined canine fossa, no strong canine jugum, maxillary portion lateral to the pyriform aperture is small, less robust. Specimen: Modern humans, Upper cave, Liujiang, Djebel Irhoud, Cro-Magnon, Wadjak, Qafzeh, Skhul, Tabun, Omo I, Kabwe (Broken Hill) 2, Laetoli 18, ?Border Cave1, ?Florisbad. Facial Pattern V: No first flexion, total central projection. Mid-face is high and puffy. The orbit is high and round, the whole infraorbital region forms a single plane sloping from the orbit antero-inferiorly and from the pyriform aperture postero-laterally. Europe: the Neandertal specimens Western Asia: Amud, Shanidar Eastern Asia: None Africa: None A remarkable phenomenon is that the specimens belonging to these five facial categories could be roughly grouped geochronologically or geographically, or both, which means we may be able to trace their phylogenetic relationships in the light of the facial patterns by the geochronological and geographical groups. Facial Pattern I occurs mainly in African Early H. erectus or H. ergaster. Facial Pattern II occurs mainly in Chinese hominids since the Middle Pleistocene, and also in Middle Pleistocene specimens from African Ndutu and European Gran Dolina. Facial Pattern III occurs exclusively in the Middle Pleistocene Afro-European H. heidelbergensis, and not in Eastern Asian examples. Facial Pattern IV is in Late H. sapiens and modern humans all around the world, while Facial Pattern V is in the European and Western Asian Neandertals. In addition, the five morphological types can be divided into two topographic categories: the puffy one and the non-puffy one. One keeps the short and flat, and less robust pattern, one is high with central puffiness. The non-puffy form includes Patterns I, II, IV. The puffy form includes Patterns III and V. The relationships among these facial types are clear if we consider their representatives in chronological and geographical framework. In Facial Pattern II, if it became less robust and there was modification with the emergence of the real canine fossa, the midface would be modern. In Facial Pattern III, if its emerging central puffiness intensified, Facial Pattern V would result. If Facial Pattern IV had led to Facial Pattern III, the modern one, it world have had to go back to Facial Pattern II first, with shortening of the face and flattening of the middle part; this would have been an example of an evolutionary reversal and we doubt whether this would have been likely. When we go back to the Early Pleistocene, the facial pattern we encounter is not puffy, but flat. Both Facial Pattern II and Facial Pattern III may be traced back to Facial Pattern I, the facial form of Early H. erectus or H. ergaster. Facial Pattern I is a kind of prototype, from which two kinds of faces could have been introduced through two different pathways of modification , to

ACTA ANTHROPOLOGICA SINICA

28

Supplement to Vol. 19, 2000

Facial Pattern II by lowering of the face, posterior migration of the lower part of the infraorbital region, and shortening of the dental arcade; to Facial Pattern III by projecting of the central part and prolonging of the height of the face. Type I is a logical modification from the flat and inferoanteriorly sloping face of H. habilis, the latter, introduced from the “dished face” possessed by the australopithecines by shortening of the facial width, weakening of the buttress with less masticatory demand. If the links proposed between Facial Patterns I and III, and Patterns II and IV are correct, the model of facial evolution in the Pleistocene could be reconstructed as follows:

Facial Pattern I (Prototype)

Pattern II

Pattern IV (extant)

Pattern III

Pattern V (extinct)

When we incorporate the hominid assemblages grouped by facial patterns, the relationships among the Pleistocene hominids could be formulated as follows (here we group Chinese H. erectus in Late H. erectus ): Late H. erectus

H. sapiens

H. heidelbergensis

H. neanderthalensis

Early H. erectus

Therefore, based on facial morphology, two general human evolutionary trends can be recognized in the Pleistocene. From Early H. erectus, two kinds of faces emerged during the Middle Pleistocene, a rather low and flat face of Late H. erectus and a high face and emerging central puffiness in H. heidelbergensis. They attained different evolutionary stages and headed along different evolutionary pathways, then reached different ends. The former facial patterns evolved into that of H. sapiens and the latter into that of the Neandertals followed by extinction without issue. The remarkable topographic split happens at least in the Middle Pleistocene. It is interesting to note that there are already signs of the beginnings of the separation into two facial patterns in Early H. erectus specimens. For example the maxillae lateral to the pyriform aperture face forward in ER3733 or are slightly puffy in WT15000. Maybe after them the differentiation became increasingly marked. The puffy face is a derived character, that flourished in Africa, Europe and West Asia since the Middle Pleistocene, but disappeared in the Late Pleistocene. The non-puffy face is a continuous and common phenomenon in the Pleistocene; it can be traced back to Early H. erectus and even to H. habilis. Hence the non-puffy face with flexion is not necessarily a modern or apomorphic character as favoured by de Castro et al.[40]. On the contrary, it may be seen as a plesiomorphic character. From Early H. erectus to Late H. erectus and then H. sapiens, the neurocranium continues to enlarge while the facial part retains the same non-puffy topography. Therefore, it would be safe to assert that H. erectus displays increasing evolutionary progress in the neurocranial part, while the facial topography shows a relative stasis. The nature of this dynamic combination of neurocranial and facial developments may be used as the basis for a tentative phenetic approach to the defining of H. erectus, although we are well aware that dental size and form, cranial form, thickening and pneumatization, ectocranial superstructure and mandibular traits, not to speech of postcranial morphology, which we do not consider here may all constitute valuably to the definition of H. erectus.

6 6.1

The issue of European and African H. erectus

European H. erectus No fossils unequivocally classified as H. erectus were reported before the 1990s; for quite a long time, there was virtually no solid fossil evidence of European H. erectus. However, the

WANG et al.: Review of the Phylogenetic Position of Chinese Homo erectus in Light of Midfacial Morphology

29

Spanish Gran Dolina[40] and Italian Ceprano[41] specimens found in the 1990s have helped to overcome the dearth of early hominids in Europe by extending the fossil evidence of European human history back to 780ka B.P. Also they have provided new evidence of the possible existence of H. erectus in Europe. The face of juvenile specimen ATD6-69 is the earliest occurrence of the modern face in the fossil record. As discussed above, the flexed face if of plesiomorphic nature and hence does not of itself constitute an effective taxonomic basis for the proposed new species H. antecessor in which to classify the fossil materials found at Gran Dolina. We suggest that if complete cranial specimens are found in the future, they may well be comparable to those of the Peking Man. On the facial morphology, we tentatively regard the Gran Dolina specimens together with the Ceprano cranium as falling in the category of Late H. erectus together. They represented the descendants of early immigrants from Africa at the stage of Early H. erectus or earlier, and gave birth to European H. heidelbergensis. This may be the reason that we can pick up some features usually encountered in specimens of Late H. erectus in this European assemblage, such as the angular torus in the Arago skull [23, 42]. If this is correct, Pope’s proposal that the near human midfacial morphology first appears in Eastern Asia and later spreads to other regions would not be correct [7]. This kind of face is a pan-Old World phenomenon of the Middle Pleistocene. It appears also in Africa. 6.2

African H. erectus The midface of the Tanzanian Ndutu skull is placed in our list in Facial Pattern II. It is in Rightmire’s assemblage of H. heidelbergensis. Clarke has discussed the significance of the Ndutu skull. He at first believed this 400ka old specimen belonged to H. erectus [43]. Later he re-classified it as African archaic H. sapiens [17]. However the overall skull morphology, low endocranial capacity (930-960cc) and especially the general facial configuration of the Ndutu skull as reconstructed by Clarke are very much what we encounter in Chinese H. erectus. The differences between them probably reflect geographical and geochronological variations. We therefore propose to group the Ndutu skull in the category of Late H. erectus with Chinese H. erectus. Although Late H. erectus is still poorly documented in Africa, the Ndutu skull seems to disclose the presence of this taxon. More specimens might be found if more efforts were directed to the identification and excavation of Middle Pleistocene deposits. Clarke argues also that previously classified African H. erectus is a separate species and he calls it H. leakeyi [17]. He suggests that H. erectus is confined to Eastern Asia and is not associated with Acheulean handaxe technology. These views lead him to propose that both the morphology and the cultural nature of H. erectus point to its being an evolutionary dead-end. Recently, however, an unequivocal Acheulean industry has been found in Eastern Asia and especially at Bose Basin in South China in the Middle Pleistocene.[44] This vitiates the argument - that dates back to Movius [21] - that Acheulean technology is totally absent from Eastern Asian sites. The fossils assigned to H. ergaster and H. leakeyi are on our thinking representative of Early H. erectus [16, 34]. They are simply more primitive than the specimens of Chinese and Indonesian H. erectus, and it is proper to refer to Eastern Asian Middle Pleistocene H. erectus as Late H. erectus in terms of their advanced nature. (The words “Early” and “Late” here have to do only with the evolutionary stage, not with taxonomic nomina.) From African Early H. erectus, at least two human forms emerge in Africa in the Middle Pleistocene. One is H. heidelbergensis, and the other is Late H. erectus. While the European H. heidelbergensis evolved into the Neandertals, the apparent absence of unequivocal Neandertals in Africa suggests that the African branch of H. heidelbergensis became extinct without issue. The Late H. erectus seems to have given rise to H. sapiens. The coexistence of Late H. erectus and H. heidelbergensis in Africa can provide a better interpretation of the broad spectrum of variation among African hominids of the Middle Pleistocene [25]. 6.3

The earliest hominid species out of Africa It is generally believed that the earliest hominid species to leave the African continent, cradle

ACTA ANTHROPOLOGICA SINICA

30

Supplement to Vol. 19, 2000

of humankind, was H. erectus over 1 million years ago. Now the new cultural and geochronological findings in Eastern Asia, such as Longgupo in China [45], and Modjokerto in Indonesia [46], suggest that early hominids left the African continent earlier than previously believed, and that the hominids involved in this first exodus were perhaps H. habilis or H. ergaster [47- 48]. If this is correct, the interpretation of Tobias and von Koenigswald in the 1960s that hominids at the stage of H. habilis existed in Asia would be vindicated [49]. However, because these new scenarios of early human evolution are not yet widely supported, we tentatively place all the European and Asian hominids of the Early Pleistocene in Early H. erectus even though they are not yet well documented by fossil materials.

7

A hypothesis of human evolution in the Pleistocene

By incorporating the fossil findings of all the Old World according to their classification into the general ancestor-descendant links, we hypothesize from the facial evidence a general evolutionary model in three geographical zones as follows (E: Early; L:Late; X:Extinct): Eastern Asia Terminal Middle Pleistocene

H. sapiens

Africa H. sapiens

Europe H. sapiens H. neanderthalensis(X)

Middle Pleistocene

L. H. erectus L. H. erectus H. heidelbergensis(X)

Early Pleistocene

E. H. erectus

E. H. erectus

H. heidelbergensis

E. H. erectus

H. habilis

This model is a linear model and is based on a linear facial evolutionary model recognized above. It is admittedly open to criticism. In fact, gene exchange should be expected in the formation of H. sapiens and other species between African and Asian branches. As the second author (PVT) states[50], “Of course, the picture must have been far more complex than a simple linear progression from early to late forms in each region: gene flow and drift would have been expected.” Yet this model, based on facial evolution, can offer a basis for further investigation and interpretation and an alternative human evolutionary scenario especially in the Middle Pleistocene. On this model, the Asian and especially Chinese H. erectus occupies an ancestral position and thus does not support the narrowly-interpreted Out-of-Africa model. It can be classed as one of the variants of the Multi-regional Origin model, or another kind of “Di-regional model of human evolution” [51]. This simple model does not cover portions of Western Asia and Australasia, where the issues of human evolution may be more complicated (We are aware also that the classification of the fossil record into geographical zones is an artifact of geological sampling and fortuitous fossil discoveries. Most of the Old World was a single, diversified geographic area, almost certainly without great discontinuities). There are two key features to this model. The first is the continuity of facial continuity in Africa and Eastern Asia and discontinuity with Europe, suggesting a dual origin of modern humans; the second is the coexistence of L. H. erectus and H. heidelbergensis in Africa in the Middle Pleistocene, or a dual morphological trend, resulting from a prior split of the last common ancestor of these two clades. The story of the African part points

WANG et al.: Review of the Phylogenetic Position of Chinese Homo erectus in Light of Midfacial Morphology

31

to the continuity with Asia and discontinuity with Europe. The respective human evolutionary trends in Eastern Asia and in Europe, based on analysis of facial morphology and topography, help us to understand better the nature of the greater variation of Middle Pleistocene hominids in Africa and to formulate a dual evolutionary model in Africa since Early H. erectus. This model of dual clades awaits further finds and interpretations to be tested. It will not be surprising if more specimens of fossil hominids comparable to Chinese H. erectus, as was claimed by Arambourg for the Ternifine remains form Algeria under the original name of Atlanthropus mauritanicus [52], or specimens with features related to the Neandertals, come to light in the African continent. An issue introduced by this analysis is the mechanism of the facial split. Facial growth is not a simple thing. It relates to neural (brain and eyes), nasal (respiratory and olfactory systems), and oral (dental, masticatory and spoken language) complexes. Facial modification is achieved by genetic, biomechanical and clonal factors [53]. How to explain the split during the Middle Pleistocene, including the functional, developmental and selective aspects, and the extinction of the puffy pattern, is challenging, we do not attempt it here. Another problem would be the issue of the co-existence of two human species in Africa, their ecological associations and mutual relationship. The phenomenon of the synchronic and even sympatric co-existence of two or more hominid species exists also at the stages of Australopithecus (of which we have at least nine species have been recognized), and early Homo in the Plio-Pleistocene. The reason for the different fates of H. heidelbergensis in Africa and in Europe awaits interpretation too. Perhaps competition with H. erectus may be responsible for the earlier extinction of H. heidelbergensis in Africa.

8

Conclusion

From the analysis of facial morphology and topography starting with Chinese H. erectus, some preliminary conclusion or extrapolations may be drawn. The result supports the ancestral status of Chinese H. erectus and appears to be in keeping with a modified version of the “Multiregional Evolution Hypothesis” for the origin of modern humans. 1. The similarities in the possession of low and flexed faces by Chinese H. erectus and at least some modern humans (especially Mongoloids) strongly support their ancestor-descendant link, while the similarities in central puffiness reflect a close link between H. heidelbergensis and the Neandertals. But only the European branch evolved into H. neanderthalensis, while the African populations of H. heidelbergensis seems to disappeared from the scene without issue. 2. H. erectus is not confined to East Asia. It is a pan-Old World phenomenon, occuping almost the whole Old World during the first half of the Pleistocene. 3. The African hominids show more variants than those of the other two major continents in the Old World and probably embrace at least two clades which appear in Eastern Asia and in Europe respectively during the Middle Pleistocene. The Ndutu skull is probably an African representative of Late H. erectus. 4. The facial morphology is more informative in disclosing evolutionary trends than neurocranial morphology. The latter can better disclose the evolutionary stage. The facial pattern could be predicated on evolutionary stages reflected by neurocranial morphology in certain regions. Acknowledgment: We are grateful to Professor Wu Xin-zhi for reviewing this article. The first author (QW) sincerely thanks the National Research Foundation of South Africa (NRF), and for Special Subjects of the Natural Scientific Foundation of China (NSFC) for financial support, and also thanks the Institute of Vertebrate Paleontology and Paleoanthropology (IVPP), Chinese Academy of Sciences and the Department of Anatomical Sciences of the University of the Witwatersrand, Johannesburg. The second author (PVT) is grateful to the PAST Fund, the

ACTA ANTHROPOLOGICA SINICA

32

Supplement to Vol. 19, 2000

Department of Arts, Culture, Science and Technology of the South African Government and the Department of Anatomical Sciences of the University of the Witwatersrand, Johannesburg. We appreciate the help of Mrs. Heather White. References: [1]

WOLPOFF MH, WU X,THORNE AG. Modern Homo sapiens origins: A general theory of hominid evolution involving the fossil evidence from East Asia [A]. Origins of Modern Humans: A World Survey of the Fossil Evidence. .New York: Alan R. Liss, 1984, 414-484.

[2]

STRINGER CB. Middle Pleistocene hominid variability and the origin of Late Pleistocene humans[A]. Ancestors: The Hard Evidence. New York: Alan R. Liss, 1985, 289-295.

[3]

WEIDENREICH F. The skull of Sinanthropus pekinensis: A comparative study on a primitive hominid skull [M]. Pal Sin Ser D, Vol 7, 1943.

[4]

WU R. Chinese human fossils and the origin of Mongoloid racial group [J]. Anthropos (Brno), 1986, 23:151-155.

[5]

THORNE AG,WOLPOFF MH. Regional continuity in Australasian Pleistocene hominid evolution [J]. Am J Phys Anthropol, 1981, 55:337-349.

[6]

WU X. The evolution of humankind in China [J]. Acta Anthropol Sin, 1990, 9:312-321.

[7]

POPE GG. Craniofacial evidence for the origin of modern humans in China [J]. Year Book of Phys Anthropol, 1992, 35:243-298.

[8]

WU X, POIRIER FE. Human Evolution in China: A Metric Description of the Fossils and a Review of the Sites [M]. New York:Oxford University Press, 1995.

[9]

LIU W, YANG M. The changes of tooth size of Chinese and the systematic status of Homo erectus in East Asia [J]. Acta Anthropol Sin, 1999, 18:176-192.

[10] LING S. Comparison of technological mode of Paleolithic Culture between China and the West [J]. Acta Anthropol Sin, 1996,15:1-20. [11] ZHANG S. On the important advancement of the Paleolithic archeology in China since 1949 [J]. Acta Anthropol Sin, 1999, 18:193-214. [12] WU X. Chinese human paleontological study in the 20th century and prospects [J]. Acta Anthropol Sin, 1999, 18:165-175. [13] DELSON E , ELDREDGE N, TATTERSALL I. Reconstruction of hominid phylogeny: a testable framework based on cladistic analysis[J]. J Hum Evol, 1977, 6:263-278. [14] WOOD BA. The origin of Homo erectus [J]. Cour Forsch Inst Senckenberg, 1984, 69:99-111. [15] STRINGER CB. The definition of Homo erectus and the existence of the species in Africa and Europe [J]. Cour Forsch Inst Senckenberg, 1984, 69:131-143. [16] RIGHTMIRE GP. The Evolution of Homo erectus: Comparative Anatomical Studies of an Extinct Human Species [M]. Cambridge: Cambridge University Press, 1990. [17] CLARKE RJ. The Ndutu cranium and the origin of Homo sapiens [J]. Am J Phys Anthropol, 1990, 19:699-736. [18] RIGHTMIRE GP. The human cranium from Bodo, Ethiopia: evidence for speciation in the Middle Pleistocene? [J] J Hum Evol, 1996, 31:21-39. [19] CANN RL, STONEKING M, WILSON AC. Mitochondrial DNA and human evolution [J]. Nature, 1987, 325:31-36. [20] STRINGER CB, ANDREWS P. Genetic and fossil evidence for the origin of modern humans [J]. Science, 1988, 239:1263-1268. [21] MOVIUS HL. Lower Paleolithic archaeology in southern Asia and the far East [J]. Studies in Anthropol, 1948, 1:1781. [22] IKAWA-SMITH F. Early Paleolithic in South and East Asia [M]. Mouton: The Hague, 1978. [23] WU R, WU X. Comparison of Tautavel man with Homo erectus in China [A]. L’Homo erectus et la place de l’Homme de Tautavel parmi les hominides fossiles. Nice: 1er Congres International de Paleontologie Humaine, 1982, 606-616. [24] WU X. Comparative study of early Homo sapiens from China and Europe [J]. Acta Anthropol Sin, 1988, 7:287-293. [25] WU X, BRAUER G. Morphological comparison of archaic Homo sapiens crania from China and Africa [J]. Acta Anthropol Sin, 1994, 13:93-103. [26] TOBIAS PV. Single characters and the total morphological pattern redefined: the sorting effected by a selection of morphological features of the early hominids [A]. Ancestors: The Hard Evidence. New York:Alan R. Liss, 1985, 94101.

WANG et al.: Review of the Phylogenetic Position of Chinese Homo erectus in Light of Midfacial Morphology

33

[27] TOBIAS PV. The Skulls, Endocasts and Teeth of Homo habilis (Olduvai Gorge:Volume 4) [M]. Cambridge: Cambridge University Press, 1991. [28] LE GROS CLARK WE. The Fossil Evidence for Human Evolution [M]. Chicago: Chicago University Press, 1964. [29] TOBIAS PV. The Cranium and Maxillary Dentition of Australopithecus (Zinjanthropus) boisei (Olduvai Gorge: Volume 2) [M]. Cambridge: Cambridge University Press, 1967. [30] RAK Y. The Australopithecine Face [M]. New York: Academic Press, 1983. [31] RAK Y. The Neanderthal: A new look at an old face [J]. J Hum Evol, 1986, 15:151-164. [32] POPE GG. Evolution of the zygomaticomaxillary region in the genus Homo and its relevance to the origin of modern humans[J]. J Hum Evol, 1991, 21:189-213. [33] KRAMER A. Human taxonomic diversity in the Pleistocene: Does Homo erectus represent multiple hominid species[J]? Am J Phys Anthropol, 1993, 91:161-171. [34] RIGHTMIRE G. Evidence from facial morphology for similarity of Asian and African representatives of Homo erectus[J]. Am J Phys Anthropol, 1998, 106:61-85. [35] LOCKWOOD CA. Variation in the face of Australopithecus africanus and other African hominoids [D]. Johannesburg: University of the Witwatersrand, 1997. [36] LOCKWOOD CA, TOBIAS PV. A large male hominin cranium from Sterkfontein, South Africa, and the status of Australopithecus [J]. J Hum Evol, 1999, 36:637-685. [37] LU Z. Hominid fossils[A]. Nanjing Hominid Site. Beijing: Wenwu Press, 1996, 15-82. [38] WOO JK.Human fossils found in Liujiang, Kwangsi, China [J]. Vertebr PalAsiatica, 1959, 3:109-118. [39] BRAUER G. “The Afro-European sapiens-Hypothesis” and hominid evolution in East Asia during the Late Middle and Upper Pleistocene[J]. Cour Forsch Inst Senckenberg, 1984, 69:145-165. [40] DE CASTRO JMB, ARSUAGA JL, CARBONELL E et al. A hominid from the Lower Pleistocene of Atapuerca, Spain: possible ancestor to Neandertals and modern human [J]. Science, 1997, 276:1392-1395. [41] ASCENZI A, BIDDITTU I, CASSOLI PF et al. A calvarium of late Homo erectus from Ceprano, Italy [J]. J Hum Evol, 1996, 31:419-423. [42] GRIMAUD D. La parietal de l’Homme de Tautavel [A]. L’Homo erectus et la Place de l’Homme de Tautavel parmi les Hominides Fossiles. Nice: 1er Congres International de Paleontologie Humaine. 1982, 62-109. [43] CLARKE R J. New cranium of Homo erectus from Lake Ndutu,Tanzania [J]. Nature, 1976, 262:485-487. [44] HOU Y, POTTS R, YUAN B et al. Mid-Pleistocene Acheulean-like stone technology of the Bose basin, South China [J]. Science, 2000, 287:1622-1626. [45] HUANG W, CIOCHON R, GU Y et al. Early Homo and associated artifacts from Asia [J]. Nature, 1995, 378:275278. [46] SWISHER CC, CURTIS GH, JACOB T et al. Age of the earliest known hominids in Java, Indonesia [J]. Science, 1994, 263:1118-1121. [47] CIOCHON R. The earliest Asians yet [J]. Nat Hist, 1995, 104:50-54. [48] WOOD BA, TURNER A. Out of Africa and into Asia [J]. Nature, 1995, 378:239-240. [49] TOBIAS PV, VON KOENIGSWALD GHR. A comparison between the Olduvai hominines and those of Java and some implications for hominid phylogeny [J]. Nature, 1964, 204:515-518 [50] TOBIAS PV. Africa-derived skulls and Africa-derived mitochondrial DNA: Towards a reconciliation [A]. The Origin and Past of Modern Humans as Viewed from DNA. Tokyo: World Scientific Press, 1995, 189-215. [51] BABA H. Diregional model of human evolution [A]. The Origin and Past of Modern Humans as Viewed from DNA. Tokyo: World Scientific Press, 1995, 244-266. [52] ARAMBOURG C. Le gisement de Ternifine I. Deuxieme Partie:L’Atlanthropus mauritanicus [M]. Arch De l’Inst De Pal Hum, 1963, 32: 37-190 [53] MOORE WJ, LAVELLE CLB. Growth of the Facial Skeleton in the Hominoidea [M]. London: Academic Press, 1974.

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

34-40

Restoration of the Face of Javanese Homo erectus Sangiran 17 and Re-evaluation of Regional Continuity in Australasia Hisao BABA1, Fachroel AZIZ2, Shuichiro NARASAKI3 (1. Department of Anthropology, National Science Museum, Tokyo, 3-23-1 Hyakunincho, Shinjuku-ku, Tokyo, 169-0073 Japan; 2. Quaternary Geology Laboratory, Geological Research and Development Centre, Bandung, Jl. Diponegoro 57, Bandung, 40122 Indonesia; 3. Laboratory of Biological Anthropology, Gunma Museum of Natural History, 1674-1 Kamikuroiwa, Tomioka, Gunma, 370-2345 Japan)

Abstract The Javanese Homo erectus Sangiran 17 skull was reconstructed, using the plaster casts made from six portions of the original skull. The mandible was newly restored, modifying the Zhoukoudian Homo erectus mandible XII. The facial skeleton is much robust and less prognathic, compared to that of Zhoukoudian Homo erectus reconstructed by Tattersall and Sawyer [1]. Then, the face of Sangiran 17 was restored based upon the reconstructed skull, arranging the chewing and facial muscles and other organs. The restored face is wide, moderately low, and less prognathic, having a thick eyebrow and a small nose. Within the seven facial characters considered to be shared between Sangiran 17 and Australians by Thorne and Wolpoff [2], only two characters (eversion of the lower border of the zygoma and lack of a nasal sill) are present but the other five are not present in both original and reconstructed Sangiran 17 skulls, which does not support the hypothesis of regional continuity in Australasia.

Key words:

Homo erectus; Sangiran 17; Face restoration; Australasia; Regional continuity

1

Introduction

So far as we know, the Sangiran 17 skull is the best preserved skull of Asian Homo erectus, because it has a face (Fig. 1). The Sangiran 17 skull was found in 1969 from Puchung site in Sangiran, Central Java, by Mr. Tukimin, a local farmer [3]. The skull is believed to be derived from the Kabuh (Bapang) Formation and dates to 0.7 to 0.8 million yr. BP [4-5]. The original Sangiran 17 skull is kept in the Geological Research and Development Centre, Bandung, under the care of the second author (FA). The skull had partly been reconstructed and studied by several authors [1, 6-10]. Recently the skull was re-arranged and stabilized by us [11]. However, the face remains distorted considerably.

2

Morphology of Sangiran 17 Skull

The Sangiran 17 skull provides an almost complete skull vault (Fig. 1). But its face is more or less damaged, the mandible is missing, and the left one third of the upper face is broken off. The remaining portions of the face are also distorted considerably [2, 9-11]. The skull shows typical characteristics of Asian Homo erectus from Lower / Middle Pleistocene [9-11]. The vault is long, wide and low, with remarkable development of the superstructures. The vault is large, but the endocranial volume of 1,004 ml [12] is moderate due to thickness of the bones (10mm at the bregma). It provides thick brow ridge and occipital torus, a clear bregmatic eminence, bulged supramastoid crests, and distinct temporal crests, which makes the vault outline rhombic in the lateral aspect and heptagonal in the occipital aspect. The frontal sinus is well developed in the medial half of the brow ridge. The Sangiran17 face is wide and low. It exhibits large orbital openings, a low nasal bridge, a small nasal aperture, a swollen and flared right zygomatic bone (left one is missing), a considerably projected palate, and moderately sized teeth [11, 13].

Biography: Hisao BABA, 1968, graduated Faculty of Science, University of Tokyo; 1970, Master of Science (University of Tokyo); 1983, Doctor of Medical Science (University of Tokyo); 1995, Director of the Department of Anthropology, National Science Museum and Professor of the Department of Biological Sciences, University of Tokyo.

BABA et al.: Restoration of the Face of Javanese Homo erectus Sangiran 17 and Re-evaluation of Regional Continuity in Australasia

Figure 1

Re-arranged original skull of Sangiran 17

35

ACTA ANTHROPOLOGICA SINICA

36

Supplement to Vol. 19, 2000

In the Sangiran 17 skull, the temporal fossae are deep in both anterior and posterior portions, the temporal crests are well ridged, and the supramastoid crests are thick and projected about 8mm high from the temporal fossa. This indicates marked development of the temporalis muscles. The right zygomatic bone of the Sangiran 17 skull is located anteriorly and flares laterally. It is extremely large and bulged so that the inferior border, from which the masseter muscle arises, is wide and situated low, near to the alveolar surface. These features mean that the masseter muscle had to be thick and located low, which implies that the region around the mandibular angle, to which the masseter inserts, might have been well developed and located low [13].

3

Reconstruction of Sangiran 17 Skull Cast

Negative molds of the Sangiran 17 skull were made divided into six separated portions by Mr. Shokichi Miyamoto during joint research supported by the Japanese International Cooperation Agency (JICA) and stored in the Geological Research and Development Centre, Bandung, under care of the second author (FA). The original Sangiran 17 skull was distorted and crushed considerably. However, since the skull is too hard and brittle to modify, we had to give up to use the original skull and decided to make up the precise cast in a supposed original shape, in the following procedure. At first, we made positive casts using these molds. Then, in order to correct displacements and distortions, we cut casts into several pieces and adjusted the position, referring to the original specimen, according to our knowledge of Homo erectus morphology (Fig. 2). Thirdly, we restored missing portions of the left face based on the shape of the right face .

Figure 2

4

Reconstructed cast of the Sangiran 17 skull

Restoration of the Sangiran 17 Face

We asked Mr. Yoichi Yazawa, who has been working on reconstruction and restoration under our supervision for these ten years, to carry out further restoration with us. At first, we restored a mandible of Sangiran 17 using the Sinanthropus mandible (XII) [14]. That is, we made the dental arcade wider and the mandibular ramus a little higher. Thus we obtained a presumed whole restored skull of Sangiran 17 (Fig. 3). Second, using clay, we attached the temporalis and masseter muscles to the skull, of which area and thicknesses were determined based upon the muscle markings on the skull and upon our

BABA et al.: Restoration of the Face of Javanese Homo erectus Sangiran 17 and Re-evaluation of Regional Continuity in Australasia

37

knowledge of human and ape anatomy. The attachment area and thickness of the muscles are much larger in the Sangiran 17 skull than in recent humans and more or less larger than in other fossil hominids, as discussed above. Third, we attached facial muscles to the skull, using wax sheets of various thicknesses. There are no significant difference in the structure of facial muscles between apes and humans, as might be true in Homo erectus. Thus, we restored the facial muscles according to our knowledge of human anatomy (Fig. 3).

Figure 3

Process of reconstruction and restoration of the Sangiran 17 face. Skull reconstruction (upper right), masticatory muscle restoration (lower left), facial muscle restoration (upper left), and a live face (lower right)

Fourth, we attached ears by using wax sheets and salivary glands using clay to the face. We set eye balls in the orbital cavities, using ready made eye balls for mannequins, which are a little larger than those of the actual humans and fit well with the large orbital cavities in the Sangiran 17 skull (Fig. 3). Fifth, we restored the skin, using wax sheets covering all of the head and face (Fig. 3). The thickness of the skin including subcutaneous fat varies from portion to portion, e.g. it is thin in the nasal region (2-3 mm), medium over the skull (5-10 mm), and thick in the cheek (15-30 mm).

ACTA ANTHROPOLOGICA SINICA

38

Supplement to Vol. 19, 2000

Finally, we colored the skin dark brown, as is seen in recent humans in low latitudes. We did not reconstruct the hair and beard, because they conceal the details of shape in the head and face. The Sangiran 17 head, of which shape directly reflects the skull vault shape, is low, long and wide with its maximum width at the lower part of the vault (Fig. 3). In the lateral view, there are three projections (the brow ridge, occipital torus and bregmatic eminence), which make the vault outline rhombic in shape. All these characteristics should have been more or less shared in other Homo erectus in Asia and Africa. The forehead is flat and strongly receded, showing Javanese Homo erectus features. The Sangiran 17 face is wide and moderately high, with a rounded rectangular outline, which might represent presumed common features shared in Homo erectus (Fig. 3). The mouth is, however, moderately projected compared with those of other Homo erectus. In addition, the cheeks are much swollen, and the nose is small. Consequently, the face looks very flat [13]. These features in Sangiran17 show the uniqueness of this specimen, and curiously resemble those of recent Northeast Asians, which imply some similarity in masticatory adaptation rather than direct phylogenetic relationship between them [13].

5

Re-examination of the Australasian Regional Traits in Sangiran 17

Thorne and Wolpoff [2] stated that there were twelve morphological characters shared in the Sangiran 17 and Australians, including fossil and recent specimens, which reflects regional continuity in Australasia. Among them, we re-examined seven facial characters, such as marked prognathism, presence of a marked ridge paralleling the course of the zygomaxillary suture (or malar tuberosity), the eversion of the lower border of the zygoma, the rounding of the inferolateral border of the orbit, lack of the nasal sill, marked expression of the dental plane curvature, and the degree of facial and dental reduction. These characters had been obscure, but were revealed clearly by the present observation and reconstruction of the Sangiran 17 skull (Fig. 1, 2). The degree of facial prognathism is moderate in the reconstructed Sangiran 17 skull face as in other later Homo erectus (Fig. 2). Moreover, the zygoma is extraordinary bulged anteriorly as well laterally, exceeding the condition of flat “generalized” faces in other Homo erectus and archaic Homo sapiens (cf. [15]). Consequently, the Sangiran 17 face looks flat. Fenner [16] originally described the malar tuberosity in Australians, which is a prominent ridge located on the external surface of the zygoma. Wolpoff argued that the tuberosity did exist in the original (personal communication in Habgood [17]). Thorne and Wolpoff [2] also pointed out the presence of the ridge paralleling the course of the zygomaxillary suture. We confirmed that there was neither a true tuberosity nor the ridge paralleling the suture. In our opinion, the ridge described by Thorne and Wolpoff [2] is actually a postmortem deformation as described above, but a few small (1-2 mm in length) processes were seen on the inferolateral border of the external surface of the zygoma (Fig. 1). As for the eversion of the lower border of the zygoma, it exists in Sangiran 17. But in Sangiran17, the whole zygoma is bulged laterally, and the partial eversion is not seen on the lower border (Fig. 1). Weidenreich [18], Howell [19], Thorne and Wolpoff [2], and Habgood [17] have mentioned that the rounding of the inferolateral border of the orbit is often seen widely in Homo erectus and archaic Homo sapiens. According to the present observation, Sangiran 17 does not exhibit this rounding on the inferolateral border (Fig. 1). Some rounding is, however, seen on the inferomedial border, although it is somewhat obscured by the deformation. In most Homo erectus and archaic Homo sapiens, there is no distinct sill (line or ridge) in the lower border of the nasal aperture which divides the nasal floor and the nasoalveolar clivus [18-19]. Our observation on the preserved right half of the lower border confirmed Thorne and Wolpoff’s claim that neither a line or ridge existed. The marked expression of the curvature of the posterior alveolar plane of the maxilla was described by Thorne and Wolpoff [2]. In addition, Rightmire [9-10] stated that the occulusal surface

BABA et al.: Restoration of the Face of Javanese Homo erectus Sangiran 17 and Re-evaluation of Regional Continuity in Australasia

39

slopes sharply (unnaturally) upward from front to back in the Wolpoff’s reconstruction. The occulusal as well as alveolar surface, however, slopes moderately or almost horizontally in the present reconstruction (Fig. 2). It does not make sense to compare the degree of facial and dental reduction between Sangiran17 and Australians, because masticatory adaptation pattern is completely different from each other. That is, Sangiran 17 has the largest middle face in Homo erectus ever recovered and small teeth as a later Homo erectus, which suggests a certain dietary adaptation, for example to eat tough and clean food substances in wet area such as forest. On the other hand, Australians’ teeth are the largest in the living humans but the middle face is not so large as those of recent Northeast Asians, which suggests that the dental attrition in Australians is mainly a result of contamination of sand grains in the food in arid areas such as desert. As a summary of our observations on these seven characters, the Sangiran 17 face shows the smooth lower border of the nasal aperture and eversion of the lower border of the zygoma. But, we confirm that Sangiran 17 face does not possess the marked facial prognathism, rounding of the inferolateral border of the orbit, malar tuberosity, and steep curvature of the posterior alveolar plane. Moreover, the comparison of facial and dental reduction does not make sense. Thus, so far as the facial morphological characters are concerned, the regional continuity in Australasia is far less evident than Thorne and Wolpoff [2] argued. Acknowledgements: We thank to Geological Research and Development Centre for the permission of the study of the Sangiran 17 skull. References: [1] TATTERSALL I, SAWYER GL. The skull of “Sinanthropus” from Zhoukoudian, China: a new reconstruction [J]. J Hum Evol, 1996, 31:311-314. [2] THORNE AG, WOLPOFF MF. Regional continuity in Australasian Pleistocene hominid evolution [J]. Am J Phys Anthropol, 1981, 55: 337-349. [3] SARTONO S. Discovery of another hominid skull at Sangiran, Central Java [J]. Curr Anthropol, 1972, 13:124-126. [4] MATSU7URA S. A chronological framing for the Sangiran hominids; Fundamental study by the fluorine method [A]. Bull Nat Sci Mus, Tokyo, Series D, 1982, 8: 1-53. [5] ITIHARA M, SUDIJONO, KADAR D et al. Geology and stratigraphy of the Sangiran area [A]. In: WATANABE N, KADAR D eds. Quaternary Geology of the Hominid Fossil Bearing Formations in Java. Geological Research and Development Centre, Special Publication, 1985, No. 4, 11-43. [6] SARTONO S. Implications arising from Pithecanthropus VIII [A]. In: TUTTLE RH ed. Paleoanthropology: Morphology and Paleontology. The Hague: Mouton, 1977, 326-360. [7] JACB T. Morphology and palaeoecology of early man in Java [A]. In: TUTTLE RH ed. Paleoanthropology: Morphology and Paleontology. The Hague: Mouton, 1975, 311-325. [8] SANTA LUCA AP. The Ngandong Fossil Hominid [M]. Yale University Publication on Anthropology, 1980, No. 78. 1175. [9] RIGHTMIRE BP. The Evolution of Homo erectus: Comparative Anatomical Studies of an Extinct Human Species [M]. Cambridge: Cambridge University Press, 1990, 1-260. [10] RIGHTMIRE BP. Evidence from facial morphology for similarity of Asian and African representative of Homo erectus [J]. Am J Phys Anthropol, 1998, 106: 61-85. [11] AZIZ F, BABA H, WATANABE N. Morphological study on the Javanese Homo erectus Sangiran 17 skull based upon the new reconstruction [A]. Geological Research and Development Centre, Paleontology Series, 1996, 8: 11-25. [12] HOLLOWAY RL. Early hominid endocasts: volumes, morphology and significance for hominid evolution [A]. In: TUTTLE RH ed. Primate Functional Morphology and Evolution. The Hague: Mouton, 1981, 393-415. [13] BABA H. Diregional model of human evolution [A]. In: BRENNER S, HANIHARA K eds. The Origin and Past of Modern Humans as Viewed from DNA. Singapore: World Scientific, 1995, 244-266. [14] WEIDENREICH F. The Mandible of Sinanthropus pekinensis, A Comparative Study [M]. Palaeontol Sin, Series D, 1938, VI-3, 1-169.

40

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

[15] SMITH FH, PAQUETTE ST. The adaptive basis of the Neandertal face form, with some thought on the nature of modern human origins [A]. In: TRINKAUS E ed. The Emergence of Modern Humans. Cambridge: Cambridge University, 1989, 182-210. [16] FENNER FJ. The Australian Aboriginal skull: its non-metrical morphological characters [J]. Trans R Soc South Aust, 1939, Vol. 63, Part 2, 248-306. [17] HABGOOD PJ. The origin of anatomically modern humans in Australasia [A]. In: MELLERS P, STRINGER CB eds. The Human Revolution: Behavioral and Biological Perspectives on the Origin of Modern Humans. Edinburgh: Edinburgh University Press, 1989, 245-273. [18] WEIDENREICH F. The Skull of Sinanthropus pekinensis: A Comparative Study on a Primitive Hominid Skull [M]. Palaeontol Sin, New Series D, 1943, 10: 1-484. [19] HOWELL FC. Hominidae [A]. In: MAGLIO VJ, COOKE HB eds. Evolution of African Mammals. Cambridge: Harvard University Press, 1978, 154-248.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

41-45

Two New Human Fossil Remains Discovered in Sangiran (Central Java, Indonesia) Dominique GRIMAUD-HERVE 1, Harry WIDIANTO 2, Teuku JACOB 3 (1. Laboratoire de Préhistoire du MNHN, UMR 6569, Institut de Paléontologie Humaine, 1 rue rené Panhard, 75013 Paris, France, UAMI: Université de Provence, UAMII: Université de la Méditerranée; 2. Balai Arkeologi, Jl Gedongkuning 174, Kota Gede, 55171 Yogyakarta, Indonesia; 3. Department of Physical Anthropology, Faculty of Medicine, Gadjah Mada University, Yogyakarta, Indonesia)

Abstract The discovery of two new human fossil remains in the Sangiran dome confirms that observed morphological characters such as a low and elongated skull with distinct maximal and biparietal breadths, the presence of sagittal keeling and supra-orbital, angular and occipital toruses, and nucchal plane and occipital squama angulation are common to the asian Homo erectus population.

Key words:

Homo erectus; Sangiran dome

1

Grogol Wetan

1.1

Discovery and stratigraphical position of the hominid fossil This human fossil remain was discovered in 1993 by Sugimin in the Grogol Wetan site, in the Sangiran dome (Central Java).The fossil was found in the Kabuh layers whose base is dated at approximately the lower/middle Pleistocene boundary. A first attempt to date by the Argon method gives an estimated age of 0.78 ± 0.29 million years [1]. This quite fragmented skull is now kept in Suakade Prambanan. 1.2

Conservation, age and sex The general preservation of the fossil is quite good. The fossil, which is in need of a careful reconstruction, possesses almost the whole skullcap. Several isolated bone fragments, broken at the time of discovery, were also discovered. The well conserved parts of the cranium are: the entire left side of the frontal bone, the sagittal part of the right frontal region, both parietal bones, the occipital bone (nuchal part is missing), the left and right posterior parts of the temporal bones, together with the upper part of the right temporal squama. The anterior right part of the broken maxillar bone shows both incisors, the canine, the first premolar and the root of the second premolar. Another maxillar fragment has conserved the second and third right molars. The only unbroken isolated teeth found are the first and the third left molars. The synostosis of the cranial sutures is complete. All the superstructures (occipital torus, metopic and bregmatic eminences, sagittal keeling and torus angularis) defined by Anton [2] are present; thus, we may infer that the remains of the Grogol Wetan hominid represent an adult specimen. The cranial superstructures are not well developed and the bones are quite thin (7mm). In essence, this fossil is very similar to Sangiran 2. These characteristics would indicate that this specimen is a female hominid. 1.3

Morphological description This morphological description is preliminary; a complete restoration ought to improve the first observations presented here. 1.3.1 Lateral view The preserved part of the superior supra-orbital torus shows a thickness identical to Sangiran2’s in norma lateralis. The supra-orbital sulcus is not marked. The frontal bone in the region of the forehead does not show a marked convexity.

42

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

The sagittal curvature is regular up to the prelambda plane. The occipital squama’s convexity is slight, similarly to other Sangiran skulls. It is separated from the transversal occipital torus by a discrete relief. This torus presents a clear angulation between the inferior and superior occipital planes, which is characteristic of the Homo erectus group sensu stricto[3]. The temporal lines present an apparent but blunt relief, and are in a high position on the parietal bones, as we have observed on Sangiran 2, 10, 12, 17 and on Trinil 2 [4-5]. The superior and inferior lines are separated as they rise from the frontal, their relief stays constant until its termination, at which point the superior line thickens in a slight angular torus. The convexity of the superior temporal border, well preserved on the right, is not marked; this character is common to the Sangiran hominids from this stratigraphic level, such as Sangiran 12 or 17.

Figure 1

Grogol Wetan human fossil skull, left lateral view

1.3.2

Superior view The superior outline of the Grogol Wetan skull has a sphenoïd shape, with a low and posterior position of the maximal cranial breadth and an accentuated narrowing in the post-orbital region. The beginning of a very marked sagittal keeling is noticeable, just after the supra-orbital sulcus. The parietal eminences do not constitute individualized protrusions; the same state applying to the preserved left frontal part. 1.3.3

Posterior view The cranial outline in posterior view is trapezoïdal as observed on the other Sangiran skulls, with a very low maximal cranial breadth at the level of the supramastoidal crests. This breadth is different from the biparietal maximal breadth which is situated at the end of the inferior temporal lines. The superior region of the skull is rather horizontal and not at all convex. The occipital torus, clearly apparent, is ligthly marked, but less than on Sangiran 12 ( 17 ); it shows a slight superior sulcus.

GRIMAUD-HERVE et al.: Two New Human Fossil Remains Discovered in Sangiran (Central Java, Indonesia)

43

1.3.4

Anterior view The two preserved elements of the supra-orbital torus are totally fused. We suggest that this torus corresponds to type III of Cunningham [6]. This character is again common to the others Homo erectus from Sangiran and Trinil.

2 Bukuran 2.1

Discovery of the fossil In 1996, a new hominid skull was discovered in the Bukuran site, in the Sangiran dome (Central Java). It was found in the Kabuh layers. This fossil is kept in the department of Physical Anthropology of Gadjah Mada University, in Yogyakarta. 2.2

Conservation, age and sex The state of preservation of the Bukuran fossil is quite good. It consists of two parts, the parieto-occipital region and the frontal region. All the cranial sutures are clearly apparent, the lambdoïdal one being open. The robustness of the supra-orbital and occipital toruses is very marked. These characters allow us to attribute this hominid to a young male. 2.3

Morphological description

2.3.1

Lateral view The skull outline is elongated and low. The supra-orbital torus and the glabellar region are strongly developped. The post-orbital sulcus is marked. The sagittal convexity of the frontal and parietal bones is regular. The convexity of the occipital squama is low, and so is the nuchal plane’s; these two regions are separated by a marked occipital torus with the characteristic angulation of the Homo erectus group. The temporal lines are situated in the middle part of the parietal. Their relief is discrete and blunt. There seems to exist a pathology, given that the relief of the inferior temporal line is more pronounced that the superior one. We observe an angular torus separated from the superior line by a depression; this is rather uncommon. A large number of wormians bones is noticeable: 5 in the lambdatic region, 2 on the left lambdatic suture and 1 on the right. No wormian bones have been observed on the others cranial regions. The convexity of the superior temporal border is not very accentuated. It is similar to the one of Trinil 2 or 17. 2.3.2

Superior view The cranial outline of Bukuran is sphenoïdal, similar to those of the hominids discovered in the Kabuh layers of Sangiran, such as Trinil 2, Sangiran 2, 10, 12 or 17. The maximal cranial and biparietal breadths are distinct and situated in a low and posterior position. The first one is at the level of the supra-mastoïdal crest while the second is in the posteroinferior part of the parietal, on the angular torus. The post-orbital constriction is very marked, and the narrowing of the frontal squama is accentuated. A sagittal keeling is observed on the medial part of the frontal squama; it disappears up to the bregmatic eminence and continues from there on. Parietal protrusions have not been noted. The observed occipital convexity is regular.

44

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

2.3.3

Anterior view No frontal eminences has been noted on the Bukuran skull. All the elements of the supra-orbital torus are fused. This torus corresponds to the type III of Cunningham. It is robust with a deep sulcus on it, similarly to Grogol Wetan, Trinil 2, Sangiran 2 or Sangiran 17. 2.3.4

Posterior view The outline is pentagonal with a horizontal superior vault. The sagittal keeling is apparent from this view. The occipital torus is developped and shows a strongly individualized external occipital eminence.and a supra-toral sulcus. Like the others skulls of the same of this site, the inferior part of the occipital (the nucchal part) is more developped than the superior part. The convexity of these two occipital regions is not accentuated.

Figure 2

Bukuran human fossil skull, ¾ left view

3

Conclusion

The two new hominid fossil remains found in the Sangiran dome at the Grogol Wetan and Bukuran sites represent adult specimens. The presence of the characteristic occipital torus, the metopic eminence, the bregmatic eminence, a sagittal keeling and an angular torus are well differentiated on these two subjects. We know, according to Anton [2, 7], that the differentiation of these superstructures appear ontogenitically in this order as a function of age. Thus, these two skulls represent the adult stage. The first observed morphological characters show a low and elongated skull for both fossils, with namely a sphenoïdal type outline, and a maximal cranial breadth (situated in a low and posterior position at the level of the supramastoïdea crests) distinct from the maximal biparietal one (situated on or near the torus angularis). All these characters are common to the asian Homo erectus population [8]. We observed a supra-orbital torus whose elements are fused both in Grogol Wetan and Bukuran,with no apparent frontal or parietal protrusions.

GRIMAUD-HERVE et al.: Two New Human Fossil Remains Discovered in Sangiran (Central Java, Indonesia)

45

The presence of a sagittal keeling and angular torus have been noted, so are a strong occipital torus along with the characteristic angulation found in the Homo erectus group. Finally a more developped nucchal plane than the occipital squama is apparent. The morphological characters described are common to the hominids discovered in this geological region (Kabuh layers). These observations allow us to integrate the two fossils in the Trinil - Sangiran population; this illustrates the very marked homogeneity of this fossil group. Acknowledgments: We are grateful to Dr Teuku Jacob, Gadjah Mada University, and to the director of Ditlinbinjarah, Yogyakarta, for access to specimen in their care, François Semah and Miguel Caparros for their criticism and helpful discussion. References: [1] SALECKI H. Apport d’une Intercomparaison de Méthodes Nucléaires (230 th/234u, ESR et 40Ar/39Ar) à la Datation de Couches Fossilifères Pléistocènes dans le Dôme de Sangiran (Java Central) [D], Thèse de Doctorat du M.N.H.N., Paris: 1997, 238. [2] ANTON SC. Developmental age and taxonomic affinity of the Modjokerto child Java, Indonesia [J]. Am J Phys Anthropol, 1997, 102, 497-514. [3] WIDIANTO H. Unité et Diversité des Hominidés Fossiles de Java: Présentation de Restes Humains Fossiles Inédits [D]. Thèse de Doctorat du M.N.H.N., Paris: 1993, 284. [4] GRIMAUD-HERVÉ D, JACOB T. Les pariétaux du Pithécanthrope Sangiran 10 [J]. L ’Anthropologie, 1983, 87:469-474. [5] SARTONO S, GRIMAUD-HERVÉ D. Les pariétaux des Pithécanthropes Sangiran 12 et 17 [J]. L ’Anthropologie, 1983, 87: 475-482. [6] CUNNINGHAM DJ. The evolution of the eyebrown region of the forehead with special reference to the successive supraorbital development in the Neanderthal race [J]. Trans R Soc Edinb, 1908-1909, 46:283-311. [7] ANTON SC. Cranial growth in Homo erectus: How credible are the Ngandong juveniles [J]? Am J Phys Anthropol, 1999, 108:223-236. [8] GRIMAUD-HERVÉ D. The parietal bone of indonesian Homo erectus [J]. Hum Evol, 1986, 1:167-182.

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

46-51

Finding of a Hominid Lower Central Incisor During the 1997 Excavation in Sangiran, Central Java Hisao BABA1, Fachroel AZIZ2, Shuichiro NARASAKI3, SUDIJONO2, Yousuke KAIFU1, Agus SUPRIJO4, Masayuki HYODO5, Eko Edi SUSANTO2, Teuku JACOB4 (1. Department of Anthropology, National Science Museum, Tokyo 169-0073, Japan; 2. Geological Research and Development Centre, Bandung 40122, Indonesia; 3. Laboratory of Biological Anthropology, Department of Biology, Gunma Museum of Natural History, Gunma 370-2345, Japan; 4. Laboratory of Bio- and Palaeoanthropology, Faculty of Medicine, Gadjah Mada University, Yogyakarta 55002, Indonesia; 5. Research Center for Inland Seas, Kobe University, Kobe 657-8501, Japan)

Abstract A new hominid lower central incisor with incomplete root formation was found during our excavation in 1997 in the Sangiran area. The stratigraphic level of the specimen was tentatively inferred as between the Upper and Lower Tuffs of the Bapang Formation. Comparisons with various Plio-Pleistocene hominid specimens from Eurasia and Africa were made and the specimen showed morphological affinity with Lower/Middle Pleistocene Homo erectus or Homo aff. H. erectus, as expected from its assigned time frame. The background of the find was explained and some information concerning developmental stage of the owner of the tooth was given.

Key words:

Homo erectus; Hominid incisor; Sangiran; Indonesia

On June 24, 1997, a hominid lower left central incisor with incomplete root formation was unearthed from the Bapang (Kabuh) Formation in the Sangiran area by the Indonesia-Japan joint research team. This find is important in that it was found through a systematic excavation, and therefore the find spot is precisely known. So far almost all the Javanese hominid fossils have been found and collected by local inhabitants, and therefore their stratigraphic positions are variously ambiguous [1-3]. This new specimen has been provisionally designated as Bs 9706 following the numbering system of the Geological Research and Development Centre, Bandung. Basic description, systematic comparisons with other fossil specimens from Indonesia, China, Africa, and Europe, and discussion on the affinity of the specimen have been published elsewhere, together with a discussion on the stratigraphic level of this specimen [4]. This paper presents more detailed information on the background of the find and some additional information concerning developmental stage of the owner of the tooth.

1

Excavation at the Bukuran Site

The Bukuran excavation site (Figure 1, 2) is located in the ESE area of the Sangiran Dome (7º27 ’23”S; 110º51 ’17”E), where the uppermost portion of the Sangiran (Pucangan) and most of the overlying Bapang (Kabuh) Formations are exposed. Several hominid fossils have been previously discovered from the Bapang Formation within a one km radius from this excavation site. Such previous finds include skull fragments (Sb 7904 a-d) [5], fragment of the right mandibular corpus with P4-M3 (Sb 8103) [6], and an unpublished small skull fragment. The excavation was conducted from June 10th to 28th with the help of 39 local inhabitants. Three stair-like trenches were excavated at two cliffs (Trench-I-1, -I-2, and –II) (Figure 1). The square grid was set 1.5 m by 1.5 m in size for the three trenches as shown in Figure 3. The depth was measured from the highest point of each trench. The total area and mass volume of the excavation amounted to 144 m2 (64 grids) and about 400 m3, respectively. The sandy part of the excavated earth was screened by dry screening with 5 mm mesh screens. The screening was done by local inhabitants under the supervision of our team staffs. About 1800 vertebrate fossils including Biography: Hisao BABA, 1968, graduated Faculty of Science, University of Tokyo; 1970, Master of Science (University of Tokyo); 1983, Doctor of Medical Science (University of Tokyo); 1995, Director of the Department of Anthropology, National Science Museum and Professor of the Department of Biological Sciences, University of Tokyo.

BABA et al.: Finding of a Hominid Lower Central Incisor During the 1997 Excavation in Sangiran, Central Java

47

small fragments were unearthed. Among these, 66 were found in situ. The taxa represented among this collection were typical of the Bapang Formation [7]. The vertebrates identified were Stegodon trigonocephalus, Axis lydekkeri, Cervus zwaani, Sus brachygnathus, Rhinoceros sondaicus, Bubalus palaeokerabau, Bibos palaeosondaicus, crocodile, and tortoise. Bs 9706 was found during the screening of the sediment from a medium to coarse grained sand layer with granules in one of the trenches (Trench-I-1) (Figure 4). No artifacts were found from the excavation.

Figure 1

Upper: Schematic plan view of the Bukuran excavation site. TR-I-1, -I-2, and –II are the trench number. H(79) indicates the location where an unpublished hominid cranial remain was found in 1979. Scales in meter

2

Stratigraphy and Chronology

Relative stratigraphic level of Bs 9706 within the Bapang Formation can not be precisely determined at the present stage of our research mainly because all the three key tuff layers of the Bapang Formation (Upper, Middle, and Lower Tuffs) [8] are missing owing to their more or less discontinuous character in the Bukuran excavated area. However, the stratigraphic level of Bs 9706 could be inferred with some reliability through correlation with the stratigraphy in the neighboring Bapang region, the type locality of the Bapan (Kabuh) Formation. Our analyses on paleo- and rock magnetism of several sediment samples from both the Bukuran excavation site and the Bapang region, as well as relative vertical position of the find spot of Bs 9706, strongly suggested that the stratigraphic level of Bs 9706 was between the Upper and Lower Tuffs of the Bapang Formation [4]. Thus far, there are no published radiometric dates for the Upper and Lower Tuffs. The magnetostratigraphic study by Hyodo et al. [9] identified the location of the Brunhes/Matuyama boundary (0.78 Ma [10]) as being just below the Upper Tuff. These same authors placed the top of the Jaramillo subchron (0.99 Ma [10]) just below the Grenzbank zone, which is situated at the lowermost part of the Bapang Formation. The above chronological data, thus, tentatively assign the Bs 9706 specimen to the latest phase of the Matuyama chron or to the latest Early Pleistocene.

48

ACTA ANTHROPOLOGICA SINICA

Figure 2

Figure 3

Excavation at Trench-I-1 and –I-2

Grid setting of the excavated trenches

Supplement to Vol. 19, 2000

BABA et al.: Finding of a Hominid Lower Central Incisor During the 1997 Excavation in Sangiran, Central Java

3

49

Affinity and Developmental Stage of Bs 9706

The specimen is a lower left central incisor (Figure 5). The crown of Bs 9706 preserves its original dimensions thanks to virtually complete preservation, minimal incisal wear, and probable absence of interproximal wear. The size, shape, and external morphology of the crown show affinity with Lower/Middle Pleistocene Homo erectus or Homo aff. H. erectus from Asia, Africa, and Europe, though it is somewhat smaller in its mesiodistal and labiolingual diameters. It is smaller compared to a small sample of Homo habilis and differs from australopithecines in the combination of high crown shape index, relatively low crown height, the possible presence of five mammelons, and a moderate degree of shoveling [4].

Figure 4

Lithological succession in the east wall of the Trench-I-1. Views of the Bs 9706 lower central incisor

Figure 5 Labial (left), distal (central left), lingual (central right), and incisal (right)

ACTA ANTHROPOLOGICA SINICA

50

Supplement to Vol. 19, 2000

The root formation of Bs 9706 is incomplete. Here we describe some features that permit insight into developmental stage of the owner of the tooth. The root of Bs 9706 is unusual in that its transverse diameters are markedly less than those at the cervical portion of the crown. This morphology and the existence of traces of surface erosion on both the crown and root surfaces indicate that the root had been disproportionately severely eroded relative to the damage on the crown. This extensive surface damage of the root makes us hesitate to estimate root formation stage of this specimen from its preserved outer morphology, but it may be inferred with some degree of reliability from the existing root height. Most of the published root heights of the lower central incisors of the known Plio-Pleistocene hominids are around 18 mm (Table 1). Therefore, if the height of the mature root of this specimen is accepted as having been approximately 18 mm, the labial height of the existing root of Bs 9706, 8.4 mm, would indicate that the root is approximately half complete. Table 1

Root heights of Plio-Pleistocene hominid lower central incisors (mm)

Locality/ specimen

Side

Root height

Measurement method

Reference

Homo erectus and Homo aff. H. erectus Zhoukoudian 5

L

>17.2

Labial height

[11]

Zhoukoudian 54 (BI)

R

>16.8?

Labial height

[11]

Zhoukoudian 55 (BI)

L

>16.8?

Labial height

[11]

Zhoukoudian 57 (BV)

R

18.0

Labial height

[11]

Zhoukoudian 58 (BV)

L

18.0

Labial height

[11]

Zhoukoudian 135’

R

>14.2

Labial height

[11]

KNM-WT 15000

R

20.4

No description

[12]

KNM-WT 15000

L

19.9

No description

[12]

Tighenif (isolated I1)

R

12.2*

No description

[13]

R

(18.0)

Lingual height

[14]

SKX 3559

R

17.2

No description

[15]

SKX 5004b

L

>17.7 (20.0)

No description

[15]

SKX 26967

L

>14.4

No description

[15]

SKX 35416

R

>16.1 (16.5)

No description

[15]

KNM-ER 3230

R

19.6

Labial height

[16]

KNM-ER 3230

L

21.5

Labial height

[16]

Australopithecus afarensis LH 14 Paranthropus robustus

Paranthropus boisei

* Calculated as (total tooth height) – (crown height).

There is a distinct worn facet along the incisal edge, but it is minimal in degree and restricted within the labiolingually narrow incisal surface. This incisal worn facet does not form a distinct angle with the long axis of the tooth in lateral view. Interproximal contact facets may well have been absent both on its mesial and distal surfaces. Numerous small erosional pits on the entire crown surface obscure the judgment, but at least distinct facets are not identified on the unrecorded portions of the interproximal surfaces of the specimen. Therefore, we infer that the individual died

BABA et al.: Finding of a Hominid Lower Central Incisor During the 1997 Excavation in Sangiran, Central Java

51

shortly after the accomplishment of preocclusal eruption of the permanent lower central incisors, and possibly during the eruption of the permanent lower lateral incisors. Acknowledgments: This study was conducted under the cooperation of the Geological Research and Development Centre, Bandung; Gadjah Mada University, Yogyakarta; National Science Museum, Tokyo; and Gunma Museum of Natural History, Gunma. We thank the many individuals of these institutions for their support. We also thank Mr. Slamet Sudjarwadi, Mr. Iwan Kurniawan (GRDC), and Mr. Hisao Kato (University of Tokyo) for their assistance in the field. Grant sponsor: Japanese Ministry of Education, Science, Sports and Culture; Grant number: 08041163. References: [1] HOOIJER DA. The geological age of Pithecanthropus, Meganthropus and Gigantopithecus [J]. Am J Phys Anthropol, 1951, 9: 265-281. [2] KOENIGSWALDT GHR. Australopithecus, Meganthropus and Ramapithecus [J]. J Hum Evol, 1973, 2(6): 487-491. [3] POPE GC, CRONIN JE. The Asian Hominidae [J]. J Hum Evol, 1984, 13(5): 377-396. [4] BABA H, AZIZ F, NARASAKI S et al. A new hominid incisor from Sangiran, Central Java [J]. J Hum Evol (in press). [5] BABA H, AZIZ F. Hominid skull fragments found in 1979 from Sangiran, Central Java [J]. Bull Natn Sci Mus, Series D, 1991, 17: 1-8. [6] AZIZ F, BABA H, NARASAKI S. Preliminary report on recent discoveries of fossil hominids from the Sangiran area, Jawa [J]. J Geol Min Resources, 1994, 29(4): 11-14. [7] AIMI M, AZIZ F. Vertebrate fossils from the Sangiran Dome, Mojokerto, Trinil and Sambungmachan area [A]. In: Quaternary Geology of the Hominid Fossil Bearing Formations in Java. Geological Research and Development Centre, Special Publication 4, 1985, 155-197. [8] ITIHARA M, SUDIJONO, KADAR D et al. Geology and stratigraphy of the Sangiran area [A]. In: Quaternary Geology of the Hominid Fossil Bearing Formations in Java. Geological Research and Development Centre, Special Publication 4, 1985, 11-43. [9] HYODO M, WATANABE N, SUNATA W et al. Magnetostratigraphy of hominid fossil bearing formations in Sangiran and Modjokerto, Java [J]. Anthropol Sci, 1993, 101(2): 157-186. [10] CANDE SC, KENT DV. Revised calibration of the geomagnetic polarity timesclae for the Late Cretaceous and Cenozoic [J]. J Geophys Res, 1995, 100(B4): 6093-6095. [11] WEIDENREICH F. Dentition of Sinanthropus pekinensis: A Comparative Odontography of the Hominids [M]. Paleontol Sin, New Series D 1, 1937. 1-180. [12] BROWN B, WALKER A. The dentition [A]. In: The Nariokotome Homo erectus Skeleton. Cambridge: Harvard University Press, 1993, 161-192. [13] ARAMBOURG C. Le gisement de Ternifine, II. L’Atlanthropus mauritanicus [A]. Archives de l’Institut de Paléontologie Humaine, Mémoire 32, 1963, 37-190. [14] WHITE TD. New fossil hominids from Laetolil, Tanzania [J]. Am J Phys Anthropol, 1977, 46(2): 197-230. [15] GRINE FE. New hominid fossils from the Swartkrans Formation (1979-1986 excavations): Craniodental specimens [J]. Am J Phys Anthropol, 1989, 79(4): 409-449. [16] WOOD, B. Koobi Fora Research Project: Volume 4. Hominid Cranial Remains [M]. Oxford: Clarendon Press, 1991.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

52-61

Thickness Mapping of the Occipital Bone on CT-data –- a New Approach Applied on OH 9 Gerhard W. WEBER1, Johann KIM2, Arnold NEUMAIER2, Cassian C. MAGORI3, Charles B. SAANANE3, Wolfgang RECHEIS4, Horst SEIDLER1 (1. Institute for Anthropology, University of Vienna, Althanstr. 14, A-1090 Vienna. AUSTRIA; 2. Institute of Mathematics, University of Vienna, Strudlhofgasse 4, A-1090 Vienna, AUSTRIA; 3. Dept. of Anatomy & Histology, Muhimbili University College of Health Science, P.O. Box 65453, Dar-es Salaam, TANZANIA; 4. Dept. of Radiology II, University Hospital Innsbruck, Anichstr. 35, A-6020 Innsbruck, AUSTRIA)

Abstract A new approach for the analysis of cranial bone thickness is introduced. The study focuses on the occipital bone of modern humans and of a 1.25 Myr-old H. ergaster/erectus specimen from Olduvai Gorge (OH 9). A semiautomatic algorithm detects a multitude of thicknesses from CT-data of the investigated bones. We find that every bone is characterized by its own distribution pattern of cranial thickness, which is then analyzed statistically. The results demonstrate that the thickness distribution of the occipital bone of OH 9 is within the normal range of the H. sapiens sample (which itself shows a remarkably high variance). This contributes to a further analysis of phyletic differences of hominid morphology by including distribution patterns of thickness combined with aspects of functional anatomy.

Key words: Occipital bone; Bone thickness; Hominid evolution; Virtual anthropology; Computed tomography

1

Introduction

Information about the thickness of cranial bones are not only of great medical interest, particularly for preoperative surgical planning [1], but can be as useful for investigations of fossil hominid material [2]. However, not much data is available for both these disciplines. Intra- and interspecific variation [3-4] is poorly known and mostly depends on measurements taken on a handful of landmarks or other not well defined points. These studies fall short of offering adequate information about the structural details of skulls, especially when considering specific endo- and exocranial qualities. Other authors [5-6] have undertaken efforts to develop thickness maps of the cranial vault but again with the restriction of a very limited number of measuring points. We demonstrate a new approach, using CT-data of modern Homo sapiens skulls and of a skull of Homo ergaster/erectus from Tanzania, namely the 1.25 Myr-old OH 9. As the method is still under development we had to limit the formulation of the problem and so we are focusing on the occipital bone only. The points we use for the analysis are defined by the resolution of the CT-scan (0.4 mm). From every point on the surface of the bone we can measure the occipital thickness by using a new algorithm and the results can be given as a matrix as well as a topographical thickness map of the investigated bone. The method is part of our “Virtual Anthropology” project [7-9] concerned with the 3D-analysis of CT-data of recent and fossil hominids. The aim of the present study is to obtain a sufficient number of thickness measurements by a semiautomatic process to characterize the thickness distribution and to analyze the morphology variation of phyletic species.

2

Material

Our sample consists of 12 H. sapiens occipital bones (Table 1) and we tried to get as much variation as possible by including skulls from different populations (4 Europeans, 3 Africans, 3 Asians, 2 Australiens) of different sex (5 female, 7 males). Our first candidate of a fossil specimen to be compared was the occipital bone of the Tanzanian H. ergaster/erectus OH 9.

WEBER et al.: Thickness Mapping of the Occipital Bone on CT-data - a New Approach Applied on OH 9

Table 1 Skull

Sex

VA 1

53

Individuals in the study

Age

Origin

Female

25

Europe

335

VA 2

Male

45

Europe

LE 54

VA 3

Male

25

Europe

NL 320

VA 4

Female

30

Europe

NL 371, HY

VA 13

Female

20

Australia

Aboriginal, C.80

VA 20

Male

40

Australia

Aboriginal, C.70

VA 23

Male

50

Africa

Bantu, S 66

VA 24

Female

20

Africa

Bantu, S 157

VA 25

Male

30

Africa

Bushman, S 60

VA 26

Male

35

Asia

Chinese, 2587

VA 27

Male

35

Asia

Chinese, 2576

VA 28

Female

30

Asia

Chinese, 2584

OH 9

-

-

Africa

Tanzania

3

Notes

Methods

3.1

Data The occipital bones were isolated from the CT-data sets of the complete skulls with the software package ANALYZE AVW in such a way that the cut was roughly orthogonal to the surface along the sutures of the occipital bone. To avoid thickness artefacts of the occipital condyles, we decided to cut cross the foramen magnum approximately in the region of the condylarian canals. The data was then thresholded according to the HMH-standard method [10] and exported for the further mathematical procedure into a binary three dimensional matrix. Subsequent computations were performed in MATLAB 5.2.

Figure 1

Computation of thickness on CT-scans of the occipital bone (here one slice); thickness measurements are not symmetric (blue: from outside to inside; yellow: from inside to outside)

ACTA ANTHROPOLOGICA SINICA

54

Supplement to Vol. 19, 2000

3.2

Computation of the thickness As many thicknesses as possible of the occipital bone were measured. For computation purposes mathematical definitions of biological categories included reductions. An occipital bone is a connected and bounded subset of the three dimensional space

OCC ⊆ ∇3 The complementary set

∇3 \ OCC is called outside. A surface point (SP) is an element of OCC and has a border to outside. An occipital bone has a surface S which is the set of all SP. Defining the bone by its interior and exterior surface is a first reduction because the cutting surface of the occipital bone is not considered. For the definition of the surfaces we need to introduce the observer’s viewpoint which can be chosen as the center of gravity within the skull. A surface point is called interior surface point (ISP) if the ray from the viewpoint to SP is a subset of the outside, i.e. no further point of OCC lies between the viewpoint and ISP. The interior surface (IS) is the set of all ISP. The exterior surface (ES) is the complementary set.

S \ IS i.e. an exterior surface point (ESP) is a SP which is not an ISP. The thickness d(ISP) at an ISP is the minimal distance from the ISP to ES.

d ( ISP) = min ∆( ISP, ESP ) ESP∈ES

The measurement of thickness is not symmetric because, in general, the thickness measured from the interior surface is not the same as that measured from the exterior surface.

∆(ISP,ESP) ≠ ∆(ESP,ISP) In practice, the data acquisition by CT-scanning includes a discretization of the set OCC from all points to a finite subset because OCC is approximated by voxels. A voxel is a volume element with x/y/z dimensions, comparable to pixels in 2D. Each voxel is represented by a triple of integers, so OCC is then a subset of ∧3, defining a discretized surface (DS). On an intact occipital bone of a H. sapiens approximately 100,000 points remain on the discretized interior surface (DIS). The computational costs for this number of points are still too high, so a reduction in the cardinality of DIS is necessary, ensuring no loss in geometric information. The reduction of the DIS is obtained by a stepwise refinement of a starting triangulation. 1. A set of arbitrary points (lattice points) is initialized and the lattice is triangulated. 2. The edge is divided into two parts if the edge is longer than a given mesh size specified. 3. Step two is repeated until every edge is smaller than the mesh size.

WEBER et al.: Thickness Mapping of the Occipital Bone on CT-data - a New Approach Applied on OH 9

55

Figure 2

3D-reconstruction of the CT-data on the computer screen and scheme of triangulation procedure to further reduce the discretized interior surface (DIS)

Figure 3

Thickness optimization procedure leads to the minimal distance between an interior surface point and its corresponding exterior surface point

The remaining points of DIS are called the reduced interior surface points (RISP). In order to compute the thickness over a RISP it is now necessary to find its corresponding reduced exterior surface point (RESP).

ACTA ANTHROPOLOGICA SINICA

56

Supplement to Vol. 19, 2000

1. A first approximation of RESP (pRESP – provisional reduced exterior surface point) is found as the intersecting point of the discretized exterior surface (DES) and the line through the viewpoint and RISP. 2. pRESP is the starting point for a local optimization process which leads to the oRESP (optimal reduced exterior surface point). Starting with pRESP we find the adjacent neighbor DESP (not RESP) with the smallest distance to RISP. pRESP is overwritten by the neighbor DESP if the distance between RISP and DESP is smaller than the distance between RISP and pRESP. 3. Step two is repeated until no better DESP is found.

4

Results

In a first run, using a resolution of four voxels (thus we have a measuring point at least every 1.6 mm) the semiautomatic thickness measuring algorithm resulted in a number of measurements between 770 and 1,973 per occipital bone. Later, the number will be increased to 5,000 – 40,000 points.

Figure 4

OH 9 lies within the range of variation of Homo sapiens (distribution of the normalized thicknesses)

For statistical analysis, the thickness distribution was approximated by a histogramm with 13 classes, ranging from 0 to 26 mm. To compensate for the different number of individual measurements on the bone, we normalized the data by computing the corresponding percentage for each class. The distributions of the normalized data are surprisingly heterogeneous among the moderns, but more surprising is the fact that OH 9 lies within this range. If parameters like the mean thickness, the median or the maximum of thickness are considered, OH 9 turns out to be on the upper end of the general distribution but has not one extreme value. VA 20, a male individual from Australia, and VA 23, a male individual from Africa, have higher means than OH 9. VA 1, VA 20, VA 23, VA 24, VA 25, VA 26, VA 27and VA 28 have higher maxima than does OH 9. Among these skulls are also female specimens (Table 2).

WEBER et al.: Thickness Mapping of the Occipital Bone on CT-data - a New Approach Applied on OH 9

Table 2

57

Thickness of occipital bone

Skull

N

Mean

Std.dev.

Median

Min.

Max.

VA 1

1407

6.35

3.39

5.67

1.59

22.35

VA 2

770

4.34

2.14

4.13

1.49

15.94

VA 3

1513

5.37

1.87

5.13

1.59

12.62

VA 4

1412

5.73

2.27

6.01

1.49

14.14

VA 13

1160

4.53

1.73

4.47

1.59

11.46

VA 20

1350

8.52

3.18

8.70

1.82

20.75

VA 23

1311

9.49

3.36

9.14

1.82

24.22

VA 24

1383

6.97

3.25

6.19

1.67

22.30

VA 25

1286

6.38

3.43

5.78

1.49

19.46

VA 26

1298

7.01

3.00

6.92

1.89

20.39

VA 27

1440

6.01

3.08

5.58

1.67

20.62

VA 28

1101

7.03

3.62

6.64

1.89

23.78

OH 9

1973

8.16

3.30

8.23

1.49

18.70

The plot of the distributions of accumulated percentages (Fig. 4) makes it clear that the distribution of OH 9 is somewhat special compared with most H. sapiens skulls but two modern skulls show more deviation from these group than OH 9 does. In a factor analysis (principal component) where all the relative frequencies in the 13 classes of thickness are variables, two major components which explain more than 88 % of the variance are extracted. Component one has high loads for the thickness classes 1, 2, 3, 5, 6 and 7 (0 – 6 mm, 8 – 14 mm), component two for 4, 5 and 6 (6 – 12 mm). The scatterplot of the two regression factor scores (Fig. 5) again shows that OH 9 lies well within the normal variation of our H. sapiens sample. Only VA 4 is remarkably distant to the cubic interpolation function plotted. A possible explanation for this is that VA 4 shows pathological changes on both the cranial base and the occipital condyles. The occipital of this female is relatively thin but has a very steep increase in thickness in the region of the internal occipital crest.

5 5.1

Discussion

Problems As the data needs to be thresholded for accurate measurements to the average HMH values, small empty holes are created within the bone in the regions where the density of the spongious bone is very low (Figure 6). Our algorithm is therefore designed to cope with this problem and was robust enough. The position of the viewpoint to find the pRESP’s can affect the results if the position is too far from the optimum. In a simulation with geometric objects we found a difference in mean thickness of around 8% for extreme positions. This problem only partly affects our results because the viewpoint was under our control and could always be placed in an acceptable location. Choosing different reasonable viewpoints and repeating the algorithm reduces the variability to an acceptably low value.

ACTA ANTHROPOLOGICA SINICA

58

< CASE Figure 5

Supplement to Vol. 19, 2000

 Total population

First two major components of the PCA explaining more than 88% of the variance. OH 9 fits well to the cubic interpolation function. The most distant individual VA 4 (H. sapiens) is a pathological case

All lattice points of IS are within the actual margin, introducing a slight error, depending on how close to the margin the furthest lattice point is. Improvement by subdivision of lattice spacing solves this problem, but at the expense of increasing the number of lattice points in the set IS. The distribution analysis is sufficiently robust so that we feel confident in our results. If the occipital bone is isolated and scanned, the CT-scan of the edges is highly serrated. As the triangulation process is restricted to a certain resolution, edge details are lost (Fig. 6, white points have no coordinates and are disregarded). 5.2

Conclusion Being thought to be more massive than those of modern humans, the occipital bone of OH 9 demonstrates the relativity of this statement in a quantitative analysis with a sufficient number of thickness measurements. It is not enough to measure a few thicknesses to characterize an occipital bone adequately. It is more important to analyze the thickness distribution and to detect the peculiarities of this distribution in comparison with others. The topographical thickness plots (Fig. 7) of our measurements show some subtle structural differences. The occipital bone thickness of OH 9 seems to vary more regularly yet shows patterns similar to H. sapiens occipital bones if the occipital torus (thickened) or the cerebellar fossae (thinned) are included. Analysis of functional anatomy characterized by thickness distributions in defined regions should distinguish between different

WEBER et al.: Thickness Mapping of the Occipital Bone on CT-data - a New Approach Applied on OH 9

59

hominid morphologies. The high variance among H. sapiens occipital bone thickness and the similarity to a H. ergaster/erectus suggests that occipital bone thickness was not a feature under selection pressure within the last million years.

Figure 6

Small holes are created by the thresholding process in regions where the density of the spongious bone is low (above). The thickness algorithm is designed to cope with this problem. Some edge details are lost because the resolution of the triangulation process is limited (below). Using an appropriate meshsize, this effect hardly affects the distribution of thickness measurements

ACTA ANTHROPOLOGICA SINICA

60

Figure 7

Supplement to Vol. 19, 2000

Topographical thickness plots of the 12 H. sapiens occipital bones and the one of OH 9; thin bone is dark gray, thick bone is white. Thickness distribution in defined regions can be analyzed

WEBER et al.: Thickness Mapping of the Occipital Bone on CT-data - a New Approach Applied on OH 9

61

References: [1] ELAHI MM, LESSARD ML, HAKIM S et al. Ultrasound in the assessment of cranial bone thickness [J]. J Craniofac. Surg, 1997, 8: 213-221. [2] GAULD SC. Allometric patterns of cranial bone thickness in fossil hominids [J]. Am. J Phys Anthropol, 1996, 100: 411426. [3] LIEBERMAN DE. How and why humans grow thin skulls: experimental evidence for systemic cortical robusticity [J]. Am J Phys Anthropol, 1996, 101: 217-236. [4] ZIPNICK RI, MEROLA AA, GORUP J et al. Occipital morphology. An anatomic guide to internal fixation [J]. Spine, 1996, 21: 1719-1724. [5] ROSS AH, JANTZ RL, MCCORMICK WF. Cranial thickness in American females and males [J]. J Forensic Sci, 1998, 43(2): 267-272. [6] HWANG K, KIM JH, BAIK SH. Thickness map of parietal bone in korean adults [J]. J Craniofac Surg, 1997, 8: 208-212. [7] WEBER GW, RECHEIS W, SCHOLZE T et al. Virtual anthropology (VA): Methodological aspects of linear and volume measurements - First results [J]. Coll Antropol, 1998, 22: 575-583. [8] CONROY GC, WEBER GW, SEIDLER H et al. Endocranial capacity in an early hominid cranium from Sterkfontein, South Africa [J]. Science, 1998, 280: 1730-1731. [9] BOOKSTEIN F, SCHAEFER K, PROSSINGER H et al. Comparing frontal cranial profiles in archaic and modern Homo by morphometric analysis [J]. The Anatomical Record, 1999, 257: 217-224. [10] Spoor CF, Zonneveld FW, Macho GA. Linear measurements of cortical bone and dental by computed tomography: Applications and problems [J]. Am J Phys Anthropol, 1993, 91: 469-484.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

62-68

The Period of Transition between Homo erectus and Homo sapiens in East and Southeast Asia: New Perspectives by the Way of Geometric Morphometrics Florent DETROIT (Laboratoire de Prehistoire du Muséum National d’Histoire Naturelle, Institut de Paleontologie Humaine, UMR 6569 du CNRS, 1, rue R. Panhard, 75013 PARIS FRANCE)

Abstract Anatomically modern humans origin is one of the most passionately debated questions of the moment. The objective of this work, using a new methodology (3D geometric morphometrics) for studying human cranial shape evolution, is to shed some light on the evolution of the genus Homo in East and Southeast Asia, with African fossils for comparisons. This is a region of major palaeoanthropological interest due to the large number of fossil remains discovered beginning at the end of the 19th century. This part of the Old World is a key geographical area as far as the debate on the origin of anatomically modern humans is concerned. We present here our first results of architectural comparisons of human skulls dating from about 2.5-2 Ma to present days (from H. habilis to extant H. sapiens). We attempt to underline the main architectural differences between Homo erectus and Homo sapiens, but also to analyse eventual geographical and/or temporal intraspecific architectural variability.

Keywords:

Modern humans origin; Homo erectus; Homo sapiens; Cranial vault; 3D geometric morphometrics; Procrustes analysis; East Asia; Southeast Asia

1

Introduction

1.1

The Origin of modern humans Two major conflicting evolutionary models are classically proposed to explain the evolution from Homo erectus to Homo sapiens. The replacement - or “Out of Africa” - model [1-3] considers that anatomically modern Homo sapiens appeared first in Africa, at the end of the Middle Pleistocene. They subsequently migrated to the Middle East and Asia about 100 ky ago and into Europe and Australia 50 ky later. During these migrations, modern humans totally replaced archaic populations without any interbreeding. This model appears to be strongly supported by the majority of recent molecular studies [4-6]. The multiregional model [7-11], inspired from Weidenreich’s work, is essentially based on the observation that, to some extent, Australian Upper Pleistocene Homo sapiens cranial morphology is inherited from Javanese Homo erectus while the Chinese human fossil record would tend to show an in situ evolution from early Homo erectus to recent Homo sapiens [12-16]. After the dispersal of Homo erectus outside Africa, about 1.5 My ago, an uninterrupted gene flow allowed the gradual and continuous evolution of Homo erectus into Homo sapiens in all the regions, without population replacement. This gene flow was stronger within than between the main regions. Consequently, this heterogeneous gene flow explains the persistence of regional phenotypic differences observed on human fossils during 1.5 My [8]. However, some scholars argue in favour of less extreme models [17-18]. They propose the possibility of replacement in Europe (Neandertals replaced by modern Homo sapiens) and at the same time of evolutionary continuity in Asia. This evolutionary continuity could be the result of interbreeding between in situ archaic populations in place (“late Homo erectus”) and the newly evolved Homo sapiens migrants from Africa. 1.2

Comparisons of human crania In palaeontological studies, two main approaches based on fossils studies are traditionally used to deal with evolutionary problems. The first one is phylogenetic (mainly cladistic) and is based on characters state analyses to construct parsimonious evolutionary trees. The second one is morphometric: with a more phenotypic approach, fossil or present skeletal parts are statistically

DETROIT: The Period of Transition between Homo erectus and Homo sapiens in East and Southeast Asia: New Perspectives by the Way of Geometric Morphometrics

63

compared to make groups of morphologically more related specimens in order to demonstrate and interpret morphological and/or morphometrical changes in the course of evolution. For years, this second approach, mainly based on analyses of linear measurements (“traditional morphometrics”), has been used in palaeoanthropoloy to assess regional or temporal differences from human cranial morphologies. But interpretations of such analyses are often very complex and unable to account for slight variations of the geometry of anatomical parts such as human skulls. Geometrical morphometrics in three dimensions, which considers shape as a whole and uses coordinates of homologous landmarks, appears to be appropriate for the shape analysis of such a complex structure [19-20]. Procrustean landmarks superimpositions allow accurate architectural comparisons [21-22]. Separating shape from size during the study provides the possibility to investigate relationships between size and shape (allometry) and prevents the confusion arising from size and shape differences in the analysis. We present here our first results of architectural comparisons of human skulls dating from about 2.5-2 Ma to present days (from H. habilis to extant H. sapiens). We compare the configurations of the skull cap, the best documented portion of fossil skeletons. It is also, without any doubt, the most studied human anatomical part with the other methodologies described above. This is an important point for testing and sometimes guiding some of the methodological choices we made during this study.

2

Material and methods

Indonesian Homo erectus from Sangiran (casts) and Ngandong (originals) and Chinese Homo erectus from Zhoukoudian are compared with fossil Homo sapiens from North Vietnam (Lang Cuom: originals), from Indonesia (Wajak, Liang Momer, Gua Nempong: originals) from China (Zhoukoudian Upper Cave) and from Australia (Cohuna). We also include comparative individuals in this sample, in particular African Homo habilis and Homo erectus / Homo ergaster and extant Homo sapiens: three extant Homo sapiens have been randomly chosen and three have been chosen for their geographical origin (Australia, China & Java). 2.1

Cranial landmarks As the skull cap is very often the only preserved part of fossil skulls, we focus on this portion for architectural comparisons. As landmarks we selected classical craniometric points that modelise the global architecture as well as possible. Twenty seven landmarks were digitized on each skull (Figure 1) using the Microscribe® 3DX digitizing arm. Seven landmarks are sagittal points and twenty landmarks are parasagittal points (ten on each side). Following the classification of Bookstein, the majority of the landmarks belongs to the type I (intersection points) and type II (maxima of curvature). They are relatively easy to localise precisely on fossil or extant skulls. Only euryon (n°16) and coronion (n°17) could be considered as constructed points (type III). 2.2

Procrustean superimpositions For each of both analyses we present below, a Generalized Least Squares fitting [23-24] was computed using the GRF-ND and Morpheus et al. software [25]. All the skulls are superimposed on their respective centroïds after what they are scaled down and rotated for the least squares fitting. Then the mean reference configuration for the whole sample is calculated. Superimposed coordinates were analysed by Principal Component analysis using the SAS software, version 6.11. Along the axis of the principal component analysis, we illustrated architectural variability with the presentation of the extreme configurations (extreme individuals) compared to the mean reference configuration.

ACTA ANTHROPOLOGICA SINICA

64

Figure 1

Supplement to Vol. 19, 2000

Location of the cranial landmarks and visualisation of the cranial modelisations obtained. Only the left half skulls are statistically studied (17 landmarks)

2.3

Damaged or missing anatomical parts The entire skull cap were digitized but only the left half skulls (seventeen landmarks) were statistically studied. Very often, when studying fossil skulls, points are missing only on one side. So we use here a method to reconstruct a mean half configuration for each fossil. This mean configuration is the mean between the right and the left half skull after the landmarks procrustean superimposition. When one point is missing on the left but is present on the right, it seems to be the more rigorous way to reconstruct this right point on the left side.

3

Results

The first analysis is the principal component analysis of the coordinates of the seventeen superimposed landmarks (Figure 2). The number of individuals in this analysis is very low because some of the points are missing on the left side and on the right side. So, incomplete fossils are automatically excluded from this analysis. There is some interesting observations to make onto the plane 1-2. The axis 1 which explain 53% of the total variance clearly separate both fossil and present Homo sapiens from older African fossils (KNMER 3733, 1813 & Broken Hill). On the left side along axis 1, we found individuals with very low brain cases: in proportion, the bregma and the euryon are very low, the occipital bone is angulated, the supraorbital torus is large, horizontal and straight above the orbite and behind this broad supraorbital torus, the frontal bone is very narrow. On the opposite side, the cranial architecture is more rounded: bregma and euryon are proportionally higher, the glabella is no more projecting; from the frontal view, the supraorbital complex is vertival and rounded above the orbite continuing into a wide and rounded frontal bone. Along this axis Broken Hill is clearly grouped

DETROIT: The Period of Transition between Homo erectus and Homo sapiens in East and Southeast Asia: New Perspectives by the Way of Geometric Morphometrics

65

with the Homo erectus (Homo ergaster?) KNMER 3733 and even the Homo habilis KNMER 1813 which is surprisingly very close to the African Homo erectus morphology when the size parameter is eliminated.

Figure 2

PCA of the coordinates of the 17 superimposed landmarks, axes 1 & 2 ; plot of individuals and extreme configurations on each axis, dotted lines correspond to the mean reference configuration

Along the second axis, which explain 13.6% of the total variance, the upper Pleistocene Australian fossil Cohuna appears on an extreme position. When analysing the configurations, the variability in the anteroposterior proportions of the frontal bone is clearly the dominant factor along this axis. On the top, we found the Lang Cuom 12 individual which exhibit a proportionally very short frontal bone. On the bottom, Cohuna exhibit a very particular frontal bone architecture. This frontal is proportionnaly very long and this is particularly true for the frontotemporal, the stephanion and the coronion which are in very posterior positions. In this analysis, we have the confirmation that the cranial morphology of the Cohuna cranium is abnormal certainly due to an artificial deformation as some scholars argued. In this perspective, it should have been very interesting to include in this analysis the Upper Cave 102 specimen that also look artificially deformed. But the cast we have exhibit severe post mortem damages so it was impossible to include it. For the following analysis, due to its abnormality we decided to exclude the Cohuna cranium from the calculations. The second analysis is the principal component analysis of the coordinates of only 11 superimposed landmarks (Figure 3). We eliminated some landmarks that are absent on many fossils and we kept landmarks involved in the modelisation of the sagittal outline, the frontal bone and the euryon.

ACTA ANTHROPOLOGICA SINICA

66

Figure 3

Supplement to Vol. 19, 2000

PCA of the coordinates of 11 superimposed landmarks, axes 1 & 2 ; plot of individuals and extreme configurations on each axis, dotted lines correspond to the mean reference configuration

Onto the plane 1-2, we have three groups. Along the first axis, which explain 57.7% of the total variance, we find again two groups really distinct from each other. As in the first analysis, individuals on the left side of the first axis, exhibit low brain cases (low bregma and low euryon in proportion) with broad and straight supraorbital torus whereas on the opposite side the cranial morphology is more rounded with a very weak supraorbital complex continuing into a wide and rounded frontal bone. Along this axis fossil and present Homo sapiens form a group clearly opposed to the Homo erectus (sensu lato) – Homo habilis group (including also Broken Hill) without any overlap. It is impossible to distinguish along the axis 1 any geographical or temporal trends inside one of both groups. On the other hand, the second axis which explain 9.9% of the total variance cut the so-called Homo erectus – Homo habilis group in two parts. The main architectural differences along the axis 2 appear to be localised in the frontal bone morphology. On the bottom of the axis, we find an exclusively African group (KNMER 1813, 3733, OH9 and Broken Hill) exhibiting strong and wide supraorbital torus associated with relatively narrow frontal bone. Whereas on the top of the axis individuals exhibit moderately developed supraorbital torus associated with relatively broad frontal bone: the postorbital constriction is reduced. In this second Homo erectus group there are Indonesian, Chinese as well as African specimens. It seems not possible to evoke a temporal evolutionary trend nor a sexual dimorphism because of the great diversity exhibited by both subgroups. Adding new specimens in the sample may help us to explain this dichotomy if confirmed. Observation of individuals plots onto plane 1-3 emphasis once again the clear architectural distinction between Homo sapiens and Homo erectus. The axis 3 which explain 8.5% of the total variance shows the variability of the anteroposterior proportion of the frontal bone. On the top of the axis, individuals such as Ngandong 6 exhibit very long and flat frontal bone in proportion with the overall skull whereas on the bottom, specimens like Lang Cuom 12 exhibit very short frontal bones. There is no particular trend along this axis in the Homo erectus group nor in the Homo sapiens one.

DETROIT: The Period of Transition between Homo erectus and Homo sapiens in East and Southeast Asia: New Perspectives by the Way of Geometric Morphometrics

4

67

Discussion

The methodology adopted - procrustean landmarks superposition (i.e. GLS) - seems to be very well suited for accurate cranial architectures comparisons. Interpretations of morphological variability during the course of evolution are easier and more understandable in terms of cranial geometry and separating size parameters from the shape information allows new observations. For example we have seen that KNMER 1813 which is a very small skull is architecturally very close to African and other Homo erectus / Homo ergaster. A global increase in size could be a sufficient phenomenon to evolve from an Homo habilis to an Homo erectus cranial vault architecture. Concerning the Australian Upper Pleistocene Homo sapiens cranial morphology that could be inherited from javanese Homo erectus, we pointed out without any doubt the abnormal cranial architecture of the Cohuna cranium. It is a very unusual Homo sapiens morphology but it is clearly not an Homo erectus nor an intermediate shape. Finally, we always have a very clear distinction between Homo erectus and Homo sapiens architectures. We never observed any overlap between these two groups even for specimens like Broken Hill. Broken Hill exhibit clearly an Homo erectus-like cranial architecture. We never observed any trend concerning geographical groups: Sinanthropus are never more closely to Chinese Homo sapiens than to any other fossil or extant Homo sapiens, nor are Ngandong specimens with Wajak 1 or other Indonesian or Australian Homo sapiens. Other analyses of different sets of landmarks always confirm this clear distinction: nothing seems to indicate any interbreeding between Homo erectus and Homo sapiens in East and Southeast Asia. Related with this clear separation between the erectus - shape and the sapiens - shape, we pointed here the fact that the Ngandong specimens, although sometimes considered as archaic Homo sapiens, exhibit a typical Homo erectus architecture. However our sampling is not exhaustive and we could not test in this work the position of the so-called transitional Chinese fossils. But we have a very interesting frame for future work and we intend now to include in our sample Chinese specimens like Dali, Jinniushan and Maba; new discovered Homo sapiens fossils from Indonesia and Thailand and Upper Pleistocene Australian fossils to go more precisely in the study of cranial shape evolution in East and Southeast Asia since the Middle Pleistocene. Acknowledgements: I would like to thank Prof. Teuku Jacob (Gadjah Mada University, Yogyakarta) and Dr. John de Vos (Nationaal Natuurhistorisch Museum, Leiden) for access to some of the original fossils included in this study. Dr Dominique Grimaud - Hervé and Dr Miguel Caparros are gratefully acknowledged for their help with this work. References: [1] STRINGER CB. The origin of early modern humans: A comparison of the European and non-European evidence [A]. In: BR ÄUER G, SMITH FH eds. Continuity or Replacement? Controversies in Homo sapiens Evolution. Rotterdam: Balkema, 1989. [2] STRINGER CB. Reconstructing Recent Human Evolution [A]. In: AITKEN MJ, STRINGER CB, MELLARS PA eds. The Origin of Modern Humans and the Impact of Chronometric Dating. Princeton University Press, 1993, 179-95. [3] STRINGER CB, ANDREWS P. Genetic and fossil evidence for the origin of modern humans [J]. Science, 1988, 239:1263-1268. [4] CANN RL, STONEKING M, WILSON AC. Mitochondrial DNA and human evolution [J]. Nature, 1987, 325: 31.36. [5] RELETHFORD JH, HARPENDING HC. Craniometric variation, genetic theory, and modern human origins [J]. Am J Phys Anthropol, 1994, 95:249-270. [6] TISHKOFF SA, DIETZSCH, E, SPEED W et al. Global patterns of linkage disequilibrium at the CD4 locus and modern human origins [J]. Science, 1996, 271:1380-1387. [7] THORNE AG, WOLPOFF MH. Regional continuity in Australasian Pleistocene hominid evolution [J]. Am J Phys Anthropol, 1981, 55:337-349. [8] WOLPOFF MH. Multiregional evolution: The fossil alternative to Eden [A]. In: Mellars P, Stringer CB eds. The Human Revolution. Edinburgh: Edinburgh University Press, 1989, 62-108.

68

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

[9] WOLPOFF MH, WU XZ, THORNE AG. Modern Homo sapiens origins: A general theory of hominid evolution involving the fossil evidence from East Asia [A]. In: SMITH FH, SPENCER F eds. The Origin of Modern Humans: A World Survey of the Fossil Evidence. New York: Alan R. Liss, 1984, 411-483. [10] WOLPOFF MH, SPUHLER JN, SMITH FH et al. Modern Human Origins [J]. Science, 1988, 241:772-773. [11] WOLPOFF MH, THORNE AG, SMITH FH et al. Multiregional evolution: A world-wide source for modern human populations [A]. In: NITECKI MH, NITECKI DV eds. Origins of Anatomically Modern Humans. New York: Plenum Press, 1994, 175-199. [12] WU X. The evolution of Humankind in China [J]. Acta Anthropol Sin, 1990, 9:320-321. [13] WU X. Continuité évolutive des homme fossiles chinois [A]. In: Hublin JJ, Tillier AM eds. Aux origines d'Homo sapiens. PUF. 1991, 157-179. [14] WU X, POIRIER FE. Human Evolution in China. New York: Oxford University Press, 1995, 317. [15] ETLER DA. The Chinese Hominidae: New Finds, New Interpretations [D]. Ph D in Anthropology, Graduate Division of the University of California, Berkeley. 1994, 471. [16] LI T, ETLER DA. New Middle Pleistocene hominid crania from Yunxian in China [J]. Nature, 1992, 357:404-407. [17] BR ÄUER G. The evolution of modern humans: A comparison of the African and non-African evidence [A]. In: Mellars P, Stringer CB eds. The Human Revolution. Edinburgh: Edinburgh University Press, 1989, 123-154. [18] BR ÄUER G. The occurrence of some controversial Homo erectus cranial feature in the Zhoukoudian and east African hominids [J]. Acta Anthropol Sin, 1990, 9:352-358. [19] ROHLF J, MARCUS L. A revolution in morphometrics [J]. Trends Ecol Evol, 1993, 8 (4):129-132. [20] BOOKSTEIN FL. Combining the tools of geometric morphometrics [A]. In: MARCUS LF, CORTI M, LOY A et al. eds. Advances in Morphometrics. New York: Plenum Press, 1996, 131-151. [21] BOOKSTEIN FL, SCH ÄFER K, PROSSINGER H et al. Comparing frontal cranial profiles in archaic and modern Homo by morphometric analysis [J]. The Anatomical Record (New Anat), 1999, 257:217-224. [22] PENIN X. Modélisation tridimensionnelle des variations morphologiques du complexe cranio-facial des Hominoïdea. Application à la croissance et à l'évolution [D] . Thèse de Doctorat Université Paris VI, Spécialité Sciences de la Terre. Paris, 1997. [23] DRYDEN IL, MARDIA KV. Statistical shape analysis [M]. Chichester, UK: Wiley, 1998. [24] BAYLAC M. Morphométrie géométrique et systématique [J]. Biosystema: Systématique et Informatique, 1996, 14:7389. [25] SLICE DE. 1994-2000 Software for morphometric research (GRF-ND, Morpheus et al.) available from http://life.bio.sunysb.edu/morph/.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

69-76

Neural Tube, Spheno-occipital Flexion and Semi-circular Canals in Modern and Fossil Hominids Anne DAMBRICOURT MALASSE1, Jean Pascal MARTIN 2, Eric de KERVILER 3 (1. UMR 6569 CNRS, Laboratoire d’Anthropologie, Université Aix-Marseille II, Institut de Paleontologie Humaine, 1 rue Rene Panhard 75013 Paris, France; 2. Hopital Saint Louis, Service de Chirurgie Maxillo-Faciale, Paris, France ; 3. Hopital Saint Louis, Service de Radiologie, Paris, France)

Abstract The position of the cartilaginous occipital, the sphenoidal angle, the frontalization of the petrous bone and the orientation of the foramen magnum, are major characteristics of basic-cranio-facial evolution in hominids. It has become standard to associate them and to notice their evolution or spatial organization according to the degree of spheno-occipital flexion. A study by scanning an ontogenetic series of Pan troglodytes and modern man's cranium, gives the first glimpse, with an application to complet fossilized craniums, or fragments of isolated petrous bone, such as Qafzeh (Homo sapiens), La Ferrassie and La-Chapelle-aux- Saints (Homo neanderthalensis), Sambugmacan and Solo 1 (Homo erectus). Paleontological and modern data indicate three groupings of around a same flexion: Great Apes, fossil Homo and Modern Man.

Key words: Computed tomography; Embryogeny; Hominids; Neural tube; Semi-circular canals; Sphenoidal flexion The semi-circular canals are intra-cranial structures only accessible via radiography and better via Computed Tomography (CT scan). They each occupy a plane in space and form a right angle with the two other canals (figure 3). Recent studies conducted by Spoor, Zonneveld and Wind [1-2] showed angular divergences between adult Pan, Australopithecus and Homo sapiens. The lateral canal of Homo sapiens is further frontalized in relation to the posterior than in Pan and Australopithecus. This divergence has not been explained although it represents a major criteria of distinction between Australopithecus and Homo. This angular opening reaches the angular differences which have been noticed between cranial basis. Numerous works, such as those by Delattre and Fenart in the 1960s, have showed that the Homo sapiens cranium distingues itself from the Paninae (Gorilla and Pan) by a closer sphenoidal angle. They also showed the opening of this angle after the visible birth in all primates - at the exception of the modern man which remains in flexion, around a transversal axis of rotation. They then noticed correlations between the value of the sphenoidal angle and the position of these three canals around the axis of rotation, and then between this same sphenoidal angle and the frontalization of the petrous bone. A closure of the adult sphenoidal angle, between the two present groups, seems to be correlated to a frontalization of the petrous bones and at a rotation of the two posterior and superior canal compared to the lateral canal taken as horizontal reference. What is the origin of this changes? And when do they occured in the phylogeny? Dean and Wood [3] found some of these correlations with Australopithecus. Indeed, the basis of the cranium shows a tridimensional architectural unit, distinct from the Great Apes, around a closer sphenoidal angle and a frontalization of the petrous more noticeable. We talk about the shortening of the cranial base, but the values are still superior to those of the Homo group. The correlations remain nonetheless visual and consensual, without any calculation of the correlation coefficient between the angles. In an other part, the interpretation of the differences is brought back to the hypothesis of the differential development of the cerebellar territory [4-5]. It is thus even more recently that Spoor, Zonneveld and Wind have applied the new technics of medical imaging to the study of thirteen

Biography : Anne DAMBRICOURT MALASSE (1959), paleo-anthropologist, Charge de Recherche of the National Center of Scientist Research (1990), Consulting in Archaeology, Foreign Office, Department of Social Sciences and Archaeology (1995-1996), Head of the Hindou Kouch Pak-French archeological prospection (1996-1998), Member of the National Council of University (1995-1999).

70

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

fossil hominid cranium in two types, Australopithecus and Homo. The authors note the position of the otical capsul through the lateral canal, and evaluate the straightening of the clivus, the cranium placed in the Frankfurt plan. The orientation of the lateral canal remains the same for Homo and Australopithecus, it is close to 30°. Yet, the straightening of the clivus is largely more important in the Australopithecus than in the great Apes, but there does not appear to be different positioning between the lateral and posterior canals, at the difference of the Homo sapiens. Homo sapiens shows a straightening of the clivus even more marked and a new frontalization of the petrous. To summarize, the slope of the clivus straightens with Australopithecus, according to an ascending antero-posterior rotation of the sphenoïdal corpus on one hand, and the petrous becomes frontalized, on the other hand. But the lateral and posterior canals remain linked together as in Paninae. The clivus further straightens with Homo sapiens, the petrous further frontalize again, and this time the lateral canal detaches itself from the posterior canal. Thus, there may exist a dynamic relationhip between the flexion amplitude of human and the detachment of the lateral canal. To understand the origin of these changes, the authors still remain in the hypothesis of an angular difference caused by a differential development of the neo-cortex of telencephalon or cerebellum "The premize of all brain-size hypotheses is that because the cranial base is also the floor of the cranial cavity, brain size is a fundamental constraint on basicranial form" [6-7]. This point of view is established on an late analysis of cranial ontogeny. At the first ontogenic stages, the cranial base is overlayed not by the telencephalic brain, but by the neural tube. We are going to attempt to understand the origin of the straightening of the clivus in Homo sapiens, and follow the formation of petrous, as well as the three semi-circular canals, by developping an ontogenetic approach, since the embryonic period [7]. We have discovered that the morphogenesis of the basic cranial form was well known still 1900, and that the most important angles are acquired during the seventh week. We have reach numerous works describing the neural and the chondrocranial developments, and presented a global synthesis. Then, we applied the study protocol of Monkeys, Great Apes and analyse fossilized hominids cranium. Actually, numerous works have described the embryonic chondrocranium of human and monkeys since the beginning of this century, so that the compared anatomy of embryonic stages is possible. We could have thought that the straightening position of the clivus in a new born, is directely acquired during the ossification of the embryonic tissues. Yet, it is not the case. Like other mammals, in primate embryos (human included), the chondrocranium is flat. At the first embryonic stages, cartilageous tissues constitute the planum basale that is to say the sphenooccipital area, is plan. Levi [7] wrote as early as in 1900, the dynamic of this planum : during the seventh intra-uterine week, the sphenoidal part does a pivoting from top to bottom, back to forward, which thus brakes the rectitude of the planum and provoques a "plicature". At the beginning of the eighth week, the clivus, or the spheno-occipital slope, is acquired. The cartilageous tissues situated here and there of the sphenoidal body and the basi-occipital, that are the otic capsules, accompany the flexion. They will then translate by their positionning, the amplitude of this rotation. What is the origin of the sphenoidal rotation? The study of the neural tube growth trajectory underlying the chondrocranium, enables us to establish the dynamic correlations through the definition of an orthonorm analysis. This analysis is defined in lateral view. The chord defines the X axis, horizontal, and the apex defines the origin 0. The vertical axis Y, perpendicular goes through this apex (figure 1). An elongation of the tube is observed forward of the chord, at the same time as the beginning of a prial rotation from the bottom to the top, from the front to the back, around an axis Z, perpendicular to the two other axes. The neural tissue pursues the spiral rotation and goes beyond the chordal area. It is at this precise stage, that the authors, among whom Levi, have described the spheno-occipital rotation. The chord follows the movement, the shenoidal angle then appears. And the otic capsules associated to the basi-occipital follows the movement. The differentiation of the canals begins at the end of the fourth embryonic week. At the sixth week, the three canals are

DAMBRICOURT MALASSE et al.: Neural Tube, Spheno-occipital Flexion and Semi-circular Canals in Modern and Fossil Hominids

71

distinct, but the lateral canal is still linked to the posterior; after the flexion, it is separated. We therefore conclude that strains generated by the elongation-rotation of the neural tube, determine the morphogenesis of the chondrocranium, first with the sphenoïdal flexion aimed on the apex of the chord, then with lateral angular modifications such as those of the semi-circular canals. And it is therefore during the embryonic period, that one can trace the topographical differences between the different primate neuro-sphenoïdal dynamics. The phenomenon of rotation is common to all living primates, but the rotation of the chordal segment is less important in Monkeys and it is the entire basis of the cranium which is less bent. It will remain so during ontogeny. With the Paninae, Gorilla or Pan, comparison between an 8 month foetus of Gorilla and Homo sapiens, shows a sphenoidal angle more opened in the Gorilla than in the human, as well as a frontalization of the petrous and the posterior canal less marked. The origin of these angular divergences is therefore indeed linked to the intra-uterine morphogenesis of the cephalic pole. The embryonic brain of the Gorilla develops itself as in all primates, according to a spiral rotation of the neural tube which goes to the chord segment, but the rotation is of less amplitude than in human.

Figure 1

Homo sapiens neural rotation in lateral view [8]

ACTA ANTHROPOLOGICA SINICA

72

Supplement to Vol. 19, 2000

It is obvious that fossil primates, including hominids, have developped from a flat chondrocranium and that they went through a rotation of the neural tube. It is the amplitude in the chord part which has changed, very week in the first Primate, it became more important stating with the hominids. How do fossil craniums of Homo organize themselves according to this new key of understanding? In order to do so, the study material is made of cranium studied with X-rays and the CT scan for present species. Fossils are still in great majority studied from modern casts and exocranial measures. This study show the first result of the ontogeny of Pan and Gorilla, since the fetal stage. In the first step, we test the correlation between the intra and exo-cranial angular values, in profile and transversal view, from present species, in order to establish predictions on the intracranial flexion in fossiles (figure 2). They will be verified by CTs studies, according to the degree of mineralization. The results [8] are close to May and Sheffer [9] who compare the sphenoïdal angle with an external angle, namely the cranio-facial flexion. The first sample is established for the radios of adult craniums of Gorillas, Chimpanzees and Homo sapiens taken radomly in a pluri-ethnical population. For the CTs, we study one adult Homo sapiens, a serie of six chimpanzees of growing age from the age of one year, a formol fetus and adult Gorilla. For the original fossils, the skulls are: Homo sapiens; Le Rond du Barry, Qafzeh 6, Qafzeh 7, Homo erectus from Indonesia Solo 1 and Sambungmacan (collection of Professor Teku Jacob) and Homo neanderthalensis, La Chapelle aux Saints. The second sample comes from casts only. The results show first that for present species, two groups distinguish themselves according to the adult value of internal and external flexion, and the frontalization of petrous. Both groups are Sapiens and from its fossilized representatives, since Qafzeh, on one side, and the Paninae on the other hand, Pan et Gorilla. Concerning the orientation of the petrous in external view, and the sphenoidal angle, at the adult stage, the coefficient of correlaton is the highest between Gorilla alone and Homo sapiens (0,8), it is of 0,7 between Pan and Homo sapiens. A closing of the sphenoidal angle between Great Apes and Human does indeed accompany a frontalization of the petrous.

A

B Figure 2

Khirgiz, IPH Collection, X Ray analysis

A: Basi-cranio-facial angular analysis: 1: external nasal spine, 2: basion, 3: tuberculum sella, 4: lame criblée, 5: prosthion, 6: hormion, A1, A2, internal craniofacial contraction, a1, a2, external cranio-facial contraction, spheno: sphenoïdal angle. B: Petrous bone frontalization (beta) 1: intersection between the petrous bone and the clivus, 2: center of the carotidian canal, 3 and 4: porion

DAMBRICOURT MALASSE et al.: Neural Tube, Spheno-occipital Flexion and Semi-circular Canals in Modern and Fossil Hominids

Table 1

73

angular measurements, external, internal cranio-facial contraction, sphenoïdal angle and petrous frontalization, in Homo sapiens, Homo fossilis and Pan troglodytes

Pan

A1

A2

béta1

Anglsphéno

A 194328 A1943 35 A 197415 1974 20 A 1919 9 A 194327 19 902 1974 16 1957 70 A 194326

80 78 79 80 80 82 82 85 77 76

144 145 153 150 142 145 145 152 131 138

58 55 55 54 53 52 52 51 51 50

172 150 161 186 151 167 150 148 148 162

Homo sapiens h-69-73-1 h-69-3-4 h-69-19-1 h-69-61-5 h-69-67-2 h-69-71-1 h-69-81-6 h-69-69-1 h-69-86-6 h-1872 A 1919 7 1969 3 1 1969 13 8 1969 61 7 A 1955 111 1955-101-1 1969 8 24

59 48 50 55 45 56 47 49 47 53 50 42 42 50 36 59 57

96 74 79 88 78 83 69 80 70 85 101 84 88 107 68 95 94

49 47 46 46 45 45 45 43 43 41 51 33 38 39 38 37 32

129 140 121 131 122 127 124 134 123 132 145 135 133 136 133 135 131

Homo fossilis ER3733 ER 3883 Broken Hill OH9 Nariokotome Petralona Dali Sangiran 17 Ngandong Ngawi Sangiran Pit.4 Saccopastore La Chapelle La Ferrassie SkhulV Cromagnon

77 91 88 78 100 90 108 103 88 94

101 122 98 110 122 114 132 124 115 113

52 55 45 45 49 47 40 43 46 55 50 35 35

139

ACTA ANTHROPOLOGICA SINICA

74

Figure 3

Supplement to Vol. 19, 2000

Angular measurements between semi-circular canals ; cscl: lateral semi circular canal, cscp: posterior semi circular canal, cscs: superior semi circular canal, ax: great axis of the lateral semi circular canal, sag: sagittal plane, X: cross section of the posterior and superior canals in the lateral canal plan

Concerning the semi-circular canals and the sphenoidal angle (figure 3): - For the angular opening between the lateral and posterior semi-circular canals, the first measurements indicate a large difference of 20° between the child Chimpanzee of one year old and the adult Homo sapiens. Nonetheless, the value does not change during the post-natal ontogeny of the Pan, it varies from 9 to 11°, where the sphenoidal angle goes from 138° in a child less than 1 year old, and reach on average on 10 adults 160°. There appears a very clear rise of the semicircular angle, three times superior in Homo sapiens than in Pan, with a sphenoidal angle closer which does not enter the variabilitiy of Pan. The double angular discontinuity is clear. May and Sheffer describe the same differences for the sphenoïdal angle, but they indicate an ontogenic stability in Pan, whereas it is classic to consider an increasing angulation. Nevertheless, a recent study [10] highlights a distinction during the deciduous period in Paninae, and observe a fast phasis of growth in the second part of this period. In orthodontics, a cranio-facial analysis applied to Pan and Gorilla describe the same phenomenon. At birth the Paninae are closed to the fetal growth, in terms of flexion, then during the deciduous phasis, the face begins to develop while the flexion stops. - The frontalization of the lateral canal, given by its great axis relatively to the sagittal plane is the same in human and apes. In fact, the differences between the semi-circular canals are not the lateral canal relatively to the sagittal plane but that of the posterior canal. Therefore, this angulation is acquired during the intra-uterine development. It is during their formation, at the pre-cartilageous stage, that the movement of canal translations in their own planes are possible. Since the orthogonality of the three canals are maintained, the possible margin of modification remains the position of the canals in their own planes, in the direction of verticality or horizontality. That is what we observe for the lateral canal between the Chimpanzee and the Homo sapiens. The closing of the sphenoidal angle is accompanied by this angular divergence. These results are in convergence with the thesis of a possible rotation of the lateral canal in its plane, at the moment of the spheno-occipital rotation, once it goes beyond the embryonic step which would be at the minimum the one of Australopithecus. In Homo sapiens, the greater flexion of the clivus could explain greater tension at the level of the posterior semi-circular canal which leads to a "sagittalisation", and not the frontalization of the lateral.

DAMBRICOURT MALASSE et al.: Neural Tube, Spheno-occipital Flexion and Semi-circular Canals in Modern and Fossil Hominids

75

Dental stage of Pan : 1974-65 : dm2 unerupted 1974-33, c unerupted, 1974-35 M1 unerupted, 1974-51 M1 erupted, 1974-58 M2 erupted, 1974-29 M2 erupted. Table 2 Sphenoïdal angle Pan 1974-65 1974 3 1974-35 1974-51 Ado 1974 58 Ado 1974 29 Pan troglodytes Adulte n=10 Gorilla fetus Gorilla adulte n = 12 Sapiens n= 16 Solo 1 Qafzeh 6 Sambungm

142 150 136 158 151 159,5 [12,6] 164 151 [4,3] 120 [12,6] -

cscl/cscp angle (1) Cscs/sagpl Angle (2)

Axcscl/sagpl angle (3)

11 10 16 15 12 12

38 41 40 37 36 31

55

12

45

46 50

47

58 47 49

57 50 61 51

33 33 30

Concerning the casts of fossil hominids, the first exocranial measures of fossil Homo skull, indicates an external frontalization of the petrous more important than in Australopithecus, such as observed by Dean et Wood. It distinctly separates the craniums attributed to Homo, no matter the species considered. The frontalization is however even less than in Sapiens (table 2). Concerning the originals fossils, the study of the semi-circular canals is precious because it compensates for the very frequent absence of the basi-cranium. The cranium of Qafzeh 6 shows a divergence of the lateral canal of 30°. The petrous of Qafzeh 7 is too mineralized to be analysed. The two craniums of Homo erectus, Sambungmacan and Solo 1, only Sambungmacan is usable, but the only accessible measure is the internal frontalization of the petrous of 46° which does not give any information on the flexion. Neandertals are particular. A recent tomography of the juvenile petrous Arcy-sur-Cure [11], has permitted to notice a difference of position between the posterior canal, compared to the lateral, in the direction of verticality. The posterior canal went down, which is a characteristic found in La Chapelle-aux-Saints, La Ferrassie or La Quina. Yet, the application of a cephalometrical study in dynamic orthodontics [12], showed for La Chapelle-aux-Saints, a cranio-palatin equilibrium in extention relatively to the equilibrium of Homo erectus without any common point with CroMagnon or any contemporary Homo sapiens. Significant measurements of the basal skull are not the linear values, but the angular one. The Neanderthal basal skull is no longer than in modern Homo sapiens, thus this is not a shortening of the sphenoid which distinguishes Neanderthal from modern Man, as supported by Liberman, and discuted by Spoor et al. [13] but the sphenoïdal dynamics of rotation. This is found again with the larger sella turcica. The cranio-facial equilibrium is in extension relatively to Homo erectus. In this perspective, the low position of the posterior Neanderthalian canal could go in the direction of this basi-cranio-facial extension. The frontalization of the lateral canal is more important than in oldest skull, that is correlated with a greater external frontalization of the petrous. The entire sphenoïdal area in Neanderthals develops specific dynamics which has no equivalent in past, neither in modern human. A better

76

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

understanding of the occipito-spheno-ethmoïdal dynamics, as developed by Deshayes [12] could explain the singularity of classic Neandertals. Our hypothesis is that such angular changes illustrate a modification of the embryonic neuro-sphenoidal dynamic of European Homo antecessor or habilis [7]. Then here, Homo sapiens is no longer defined on empirical adult criterias such as the cranium capacity, but on more anterior datas which preceede telencephalization, not in the phylogeny, but in the ontogeny. The skull of modern man first appears as an embryonic evolution of the chondrocranium of Homo group, and not as a sub-species of the same embryonic organization. All the cranio-facial ontogeny is different. It would be interesting to study Dali [14], in the polycentric evolutionary perspective [15] while the fossil indicates a modern basal skull, but we notice a post-mortem compression effect. It would also be necessary to compare Homo antecessors with Qafzeh and the Neandertals. To conclude, a systematic study of semi-circular canals would enable to introduce the embryonic dimension in the discourse on the evolutionary modalities, and to better define the belonging of a fossil to one phylum rather than another. In general, since the first hominids, from the Great Ape roots, it seems obvious that the chondrocranium presents itself as the reflection of an embryonic evolution relative to the neuro-chondrocranial dynamics. Acknowledgement: Many thanks to Dr. Yvette Deloison and Dr. Dominique Grimaud Hervé for their advices. References : [1] SPOOR F, WOOD B, ZONNEVELD F. Implications of early hominid labyrinthine morphology of evolution of human bipedal locomotion, [J]. Nature, 1994, 369:645-648. [2] WIND J, ZONNEVELD F. Radiology of fossil hominid skull [A]. In: Hominid Evolution : Past, Present, Future. 1985: 437442. [3] DEAN MC. Homo and Paranthropus similarities in the cranial base and developing dentition [A]. In: Wood B, Martin L, Andrews P eds. Major Topics in Primate and Human Evolution. Cambridge: Cambridge University Press, 1986, 249-265. [4] ROSS CF , HENNEBERG M. Basicranial flexion, relative brain size, and facial kyphosis in Homo sapiens and some fossil hominids [J]. Am J Phys Anthropol, 1995, 98:575-593. [5] ROSS CF, RAVOSA MJ. Basicranial flexion, relative brain size and facial kyphosis in nonhuman primate [J]. Am J Phys Anthropol, 1993, 91:305-324. [6] MAY R, SHEFFER D. Growth changes in internal and craniofacial flexion measurements [J]. Am J Phys Anthropol, 1999, 110(1):47-56. [7] STRAIT DS, ROSS CF. Kinematic data on primate head and neck posture : Implications for the evolution of basicranial flexion and an evaluation of registration planes used in paleoanthropology [J]. Am J Phys Anthropol, 108-2, 1999:205222. [8] DAMBRICOURT-MALASSE A. Continuity and discontinuity during modalities of hominization [J]. Quat Interna, 1993, 19:85-100. [9] DAMBRICOURT MALASSE A. Nouvelles approches de l'évolution crânienne des Homo erectus de Java [A]. In: Sémah F, Falguères C, Grimaud D eds. Actes du Colloque International Singer-Polignac "Origine des peuplements et chronologie des cultures paléolithiques dans le Sud-Est asiatique: récents développements". Artcom’ Paris, 1999 (in press). [10] MILLET JJ, VIGUIER B, DAMBRICOURT-MALASSE A et al. Ontogenèse crânienne de Pan et de Gorilla et hétérochronies [R]. Comptes Rendus de l’Académie des Sciences, Sciences de la Vie, (in press). [11] HUBLIN JJ, SPOOR F, BRAUN M et al. A late Neanderthal associated with Upper Palaeolithic artefacts [J]. Nature, 1996, 381:224-226. [12] DESHAYES MJ. A new ontogenetic approach to craniofacial growth [J]. J Masticat Health Soc, 1997, 7:1-104. [13] SPOOR F, O’HIGGINS P, DEAN C et al. Anterior sphenoid in modern humans [J]. Nature, 1999, 397 :572. [14] WU X, POIRIER FE. Human Evolution in China : A metric Description of the Fossils and a Review of the Sites [M]. New York: Oxford University Press, 1995, 317. [15] POPE GP. Craniofacial evidence for the origin of modern humans in China [J]. Yearbook Phys Anthropol, 1992, 35:243-29.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

77-82

Enamel Microstructure of Lufengpithecus lufengensis ZHAO Lingxia , LU Qingwu, XU Qinghua (Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, Beijing 100044 , PR China )

Abstract The enamel microstructure of 5 permanent anterior teeth of Lufengpithecus lufengensis was observed with SEM. Incremental markings perikymata were clearly showed on the entire crown surface, and the density of perikymata showed a gradual increase towards the cervix. The crown formation times were estimated respectively using 7-days and 9-days periodicity of perikymata. Compared with fossil hominoids, modern humans and apes, crown formation time of Lufengpithecus lufengensis is close to that of Australopithecus afarensis and Australopithecus africanus, and closer to modern humans and apes; it is much longer than that of Proconsul heseloni and Proconsul nyanzae, Australopithecus robustus and Australopithecus boisei. The pattern of compactness of perikymata is similar to that of modern humans. Enamel prism patterns of Lufengpithecus lufengensis were observed. Concerning prism cross section patterns, Pattern 1 prisms occur in the very outer surface layer of enamel. Under super surface layer, Pattern 3 prisms predominate in the body of enamel, pattern 2 prisms are also found somewhere. In Lufengpithecus lufengensis, variants of pattern 3 exist, such as pattern 3A, 3B. It is strange and interesting that pattern 3B occurs at Lufengpithecus lufengensis, because up to date pattern 3B are recorded only in Homo sapiens, Homo erectus and Australopithecus, not found in extant and extinct apes. On longitudinal sections of enamel, Hunter-Schreger bands occur almost throughout the thickness from the enamel-dentine junction to the tooth surface. This is similar to Homo, and different from great apes. Of enamel microstructure, the preliminary results support the suggestion that Lufengpithecus lufengensis might be one of the members of hominoids related to early hominids.

Key words:

Lufengpithecus lufengensis; Enamel microstructure; Incremental markings; Crown formation time; Prism patterns

1 Introduction Studies of enamel structures and their implication to ontology and phylogeny have recently been realized by palaeoanthroplogists [1-2]. Two types of incremental growth lines are present within enamel: daily enamel prism crossstriation and circaseptan striae of Retzius or perikymata, these incremental markings provide an absolute timetable with which we can furthermore understand dental developmental events [3-5]. The ultrastructural unit of enamel is the prism, and prism patterns are an importance taxonomic tool. Scanning election microscopic analysis of enamel can provided new insights into hominoid evolution [6-10]. In this paper, enamel microstructure of Lufengpithecus lufengensis is addressed in (1) incremental markings and crown formation time, (2) enamel prism patterns.

2 Materials and Method The materials are 4 isolated complete permanent teeth (one upper right incisor, one upper right lateral incisor, one lower right incisor and one lower left canine) and one permanent canine fragment. The observed position is placed on the lateral crown. For enamel prism pattern, according Gantt’s [11] suggestion that the area of the tooth in which prisms are best arranged to study their pattern and organization is the mid-lateral crown area, a polished facet on the midlateral crown is etched with 0.1 M phosphoric acid for 40s. In addition, One bucco-lingual longitudinal section of the canine was prepared, in order to observe the cross-striations and shape and arrangement of enamel prisms. The specimen were analyzed with SEM (JSM-1600).

Foundation item: The present work was supported by a special Fund “Early Human Origin and Its Environmental Background” granded by NSFC and MOST. Biography: ZHAO Linxia, currently carrying out the research on the microstructure of hominoids.

ACTA ANTHROPOLOGICA SINICA

78

Supplement to Vol. 19, 2000

A

B Figure 1

Perikymata (A) and striae of Retzius (B) of Lufengpithecus lufengensis

Table 1 Preikymata counts and estimated crown formation times in Lufengpithecus lufengensis Specimen

Number of Perikymata

Crown formation time(yrs)* (7-days)

(9-days)

172

3.8

4.7

PA811.2 RI

151

3.4

4.2

PA895 RI1

128

3.0

3.7

161

3.6

4.3

1 2

PA811.1 RI

PA826 lower LC

*Crown formation times are calculated using 7-days and 9-days cross-striation repeat interval between adjacent striae of Retzius and including 6 months growth prior to visible perikymata.

ZHAO et al.:

Enamel Microstructure of Lufengpithecus lufengensis

Figure 2

79

Nomenclature of enamel prism outlines (after Boyde [9])

3 Results Incremental markings perikymata were clearly showed on the entire crown surface both before and after the polished and acid treatment , and the density of perikymata showed a gradual increase towards the cervix. Perikymata counts were 172 for I1, 151 for I2, 128 for I1, 161 for the lower canine. In order to know the periodicity of perikymata or striae of Retzius and estimate crown formation time, one permanent canine tooth fragment of Lufengpithecus lufengensis was longitudinally sectioned, and the daily incremental markings cross striatons occur along the enamel prisms. The number of the enamel prism cross-striations between adjacent striae of Retzius counts 9 in this tooth. So the periodicity of perikymata is 9 days, and we use it to estimate the crown formation time of other 4 teeth, including about 6 months time for hidden growth increments in hominoid incicors [4-5]. This result is different from the result of Zhao et al. [13] which used a 7 days periodicity of Perikymata without the number of enamel prism cross-striations between adjacent Striae of Retzius. So the crown formation time is longer than before. Enamel prism cross section patterns of Lufengpithecus lufengensis were observed.. According to the nomenclature of enamel prism outlines after Boyde [9] (Fig. 2), Pattern 1 prisms reveal in the outer surface layer of enamel, the cross section of the enamel prism is circular or subcircular. Pattern 3 prisms predominate in the body of enamel under the super surface, the prisms arrange in cross rows, the head of each prism is orientated towards the occlusal, and the narrow tail orients cervically, the tails of one row of prisms fit between the heads of the next row. Pattern 2 also occur somewhere, prisms arrange in longitudinal rows (Fig. 3). In Lufengpithecus lufengensis, variants of pattern 3 exist, such as pattern 3A and 3B. Pattern 3A has a half round head with a slender tail as tadpole-shape, while pattern 3B has a more than half circle head with a wide tail like a fish tail (Fig. 4) On the longitudinal sections, Hunter-Schreger bands (Fig. 5) occur almost throughout the thickness from enamel-dentine junction to the tooth surface.

80

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

4 Discussion Perikymata are easily visible on the buccal surface of most unworn or unabraded fossil hominid teeth, as well as on newly erupted unworn modern human teeth [3]. But it is unusual to see them easily on the surface of anterior teeth in great apes, especially extending clearly from the incisal edge to cervix. Perikymata on the anterior teeth surface of Lufengpithecus lufengensis shows clearly.

Pattern 1

Pattern 2

Pattern 3 Figure 3 Enamel prism patterns of Lufengpithecus lufengensis

ZHAO et al.:

Enamel Microstructure of Lufengpithecus lufengensis

81

The pattern of compactness of perikymata of Lufengpithecus lufengensis is similar to that of modern humans. Modern human incisors have perikymata that widely spaced at the incical third of the tooth but which are close together at the cervix, incisor teeth of Autralopithecus afrensis, Australopithecus africanus and of early Homo also show this pattern [3], incisors of Lufengpithecus lufengensis show this pattern too. This pattern indicates changes in the rate of enamel secretion during the period of formation: incicor teeth begin forming quickly but gradually, slower down toward the third cervix. But incisor teeth of Australopithecus robustus have widely spaced perikymata that remain so even at the cirvix, and incisor teeth of Australopithecus boisei also do not show a marked reduction in the spacing of perikymata [3]. Although there is some debate over the periodicity of perikymata or striae of Retzius, there is considerable evidence [3, 5-6] that suggests that in extant hominoids they reflect a circaseptan(7-9) rhythm. The periodicity of enamel striae of Lufengpithecus lufengensis is within the range of extant hominoids’, it is longer than that of Proconsul’s of 5 or 6 days rhythm [8] which is resemble some extant New and Old World monkeys.

Figure 4

Enamel prism patterns 3A (left) and 3B (right) of Lufengpithecus lufengensis

Figure 5

Hunter-Schreger bands of Lufengpithecus lufengensis

Compared with fossil hominoids [4-5, 7-8], modern humans and apes, incicor crown formation time of Lufengpithecus lufengensis is close to that of Australopithecus afarensis and

82

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

Australopithecus africanus, and closer to modern humans and apes, but longer than that of Proconsul heseloni and Proconsul nyanzae, Australopithecus robustus and Australopithecus boisei. Enamel prism pattern of Lufengpithecus lufengensis occur mainly as pattern 3. It is strange and interesting that pattern 3B are seen in Lufengpithecus lufengensis, because up to now pattern 3B are recorded only in Homo sapiens, Homo erectus and Australopithecus, not found in extant and extinct apes. Gantt [11-12] proposed that hominids, including Homo, Australopithecus, have 3B enamel prism pattern, 3B enamel prism pattern and the marked increase in enamel thickness are unique features in human evolution, just as is bipedality. However, this hypothesis is still in question, it should be confirmed in future. If it is ture, how to evaluate the prism patterns of Lufengpithecus lufengensis is very interesting. Patter 3 enamel and Hunter-Schreger bands of Lufengpithecus lufengensis occur almost throughout the thickness from enamel-dentine junction to the tooth surface. This is similar to that of Homo, and different from Pongo and african great apes, Pongo has a relatively thin outer layer (less than 20% of the maximum thickness) of pattern 1 enamel overlying the deep pattern 3 enamel. Pan and Gorilla have a thick (greater than 40% of maximum thickness) outer portion of Pattern 1 overlying the deep pattern 3 enamel [10]. Above all, the preliminary investigation at present indicates: enamel microstructure of Lufengpithecus lufengensis is in some way similar to hominids, and it supports the suggestion [14-15] that Lufengpithecus lufengensis might be one of the members of hominoids related to early hominids. References: [1] WINKLER LA, SWINDLER DR. Primate dental symposium: old and new questions, new trends [J]. Am J Phys Anthropol, 1991, 86:107-111. [2] ROZZI FR. Enamel structure and development and its application in hominid evolution and taxonomy [J]. J Hum Evol, 1998, 35:327-330. [3] DEAN MC. Growth layers and incremental markings in hard tissues.A review of literature and some preliminary oberservitions about enamel structure in Paranthropus [J]. J Hum Evol, 1987, 16:157-172. [4] BROMAGE TG. The biological and chronological maturation of early hominids [J]. J Hum Evol, 1987, 16:257-272. [5] BEYNON AD, DEAN MC. Distinct dental development patterns in early fossil hominids [J]. Nature, 1988, 335:509514. [6] FITZGERALD CM. Do enamel microstructures have regular time dependency? Conclusions from the literature and a large-scale study [J]. J Hum Evol, 1998, 35:371-386. [7] MOGGI-CECCHI J, TOBIAS PV, BEYNON AD. The mixed dentition and associated skull fragments of a juvenil fossil hominid from Sterkfontein, south Africa [J]. Am J Phys Anthropol, 1998,106:425-465. [8] BEYNON AD, DEAN MC, LEAKEY MG et al.Comparative dental development and microstructure of Proconsul teeth from Rusinga Island, Kenya [J]. J Hum Evol, 1998, 35:163-209. [9] BOYDE A. The structure and development of mammalian enamel [D]. PhD dissertation, University of London, 1964. [10] MARTIN L, BOYDE A. Rates of enamel formation in relation to enamel thickness in hominoid primates [A]. In: FEARNHEAD RW, SUGA S eds. Tooth Enamel IV. London: Elsevier Science Publishers, 1984, 447-451. [11] GANTT DG. Neogene hominoid evolution--a tooth’s inside view [A]. In: KURTEN B ed. Teeth: Form, Function and Evolution. New York: Columbia University Press, 1982, 93-108. [12] GANTT DG. The enamel of Neogene hominoids —structural and phyletic implications [A]. In: CIIOCHON RL, CORRUCINI RS eds. New Interpretations of Ape and Human Ancestry. New York and London: Plenum Press, 1983, 249-298. [13] ZHAO L, OUYANG L, LU Q. Incremental markings of enamel and ontogeny of Lufengpithecus lufengensis [J]. Acta Anthropol Sin, 1999, 18:102-108. [14] WU R, XU Q, LU Q. Relationship between Lufeng Sivapithecus and Ramapithecus and their phylogenetic position [J]. Acta Anthropol Sin, 1986, 5:1-30. [15] WU R, XU Q, LU Q. A revision of classification of Lufeng Great apes [J]. Acta Anthropol Sin, 1987, 6:265-271.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

83-89

Arboreal Primates and Origin of Diagonal Gait LI Yu (Department of Human Anatomy and Cell Biology, The University of Liverpool, L69 3GE, UK)

Abstract It has long been noted that the primates normally adopt a diagonal gait in their quadrupedal walking, while other mammals use a lateral one. There is not a satisfactory interpretation about this behaviour difference though some researchers have suggested a number of hypotheses. This study is based on the analyses of kinematic and kinetic data, collected from a collection of primate and non-primate species. The results show that black and white ruffed lemurs (Varicia variegata) do not usually use a diagonal gait while on the ground, but they nearly always use this typical primate gait while on top of a branch. As regard non-primate, coatis (Nasua nasua) used lateral gait both on the ground and on the branch, but not when they are hanging-walking underneath a rope. This phenomenon suggests that the substrate is an important factor in the evolution from lateral to diagonal gaits in primates.

Key words:

Quadrupedal primates; Diagonal gait; Arboreal; Biomechanics

1 Introduction Quadrupedal gait characteristics of primates and other mammals have been attracting attentions from a large number of researchers in recent decades. One of the important topics in this field is the gait pattern. Muybridge [1] was the first to distinguish the gait sequence of primates from the other mammals. The forward cross type and backward cross type were one time used by Iwamoto and Tomita [2] as lateral and diagonal sequence. The former is the typical gait of non-primate mammal, and the latter is that of primate. Figure 1 shows the difference in these two sequences. While primate is adopting a diagonal sequence, it is also using the diagonal couplets. This means that a forelimb lands immediately after the opposite hind limb does. Therefore, a primate in walking is supported by a pair of diagonal limbs.

L

R

L

R

F(fore)

H(hind) (A) Figure 1

(B)

Diagonal (A) and lateral (B) gaits. If the counting starts from right hind limb (Rh), the sequence for diagonal gait is RhLfLhRf, and the lateral gait is then RhRfLhLf. Primates in general adopt a diagonal sequence, diagonal couplets gait

Rollinson and Martin [3] provided a detailed description of the gait pattern and a, then up-todate, review of the analyses in the quadrupedal locomotion studies. In the last 20 years, kinetics of

Foundation item: This research is carried out under a grant from BBSRC of the UK. Biography: LI Yu has obtained M.Sc in IVPP and Ph.D in Liverpool University. He is currently a research Fellow in Liverpool, with research interests in human and primates locomotion.

84

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

primates has been widely studied. With the assistance of digital computers and force plate, the contact forces during primate locomotion were measured. In this field, Kimura et al [4], Kimura [5], Reynolds [6-8], and Demes et al. [9] have made significant contributions. Schmitt [10] studied the compliant walking of primates, but the kinetic results were not reported. Primates have been considered unique to other quadrupeds for both their diagonal gaits (diagonal sequence and diagonal couplets) and larger hind-limb supporting force. While in the other mammals the forelimbs support most of the body weight, primates have a pair of dominant hind limbs. Reynolds [6-7] built a mechanical model to explain the reason, believing that the larger hind limb forces are due to larger flexion torque. His argument will be discussed later. Using the force and gait sequence data, this paper tries to explain the possible reasons for the important characteristics of primate's locomotion, the use of diagonal gait. The phenomenon of hind limb domination is also addressed.

2 Materials and Methods The data collection of primate locomotion was carried out at Chester Zoo, North-West England. A Kistle force plate (9281B) and two synchronized video cameras were set up to record the kinetic and kinematic data. The animals studied are black macaque (Macaca maurus), black and white ruffed lemur (Varicia Variegata), and tufted capuchin (Cebus apella). For comparative purposes, a non-primate species, coati (Nasua narica), was also included in the experiments. For the three primate species, the subjects were confined in their enclosures, and there was no or little interference from the researchers. As a result, the data collected were from natural and random move of the animal. For the coatis, an animal keeper provided help in data collection. Fruit was used to lure the animal to cross the force plate. However, there was no direct physical contact between the keeper and the animal when it was going over the force-recording device. An instrument simulating a horizontal branch was constructed. It consists of a strong metal frame, a wooden bar of 15 cm in length and 22 mm in diameter. The length is just enough to accommodate the two stance limbs, and the diameter is about the same as the largest horizontal branches in the enclosures of lemurs and macaques. The wooden bar was horizontally mounted on top of the frame with the support of two spring-pillar complexes. The complexes allow the bar to have a vertical movement when there is a force acting on it, but prevent any horizontal displacement over 1 mm from its central position. The stiffness of the spring was so selected that it can be depressed by about 2.5 cm with a force of 25N. In order for the animal to move on and off the measuring bar smoothly, two guiding bars, both about one metre long and one on each side, were built on two separated supporting frames. The detail of the setup is shown in Figure 2. Due to the gait characteristics, only the force/torque of the same side fore and hind limbs can be registered in a single record. The opposite limbs will not land at the measuring bar in a normal gait. One of the two video cameras is set at a right angle to the measuring bar to record the lateral view of the animals. The other camera was set in line with the bar, recording either the front or back of the animal. The signals from the two cameras were synchronized and recorded on the same video frame via a mixer. The frequency of the frame was 25 per second. The sample frequency was doubled in the lab, where a single frame was split into two fields through computer image processing. The sagittal torque for lemurs was recorded in a supplementary experiment. This was taken with a torque gauge mounted on the supporting frame. The gauge's body was fixed and its measuring head held a wooden bar with the same dimension as the measuring bar mentioned earlier (22 mm). When a lemur moved passing the system with its fore or hind limb holding the bar, the value of the torque was recorded (Fig. 2, right). The general behaviour of the animal in their enclosure was randomly recorded with a High 8 video camera.

LI:

85

Arboreal Primates and Origin of Diagonal Gait

D E

E

C

D

C

D

E

B

B A

A

E

Figure 2

A: Force plate; B: Supporting frame; C: Measuring bar; D: Guiding bars; E: Sprint-pillar complex

Experimental setup (left, general setup; right, torque-measurement equipment)

3 Results 3.1 Gait sequence The capuchins and the macaques always use the diagonal sequence, diagonal couplets (RhLfLhRf) gait, irrespective of whether they were on the ground or on the measuring structure. The pattern of lemurs differs with the substrates. On the flat surface, i.e. in terrestrial walking, the lemurs use lateral gait on most occasions, like non-primates mammals. Of the 11 observations, eight were recorded as lateral gait (RhRfLhLf); of the three occasions when the diagonal gait was observed, two were in a normal and continuous locomotion, while the other the result of an abrupt speed change. The lemurs change their gait pattern completely when walking on the measuring structure. In over 50 recorded video sequences, plus a large number of visual observations, a diagonal gait was employed when they were walking on a branch or similar structures. Coatis use the lateral gait in all the events on the ground, with the so called “singlefoot” (limbs falling time is evenly distributed). The same gait sequences were also adopted when they were walking on the measuring bar over the force plate (n>20). The coatis use the same lateral gait on top of a rope, which was hanging in their enclosure (n=6, and each case was continually recorded for from three to ten gait circles). The coatis only gave up the lateral gait when they walked underneath the rope. There were only four sequences recorded for this kind of locomotion. One of them with five gait circles showed a pattern RhLhLfRf. The other three records showed the diagonal sequences. Those sequences have from five to eight continuous cycles each. 3.2 Transverse force and its direction-cosine Figure 3 shows the direction cosines of transverse force. The direction cosine (dc) is defined as:

dcx =

Fx Fx + Fy 2 + Fz 2 2

Fx, Fy, Fz are transverse, sagittal, and vertical forces respectively. Transverse direction cosine represents the relative value of the transverse force in proportion to the limb force vector. By definition, the direction cosine has always the same sign as the force. Macaques have very small transverse forces, and the value is much more variable with no consistent pattern emerged; the value is therefore not shown here. The transverse forces of lemur have a much regular shape. For nearly all the records, the directions of the fore and hind limbs are always medial, with only 10% records showing irregular shapes.

ACTA ANTHROPOLOGICA SINICA

86

Forelimb

Supplement to Vol. 19, 2000

Forelimb Hind limb

Hind limb

Figure 3

Direction cosine values of Lemur (left) and coati (right). In these charts, a positive value indicates a laterally directed, and negative value a medial directed, force. The horizontal axis is time, which starts from touches down of a forelimb (time=0), and ends at the same side hind limb leaving the substrate (time=1)

Coati adopted a lateral gait on top of the measuring bar, which means that before a hind-limb has landed on the substrate, the same side forelimb has already lifted. As a result, the forces on fore and hind-limbs were separated. For the coati’s transverse force, the direction changes from lateral to medial shortly before middle stance for forelimb and after middle stance for the hind-limb, while in the lemur and macaque, the transverse force, especially for forelimb, does not change direction during the whole stance period.

0.2

0.2 0.1

0.1

0.1

0

0

0 0

0.5

1

1.5

0

0.5

(A)

2

-0.1

0 0.5 1 1.5 2 2.5 3 3.5 4 (C)

0.1

0.1

0

0 0

1.5

(B)

0.2

-0.1

1

-0.1 0.5

1 (D)

Figure 4

1.5

2

2.5

0

0.5

1

1.5

2

(E)

Saggital torque of lemurs. The unit of torque (vertical axis) is Nm, and the horizontal axis is time in seconds. For subcharts B-E, it should be noted that earlier part of the torque is from forelimbs, and later part from hind limb

LI:

87

Arboreal Primates and Origin of Diagonal Gait

3.3 Sagittal Torque of Lemurs When lemurs walked across the horizontal bar, their fore and hind limbs hold the bar to balance the body. A torque is then applied to the supporting bar. The value of this torque has been recorded with a torque gauge. It is shown for all the records, regardless of speed, that the forelimb rotates the bar laterally. In all the 16 records that have forelimb torque registration, only lateral torque was recorded (Fig. 4A). Hind limbs have a more complex pattern, which is largely dependent on the locomotion speed. For the high-speed excursion (the time from when the forelimb touches down to when the same side hind limb leaves the substrate is less than 1.0 second), the hind limb has the same rotation as the forelimbs, i.e. lateral (n=6, Fig. 4B). For lower speeds, in some occasions the hind limbs also have lateral rotation (n=2, Fig. 4C), while in some other occasions the rotation in both directions were shown (n=6, Fig. 4D), or medial direction only (n=2, Fig. 4E). The torque data were only recorded for lemurs, because the measuring structure is not robust enough to withstand manipulation by the other primate species. Detailed check of video records indicate that lemurs use their fore and hind-limbs differently when they walk on branches. Their hands never hold the branch closely in their normal continued walking (the distal segments of fingers do not touch the branch). Instead, their thumb and other fingers shape like an upside down U, containing the branch in the middle (Fig. 5A). Their feet hold the branch with the fingers forming a closed circle (Fig. 5B). Coatis never hold the bar like a lemur does. They simply land the limbs on top of the bar as if it is a narrower surface.

(A) Figure 5

(B)

(C)

How lemurs hold a branch: (A) Forelimb; (B) Hind limb, middle- stance; (C) Hind limb, landing

4 Discussion The gaits of lemurs and coatis provide us with a key to the question of the origin of the diagonal gait. When lemurs are walking on the ground, they like most non-primate quadrupeds, use a lateral gait. However, as soon as a lemur walks on a branch, its gait changes to diagonal exclusively. Prost and Sussman [11], and Vilensky and Patrick [12] noticed that another lower primate species, squirrel monkey (Saimiri sciureus) does not use diagonal gait in level ground walking either. It would be interesting to see what this primate would do when walking on top of a branch. The results show that lemurs change their lateral gait to diagonal when walking on top of a branch. In this situation, lemurs need to hold the thin branch for balance, and the substrate is then subjected to both torque and forces. Any deviation of the centre of gravity of the animal from the substrate in the transverse direction has to be balanced by a sagittal torque. For the coatis, a lateral gait is still used even when they are walking on a branch or a thick rope. The diagonal gait is only adopted when they are hanging-walking underneath the rope. This is partly because coatis do not hold the substrate like primates do. Coati’s transverse force (Figure 3b shows the direction-cosine of the force) indicates that each of the limbs has forces in both lateral and medial directions, so they have control for movement of both left and right. This differs fundamentally from arboreal lemurs. The force

88

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

characteristic of coatis enables them to control their transverse movement without holding the substrate. However, coatis must hold the rope when they are hanging underneath it, where a torque in sagittal direction is unavoidable. Only in this condition, coatis adopted the primate-like diagonal gait. It would thus seem that during arboreal walking, the diagonal gait is correlated with the holding of the branch and the resulted torque. Figure 5 shows the frontal view of a lemur walking above a branch. Figure 5A shows the way a forelimb is supporting on the branch. The hand does not closely hold the branch. Its thumb and the other fingers are on both sides of the branch to ensure stability, and the palm is on top for supporting the body weight. And Figure 5B shows the case for a hind-limb. Four toes are on the lateral side of the body, and the big toe is on the medial side. The hind limb holds the branch closely during the whole stance phase, except for the short periods immediately after touching down and before lifting off (Fig. 5C). It has been shown that the forelimb always produces lateral torque. Without more detailed information, we may assume from the available data that the action force of the forelimb passes the centre of the branch at the same side of the supporting limbs, resulting in a lateral torque. This may be called the default torque, which is produced by the limb's strut force [13], with no need of muscle activity. This inference is based on the fact that an open holding hand could not apply a substantial torque to the branch due to lack of friction on the contact. The hind limb holds the branch completely so that a couple may be generated apart from the torque from strut force, The results show that the torque are still lateral at fast walking speed without exception. It may be considered that during fast walking, only the default torque (lateral) is produced, for both fore and hind- limbs, because there is little demand in balancing the body in the sagittal plane. Only during slower walking, the time is long enough to accumulate a rotational movement on the lemur's centre of gravity and hence to endanger its balance. In this situation, a torque in either lateral or medial direction is required and the default torque is overridden, resulting in the complicated hind-limb torque patterns (Fig. 4C-E). In a natural environment, most of the substrates for primates are elastic, which change shape when forces/torques are applied. The balance for both substrate and the animal can be achieved by producing opposite torques with one of the another stance limb; this obviously results a diagonal gait. This is the most likely reason that primates adopted a gait of diagonal couplets and sequence, because arboreal is their most frequently used locomotion style. On a non-rigid substrate, the diagonal gait has another obvious advantage. The hind-limb is landing to a position of the substrate which is still under the control of the forelimb, and the forelimb can also be the reference point for hind-limb landing, hence the uncertainty for the landing point is greatly reduced. On the other hand, for the lateral gait, the forelimb has already lifted before the hind-limb lands. As a result, there would be more uncertainties for hind-limb landing in lateral gait. Based on observations of animal quadrupedal walking while show that the hind limb always lands at a position near the same side forelimb (Fig. 6), one of the characteristics of diagonal gait is obvious. For the diagonal gait, the step length is much larger than that of the lateral gait. For diagonal gait, when a hind limb starts to swing forward, the same side forelimb is at the most forward of all limbs. Therefore, the hind-limb will step cross the supporting position of both opposite limbs (Fig. 6A). On the other hand, for lateral gait, the hind-limb is not going to pass the opposite forelimb (Fig. 6B). For a given limb length, these characteristics mean that the diagonal gait requires a larger swing angle forward for the limb joints. Whether an advantage in locomotion or not, this explains the fact that primate has larger hind-limb forces. The fact that the hind limbs land at an anterior position in relation to other parts of the body changes the balance of the whole animal, resulting in more bearing of the force on the hind-limbs. Under this condition, the extension muscles at the hip joints must generate larger extension force to prevent it from collapsing. Reynolds [6] correctly identified this point, but wrongly reasoned it. Both hind limb support and larger extension torque are the results of the anterior landing of hind-limb, which is a consequence of a diagonal walking.

89

Arboreal Primates and Origin of Diagonal Gait

(A) Diagonal Gait

Figure 6

Distance for hind-limb to travel

Distance for hind-limb to travel

LI:

(B) Lateral Gait

Diagonal gait has a longer step length. Small circles represent fore limbs, and large circles the hind limbs. The open larger circles indicate the positions for the right hind limbs to land

What is the advantage of hind-limb supporting for primates? It seems reasonable to believe that it is the same as diagonal gait, i.e. to balance the body on a branch. With lighter supporting duty, the forelimbs will fulfil their exploring and grabbing function better. On top of a branch, the supporting condition is more variable than on the ground. As a result, the primates have to be tentative for each new supporting point, and put less weight on it to avoid possible danger. The same activity, walking on horizontal branch, has led the primates to a gait pattern with diagonal sequence, diagonal couplets, and hind-limb support. This pattern fundamentally differs from the traditional non-primate type of gait, and is possibly the first step to an even greater modification in locomotion style. References: [1] MUYBRIDGE M. Animal in Motion [M]. London: Chapman & Hall, 1899. [2] IWAMOTO M, TOMITA M. On the movement order of four limbs while walking and the body weight distribution to fore and hind limbs with standinf on all fours in monkeys [J]. J Anthropol Soc Nippon, 1966, 74:228-231 (in Japanese). [3] ROLLINSON J, MARTIN RD. Comparative aspects of primate locomotion, with special reference to arboreal Cercopithecines [J]. Symp Zool Soc London, 1981, 48:377-427. [4] KIMURA T, OKADA M, ISHIDA H. Kinesiological characteristics of primate walking: its significance in human walking [A]. Environment, Behavior, and Morphology: Dynamic Interactions in Primates. New York: Gustav Fischer, 1979. [5] KIMURA T. Bipedal and quadrupedal walking of primates: comparative dynamics [A]. Primate Morphophysiology, Locomotor Analyses and Human Bipedalism. Tokyo: University of Tokyo Press, 1985. [6] REYNOLDS TR. Mechanics of increased support of weight by the hindlimb in primates [J]. Am J Phys Anthropol, 1985, 67:335-349. [7] REYNOLDS T R. Stresses on the limbs of quadrupedal primates [J]. Am J Phys Anthropol, 1985, 67:351-362. [8] REYNOLDS T R. Stride length and its determinants in humans, early homonids, primates, and mammals [J] Am J Phys Anthropol, 1987, 72:101-115. [9] DEMES B, LARSON SG, STERN T et al. The kinetics of primate quadrupedalism: "hindlimb drive" reconsidered [J]. J Hum Evol, 1994, 26(4): 353-374. [10] SCHMITT D. Compliant walking in primates [J]. J Zool Soc London, 1999, 248:149-160. [11] PROST JH, SUSSMAN RW. Monkey locomotion on inclined surface [J]. Am J Phys Anthropol, 1969, 31:53-58. [12] VILENSKY JA, PATRICK MC. Gait characteristics of two squirrel monkeys, with emphasis on relationships with speed and neural control [J]. Am J Phys Anthropol, 1985, 68:429-444. [13] GRAY J. Studies in the mechanics of the tetrapod skeleton [J]. J Exper Biol, 1944, 20:88-116.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

90-97

Computer-Assisted Paleoanthropology: Methods, Techniques and Applications Christoph P. E. ZOLLIKOFER , Marcia S. PONCE DE LEÓ N (Anthropological Institute and MultiMedia Laboratory/Dept. of Computer Science, University of Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland)

Abstract The rapid evolution of computer-based technologies has opened up a wide area of applications in the biosciences and has led to specific developments in paleoanthropology. The methodological trio of computer tomography (CT), computer graphics and stereolithography constitutes the paradigm of computer-assisted paleoanthropology (CAP): Combining CT-based 3D data acquisition techniques with dedicated computer graphics tools, it is possible to prepare and reconstruct fragmentary fossils in virtual reality, to correct taphonomic deformation and to subject them to morphometric analysis. Virtual fossils serve as a basis to derive a variety of morphometric data such as 3-dimensional landmark coordinates, regional distribution of bone thickness and curvature, endocranial volumes, etc. CAP has proven to be particularly helpful in the reconstruction and morphometric analysis of fragmentary Neanderthal specimens and in the investigation of Neanderthal phylogeny and ontogeny.

Key words:

Computer tomography; Computer graphics; Stereolithography; Virtual reality; Fossil reconstruction; Neanderthals

1

Introduction

One of the major difficulties that commonly arises in the comparative analysis of fossil morphology is the lack of extensive samples and the incompleteness of individual specimens. Given the necessity to gain a maximum of information from a minimum of material evidence, one of the main challenges of paleoanthropology is to enhance the investigative power of morphometric methods while minimizing the invasiveness of the methods used for fossil preparation, reconstruction and data acquisition. The framework of Computer-Assisted Paleoanthropology (CAP) represents a computational approach to tackle these problems. In its essence, CAP integrates computer tools for 3-dimensional data acquisition, handling and analysis that make possible to perform the complete set of tasks involved in fossil analysis in a Virtual Reality (VR) environment. CAP provides a systematic approach to the problem of fossil reconstruction and opens up new ways of morphometric analysis of fossil remains. Ultimately, the utilization of CAP might also help to solve a long-standing problem in paleoanthropology, the generally restricted access to original specimens, as their direct examination can essentially be replaced by the computer-assisted analysis of their virtual counterparts.

2

Computer tools for fossil reconstruction and morphometry

Prior to undertaking any analysis of a fossil specimen, it is necessary to take into account all potential processes that might have contributed to its present morphological state. In this regard, our primary step is to isolate the ontogenetic and phylogenetic morphological signals against a background of “diagenetic noise” in order to infer the in vivo state of the organism. Fossil reconstruction, therefore, consists of two diametrically opposite procedures: one concerns the removal of diagenetic disturbances, the other the reconstitution and extrapolation of missing information.

Biography: C.P.E. Zollikofer did his PhD in neurobiology. His current research focuses on computational morphology and computer-assisted surgical planning; M.S. Ponce de Leon did her PhD in anthropology. Her current research focus is on Neanderthal ontogeny.

ZOLLIKOFER et al.: Computer-Assisted Paleoanthropology: Methods, Techniques and Applications

91

2.1

Data acquisition, segmentation and visualization Computer Tomography (CT) has revolutionized non-invasive 3-dimensional data acquisition techniques through its capacity to provide X-ray-based cross-sectional images of solid objects. Since the mid-eighties, CT scanning has been extensively used not only in medical diagnostics, but also for "fossil diagnostics", most notably to reveal internal anatomical features and regions still covered by matrix [1-3]. The acquisition of quantitatively reliable CT image data from fossils with the aid of medical scanners poses a specific technical problem: compared to living skeletal tissue, fossil material typically exhibits elevated X-ray densities brought about by bone re-mineralization during diagenesis. CT image artifacts caused by high-density objects can be corrected using dedicated algorithms and software tools that were originally developed to suppress artifacts caused by metallic implants in CT images of patients [4]. CT technology can also be used to acquire 3-dimensional data, essentially by combining series of consecutive cross-sectional images. Once volume data of fossil specimens have been acquired, and before 3D object representations can be generated, data segmentation procedures are applied in order to extract object regions of specific relevance. Borrowing a term from technical sciences, the entire procedure of data sampling and 3D reconstruction corresponds to a reverse engineering process: data structures are derived with the aid of computer tools from pre-existing objects, as opposed to computer-assisted design of objects. 2.2

Data manipulation in Virtual Reality It is apparent that manipulating graphical object representations on a computer screen, rather than working with real fossil specimens, is of immediate practical and theoretical benefit for paleoanthropological applications. While manual procedures such as preparing, reconstructing and measuring fossil specimens are highly invasive, analogous computer graphics manipulations are completely non-invasive. Using computer tools, virtual fossils can be re-aligned to correct for taphonomic distortion, and incomplete specimens can be reconstructed by assembling isolated fragments and completing missing parts by mirror-imaging or using data from similar specimens. Furthermore, virtual objects provide an ideal basis for the analysis and visualization of various morphometric parameters. Fossil reconstruction and morphometry in a computer graphics environment involves complex interactive manipulations and modifications of individual 3D objects on a computer screen. To achieve maximum efficiency, it is necessary to combine two basic concepts of software technology in computer graphics, Virtual Reality and Computer-Assisted Design. Virtual Reality (VR) denotes a Computer Graphics environment in which a user interacts with geometric representations of realworld or model objects, utilizing tools and performing manipulations that emulate physical tools and actions [5]. In a Computer-Assisted Design (CAD) environment, engineering principles are implemented that permit quantitative construction and mechanical analysis of user-designed objects[6]. These features turn out to be of major importance during fossil reconstruction. It is possible to plan every single reconstructive step according to predefined quantitative criteria. The necessity to explicitly formulate each stage of reconstruction renders the procedures transparent and accessible to examination or replication by other researchers. Furthermore, by performing alternative reconstructions and comparing them to each other, it is possible to evaluate the reliability of the process as a whole and to provide a range of possible reconstructions. 2.3

Reconstruction of fossil morphology Classical physical reconstruction of a fossil is a process during which fragments are isolated from rock matrix, assembled and completed to yield a best approximation of the in vivo state of skeletal morphology. This procedure seems relatively straightforward, but it is fraught with difficulties and limitations that have consequences for the interpretation of the inferred morphology. This is of special significance in paleoanthropology, where inferences drawn from fossil morphology are traditionally far-reaching with respect to the scarcity of the available material [7]. Virtual fossil reconstruction, on the other hand, combines anatomical and computational

92

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

considerations throughout the whole process and attempts to establish criteria of reliability for the reconstructed morphology [8]. 2.4

Correction of taphonomic deformation To correct taphonomic deformation of a fossil specimen, it is first necessary to distinguish between in vivo and post mortem causes that might bring about deformation. With respect to the analysis of hominid fossils, a temporal sequence of potential causes and events bringing about deformation has to be considered. Among post mortem effects, we will focus here on identification and correction of taphonomic deformation. From a physical point of view, the mechanical properties of the fossilizing bone influence the way in which a fossil is deformed. We may distinguish between two basic scenarios of deformation induced by forces exerted by the surrounding strata: In the first case, the fossil undergoes plastic deformation. Specimens deformed in this way tend to be recovered in relatively few pieces and retain distortions after reconstruction. In the second case, fracturing deformation, the fossil breaks apart during diagenesis. The formation of fractures induced by load indicates resistance against plastic deformation through localization of peak forces within the cracks and dislocation of parts. As a consequence, fossils that are heavily fragmented in situ tend to exhibit little overall deformation after virtual reconstruction. In the latter case, correction of taphonomic deformation is relatively straightforward. In the former situation, however, any attempt at inferring the original form of a fossil from its deformed state remains tentative. Nevertheless, the problem is tractable for a simple and probably common scenario: vertical compression of fossil-bearing strata. To correct deformation, the virtual fossil is positioned in situ on the computer screen and decompressed until anatomical mirror symmetry is restored (Fig. 2). 2.5

Fossil data bases During subsequent stages of CAP, various data sets are produced, each of which describes particular properties of the original objects: volumetric data of the X-ray density (i.e. serial CT images), segmentation data to separate fossils from surrounding matrix, 3D surface data for morphometric analyses, object positions in space that document different states of the reconstruction, and morphometric data. Archiving and retrieval of specific data sets plays a key role during fossil reconstruction. For example, comparative 3D surface data from similar specimens can be used to complete missing regions by application of morphing procedures. The potential benefits of this endeavor are obvious. Comparative morphometric studies on fossil hominids would greatly profit from larger sample sizes, and attempts at reconstruction of missing parts could extrapolate information from similar specimens in the database. Further, the risk of damage to original fossil specimens can be reduced, as any desired action such as taking measurements or performing alternative reconstructions can be carried out on the basis of the digital information already present in the data base. This might represent an important step towards facilitating access to original fossil specimens without exposing them to unnecessary physical risk during handling. 2.6

From Virtual Reality to Real Virtuality Virtual fossils prepared and reconstructed on a computer can be transferred back to physical reality. As opposed to Virtual Reality, Real Virtuality (RV) denotes an environment where a user interacts with physical models of 3D objects generated or modified by computer-assisted procedures. Currently, the most accurate automated replication technology available is laser stereolithography, an industrial technology that was originally devised for physical modeling of CAD-generated parts. Objects are built through consecutive polymerization of thin layers of a photosensitive liquid resin. The process resembles that of building topographical models through piling layers of cardboard. A computer-guided UV laser beam traces an outline and cross-hatches onto the surface of the resin, inducing local photopolymerization (i.e. hardening) according to the desired object structure. Building cross-sections one above the other yields models of arbitrary topological complexity. As the polymerized resins exhibit virtually no shrinkage, the accuracy of stereolithographic models matches or even surpasses that of conventional casts.

ZOLLIKOFER et al.: Computer-Assisted Paleoanthropology: Methods, Techniques and Applications

93

Stereolithographic hard-copies can be generated at different stages during the process of fossil reconstruction to complement virtual object manipulation and to monitor complex “docking” tasks such as re-establishing dental occlusion or checking the “goodness of fit” between adjacent fragments. Handling and exploring data in the form of physical objects instead of manipulating them on the computer screen offers improved perceptual equivalence: while true spatial vision can be achieved in Virtual Reality using stereo spectacles and monitors, tactile information is hardly available in a realistic form. Stereolithography therefore represents a valuable non-invasive alternative to traditional molding and casting techniques [9]. 2.7 Morphometric analysis Using VR models of fossils, morphometric characteristics can be determined in one, two and three dimensions. The spatial position of classical landmarks can be established and inter-landmark distances and angles can be derived. Features that are easy to define but difficult to measure conventionally – such as surface areas, object thickness and object volumes (cavities are identified by their negative volumes) – can be determined. Further, complex parameters such as characteristics of surface curvature, can be evaluated Deformational procedures can be used to compare homologous morphologies by transforming one object into another and similar procedures can be used to simulate growth processes. An additional possible application of computer-assisted procedures is extrapolation of missing anatomical structures on the basis of comparative data from more complete fossil specimens and/or modern human data sets.

3

Applications: Neanderthal reconstruction and morphometry

There is an ongoing debate about the evolutionary and functional significance of the morphological differences between Neanderthals and modern humans, especially with respect to the question of possible speciation events in the recent evolutionary history of Homo. Although Neanderthals can generally be distinguished from modern humans by a set of autapomorphic characters [10-12], there is a particular need for new quantitative data documenting character variation within and between groups. One important prerequisite of acquisition of new morphometric data consists in providing reliable reconstructions of the fossil specimens. – We report here on the computer-assisted reconstruction and morphometry of two fragmentary crania, the juvenile Gibraltar 2 and the adolescent Le Moustier 1 specimens. 3.1

Gibraltar 2 (Devil’s Tower) Five individual fragments represent the Devil’s Tower (Gibraltar 2) Neanderthal child skull: an incomplete mandible, the right maxilla, the right temporal, the frontal, and the left parietal [13]. To establish regions of anatomical contact between the isolated fragments, we completed missing parts through mirror-imaging of existing fragments. After rebuilding the right mandibular premolars using mirror imaging of the existing left teeth, dental occlusion with the upper jaw fragment could be established. At this stage, stereolithographic copies of the jaws were generated to check the accuracy of dental occlusion. In the next reconstructive step, the semicircular canals of the preserved right inner ear cavities served as an anatomical compass to orient the temporal bone along the sagittal plane of the skull, defined by an angle of 45° relative to the planes of the superior and posterior semicircular canals. The oriented temporal bone and its mirror copy were then placed on the mandibular condyles. Finally, the temporoparietal suture between the mirrored temporal and the original parietal bone was used to determine the anatomically appropriate position of the cranial vault bones (Fig. 1).

ACTA ANTHROPOLOGICA SINICA

94

Figure 1

Supplement to Vol. 19, 2000

Computer-assisted reconstruction of the Gibraltar 2 cranium (crossed stereo pictures). The five original fragments (dark) were positioned in anatomical space establishing points of contact between isolated fragments and mirrored components (transparent), and using internal clues (such as the preserved internal otic structures of the right temporal bone, shown here in the mirror-imaged left counterpart)

To check on the general reliability of the Devil’s Tower reconstruction, notably because of potential deviations from bilateral symmetry, parallel reconstructions were conducted using complete skulls of modern human children of comparable dental age and exhibiting a normal degree of bilateral asymmetry. In these reconstructions, only the parts corresponding to the fragments preserved in the fossil were utilized, following exactly the same procedures as for the Neanderthal child. Comparison of the original modern human skulls with the resulting reconstructions showed relatively little deviation. Measurements taken on different versions of reconstructions suggest that reconstructive errors are in the same range as anatomical departures from bilateral symmetry. 3.2

Le Moustier 1 The Le Moustier 1 fossil represents the most completely preserved adolescent Neanderthal skeleton recovered to date, although much of the original material had been lost during World War II. Restoration and reconstruction of this specimen faced a number of difficulties that are intimately connected to its convoluted history. The present state of the cranial remains is the result of at least four earlier reconstructions during which the original fragments were repeatedly disassembled and recomposed [14]. In its current physical reconstruction, the skull exhibits considerable overall deformation and anatomical inconsistencies that need correction. Moreover, the original fossil components are camouflaged with filling material, but actual physical disassembly would subject the specimen to unnecessary risk. We therefore applied our non-invasive procedure to the skull in order to generate a new reconstruction and to extract additional information [15].

ZOLLIKOFER et al.: Computer-Assisted Paleoanthropology: Methods, Techniques and Applications

Figure 2

95

Reconstruction of the Le Moustier 1 cranium. Following computerized decomposition and "cleaning" of the original fragments (top left), the skull was recomposed on the computer screen (right, scale bar is 50 mm), and distortion caused by vertical compression was corrected (bottom left)

Our virtual reconstruction of Le Moustier 1 (Fig. 2) proceeded as follows: Using CT-based 3D data, the specimen was freed from heterologous material, disassembled into its almost 100 original fragments. Following similar criteria to those established for the Gibraltar 2 reconstruction, the fragments were recomposed on the computer screen. During this process, distortions present in the current reconstruction could be eliminated by re-establishing correct anatomical correspondences between adjacent fragments. However, once completed, the virtual reconstruction still exhibited overall deformation. The slanted appearance of the cranium clearly reflected taphonomic effects that resulted in plastic deformation, notably of the cranial vault bones. We used historical photographs to determine the taphonomic in situ position of the skull. Combining morphometric and taphonomic evidence, it turned out that the skull had undergone vertical compression along an axis leading from the left frontal to the right occipital poles. This effect could be reversed on the computer screen by positioning the virtual skull in situ and applying appropriate decompression. 3.3

Morphometry: estimating cranial capacities of incomplete specimens To assess cranial capacity of incomplete Neanderthal specimens, we attempted to reconstruct missing parts by adjusting complete endocasts of modern human skulls of comparable individual age (Fig. 3). For this purpose, a series of landmarks was identified on the preserved endocranial parts of the Neanderthal skulls, and homologous landmarks were determined on the modern counterparts. Applying the 3D thin plate splines morphing technique proposed by Bookstein [16], the modern landmark constellation was transformed into the Neanderthal constellation and the modern endocranial volume was deformed accordingly. The resulting cranial capacities are 1230-1250 cc, 1370-1420 cc and 1550-1600 cc for Gibraltar 1, Gibraltar 2, and Le Moustier, respectively. It is worth noting that cranial capacities have also been determined in this way for two additional

ACTA ANTHROPOLOGICA SINICA

96

Supplement to Vol. 19, 2000

Neanderthal skulls from immature individuals comparable in age to the Gibraltar 2 skull: 1440 cc for the Engis specimen and 1325 cc for the Roc de Marsal specimen [8,10].

Figure 3

Interpolation of the Le Moustier 1 endocranial cavity by morphing data from a modern human. Equivalent landmarks on the endocasts of a modern human adolescent (left) and of Le Moustier 1 (mid) were determined. To adjust the complete modern endocast to the preserved structures of the virtual Neanderthal endocast, a 3D Thin Plate Spline mapping function was applied (right)

4

Conclusions

New possibilities for paleoanthropology have been opened up by recent advances in medical imaging technologies, in computer graphics technology and in rapid prototyping technology (stereolithography). In combination, these advances have permitted development of an entirely noninvasive 3-phase procedure of data acquisition, virtual reconstruction/morphometry and stereolithographic replication of fossils. With computer-assisted paleoanthropology (CAP), it is now possible to return to long-known fossil specimens, subjecting them to re-examination and extracting extensive additional information. Acknowledgments: We would like to thank Jeffrey Schwartz and Fred Spoor for reviewing the manuscript and providing helpful comments. Our research was supported by Swiss NSF grants #3132360.91 and #31-42419.94 to R.D. Martin and P. Stucki. References: [1] CONROY GC, VANNIER MW. Noninvasive three dimensional computer imaging of matrix filled fossil skulls by high resolution computed tomography [J]. Science, 1984, 226:457-458. [2] ZONNEVELD FW, WIND J. High-resolution computed tomography of fossil hominid skulls: A new method and some results [A]. In: Tobias PV ed. Hominid Evolution: Past, Present and Future. New York: Alan Liss: 1985, 427-436. [3] CONROY G, WEBER G, SEIDLER H et al. Endocranial capacity in an early hominid cranium from Sterkfontein, South Africa [J]. Science, 1998, 280:1730-1731. [4] PATH M, ZOLLIKOFER CPE, STUCKI P. New approaches in CT artifact suppression - a case study in maxillofacial surgery [A]. In: Lemke HU et al. eds. CAR'98, Computer Assisted Radiology and Surgery. 1998, 830-835. [5] BRESENHAM J, JACOBS P, SADLER L et al. Real Virtuality: StereoLithography - Rapid Prototyping in 3D [A]. SIGGRAPH Proceedings, 1993, 377-378. [6] ZOLLIKOFER CPE, PONCE DE LEON MS. Tools for rapid prototyping in the biosciences [J]. IEEE Computer Graphics and Applications, 1995, 15:48-55. [7] TATTERSALL I. The abuse of adaptation [J]. Evol Anthropol, 1999, 8:115-116. [8] ZOLLIKOFER CPE, PONCE DE LEON MS, MARTIN RD. Computer-assisted paleoanthropology [J]. Evol Anthropol, 1998, 6:41-54.

ZOLLIKOFER et al.: Computer-Assisted Paleoanthropology: Methods, Techniques and Applications

97

[9] SEIDLER H, FALK D, STRINGER C et al. A comparative study of stereolithographically modelled skulls of Petralona and Broken Hill: implications for future studies of middle Pleistocene hominid evolution [J]. J Hum Evol, 1997, 33:691-703. [10] ZOLLIKOFER CPE, PONCE DE LEON MS, MARTIN RD et al. Neanderthal computer skulls [J]. Nature, 1995, 375:283-285. [11] HUBLIN JJ, SPOOR F, BRAUN M et al. A late Neanderthal associated with Upper Palaeolithic artefacts [J]. Nature, 1996, 381:224-226. [12] SCHWARTZ JH, TATTERSALL I. Significance of some previously unrecognized apomorphies in the nasal region of Homo neanderthalensis [A]. Proceedings of the National Academy of Science of the USA, 1996, 93:10852-10856. [13] GARROD DAE, BUXTON LHD, SMITH GE et al. Excavation of a Mousterian rock-shelter at Devil's Tower, Gibraltar [J]. J Royal Anthropol Institute, 1928, 58:33-113. [14] WEINERT H. Der Schädel des Eiszeitlichen Menschen von Le Moustier in Neuer Zusammensetzung [M]. Berlin: Springer, 1925. [15] PONCE DE LEON MS, ZOLLIKOFER CPE. New morphometric evidence from Le Moustier 1: Computer-assisted reconstruction of the skull [J]. Anatomical Record, 1999, 254:474-489. [16] BOOKSTEIN FL. Morphometric Tools for Landmark Data [M]. Cambridge: Cambridge University Press, 1991.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

98-103

Variability of Pliocene Lithic Productions in East Africa Hélène ROCHE (Préhistoire et Technologie - UMR 7055 du CNRS, MAE(3), 21 allée de l'Université 92023 Nanterre, France)

Abstract Pliocene lithic productions are caracterized by cores, flakes and flake fragments, unmodified blocks and cobbles, and, sometimes, hammerstones. At first sight, these lithic assemblages appear as simple and static technical productions, not only during Pliocene times but, for some authors, during Early Pleistocene times as well, until the appearance of the Acheulean at 1.7/1.6 Myr. However, cores vary from cobbles and blocks from which only a few flakes are removed to intensively flaked cores, either randomly or following a more systematic knapping procedure. Only the 2.3 Myr site of Lokalalei 2C (LA2C) of the Nachukui Formation (West Turkana, Kenya) presents such an elaborate debitage scheme, which is demonstrated by more than 60 refits. If differences observed within the Pliocene lithic assemblages likely reflect environmental constraints (such as raw material avilibility), we will discuss the fact that they may also result from diversified cognitive capacity and/or technical choices, among different hominid species or within the same hominid species.

Key words:

East Africa; Pliocene; Lithic assemblages; Technology

Compared with paleontological and paleoanthropological records, pliocene archeological data unearthed in East Africa are still few. Although it has been known for almost 25 years that group(s) of hominids have made stone tools as early as 2.6 Myr, only a small number of securely dated pliocene sites have been discovered and excavated. They are distributed in two areas : along the Awash River in northern Ethiopia at 2.6-2.5 Myr, and within the Turkana Basin (southern Ethiopia-northern Kenya) at 2.4-2.3 Myr. The next group of sites known are the early pleistocene oldowan sites, at 1.9-1.8 Myr. The Pliocene archaeological record thus appears as a rare and discontinuous phenomenon. There are several possible reasons for this apparent discontinuity. 1/ First of all, it may be simply due to a lack of sufficient research within the appropriate deposits. 2/ The second reason -or group of reasons- concerns methodological principles of precaution, that I have already mentionned elsewhere [1] but which should be reiterated, given the contents of several "oldest" annoucements recently made. Indeed, it seems difficult to speak of an industry, or human -or hominid- stone tool manufacture at a given place if we are not absolutely certain about the following points : -

-

the unquestionable intentionnal manufacture of the artefacts, which should form a coherent technological assemblage, even with a small number of pieces; the reliability of the context, meaning that the material is in situ in primary context, or possibly in secondary context, with the condition that the stratigraphy of the reworked deposits is well understood and its age well established; the dating of the archaeological horizon, or embedding sediments, by several crosschecked methods.

It is only through the convergence of these different elements that a lithic assemblage can be qualified and dated. Due to its geological configuration, this convergence is certainly easier to obtain in the East-African Rift Valley, than anywhere else. However, the prehistory of this region is not exempt from misinterpretations and controversies (the most famous one being the dating of KBS tuff). Thus, precisely locating archaeological -and paleoanthropological- sites within a reliable chronological and paleoenvironmental framework has been, and still is, a long and painstaking, but absolutly necessary task. Biography: Hélène Roche is a french archaeologist attached to the Centre National de la Recherche Scientifique (CNRS). She has been conducting field work in East Africa (Ethiopia, Kenya, Uganda, etc.) for the last 25 years. From 1991 to 1994, she was Head of the Archaeology Department of the National Museums of Kenya, in Nairobi (Kenya).

ROCHE: Variability of Pliocene Lithic Productions in East Africa

99

3/ The last group of reasons is more speculative. The hominids present in East Africa between 3 and 1.6 Myr (thus before the appearence of Homo ergaster/erectus) are certainly more diverse than previously thought. However the taxonomic link between these different groups of hominids, whether Australopithecine or early Homo, is not known, nor is their actual appearance, duration and extinction. We do not know their exact geographical distribution, nor their demography. It is therefore not impossible to consider that some of these groups had mastered the knowledge of stoneknapping, while others had not. And it is not impossible that this knowledge could have disappeared in a given area for a certain time, to reappear later on. In this paper I will discuss how these early but scattered pliocene lithic productions can be considered in a technological perspective The oldest evidence of stone tool manufacture comes from the Kada Gona sites, in the Hadar region of the Upper Awah Valley (northern Ethiopia), with several sites securely dated at 2.6/2.5 Myr. The first artefacts discovered on the left bank (Kada Gona 2.3.4 site) were included in a conglomerate (named the Intermediate Conglomerate in all the sections drawn in this area), which is a rather unreliable archaeological context [2-3]. However this conglomerate had the special feature of being bordered by two tuffs, which allowed the dating of the artefacts. Subsequently, the paleomagnetic Gauss/Matuyama transition was identified within the Intermediate Conglomerate [4]. Other sites (EG -for East Gona-10, EG 12 sites) in much better clayey contexts were discovered and excavated later on [4]. The few Kada Gona 2.3.4. 'Pre-acheulean' artefacts were described as cores, flaked blocks and cobbles, and flakes. They were said to be "techonologically...closer to the Oldowan of Olduvai (Bed I) and Gombore 1 at Melka Kunture than to the Lake Turkana assemblages -notably the KBS industry∗ ...-and from the Omo sites..." [3]. Work is continuing at the East Gona sites, but the only information thus far available concerning the artefacts comes from the first lithic assemblages unearthed at the beginning of the 1990ies (n = 2970 stone artefacts). These consist of "simple cores, whole flakes and flaking debris. Unifacially and bifacially flaked cores comprise the 'flaked pieces'...'Detached pieces' are numerically dominant, with values in the range of 75-95%. There are a few 'pounded pieces', namely pieces modified or shaped by pounding or battering, like hammerstones, anvils and battered cobbles" [4]. It has also been said that "the composition of Gona assemblages is very similar to Plio-Pleistocene sites elsewhere, except for lower diversity of the cores from Gona and the high incidence of utilized pieces and manuports at Olduvai Gorge" [4]. Further south, the Turkana Basin constitutes an exceptionnal ensemble of sedimentary formations, which all together cover more than 4 Myr, with : -

to the north, the Shungura Formation, which is traversed by the Omo river to the east, the Koobi Fora Formation and to the west, the Nachukui Formation

Through tephrostratigraphy and sedimentary marker correlations, the chronostratigraphy of the three formations is now well established. This is the result of intense work by many geologists (Frank Brown, Ian Mc Dougall, Craig Feibel, etc.) over the last 20 years [5-6]. The Shungura Formation has yielded only pliocene archeological sites, the Nachukui Formation pliocene and pleistocene sites, and the Koobi Fora Formation only pleistocene sites . I will thus limit my comments to the Shungura and Nachukui formations. The Shungura Formation is made of 12 members but only member F (2.36-2.32 Myr) contains 5 archaeological sites in more or less good sedimentary context but in secure chronostratigraphic position [7]. The lithic material is made of quartz and include very few cores and whole flakes but mainly fragments and flaking debris (up to more than 95% of the whole assemblage). There is absolutely no doubt about the intentional character of these lithic assemblages, but they rather result from random percussion than from controlled knapping [8-9]. ∗

This was long before the beginning of the work on the west side of Lake Turkana

ACTA ANTHROPOLOGICA SINICA

100

Supplement to Vol. 19, 2000

The Nachukui Formation is located on the west side of the basin, and extends between the ranges that border the basin to the west and the modern Lake Turkana. The Nachukui Formation has a cumulated thickness of 730 m and consists of 8 members. In addition to a wealth of fossils (including the Australopithecus Boisei "Black skull" specimen -WT17000- and the almost complete skeleton of a young Homo ergaster -WT15000-), the Nachukui Formation has yielded many archaeological sites distributed along the sequence between 2.35 Myr and 0.7 Myr.[10-13] So far, more than 25 archaeological sites have been test-excavated or fully excavated, including: -

2 "Pre-oldowan" sites at 2.34 Myr, 12 Oldowan sites between 1.8 and 1.65 Myr 1 very early Acheulean site at 1.65 Myr 10 middle pleistocene Acheulean sites

The two pliocene sites, Lokalalei 1 and Lokalalei 2C, are stratigraphically located in the upper part of a succession of fluvial facies separated by lacustrine deposits; they are embedded within paleosols formed in a fluvial flood plain, reflecting proximal low energy environnements. Lokalalei 1 has been excavated over 67 m2 [12] and presents a low density of artefacts (n=466), compared to a large number of faunal remains (more than 3500), most of which are non identifiable splinters. The lithic assemblage consists of cores (mainly described as chopper-cores, but also discoï dal and polyhedral cores, and core scrapers), whole flakes, broken flakes and fragments, and pounded pieces (including hammerstones). The limited number of flake scars on the cores (1 to 12) and the fact that "about 80% of the flaking scars on these cores are characterized by step fractures and only a few instances of complete flake removals were observed" leads to the conclusion that the whole assemblage exhibits "crude and poor technology" [12]. Lokalalei 2C is a much smaller site (17 m2) which can be considered as a knapping location. It contains cores, whole flakes, broken flakes and fragments, worked and unworked cobbles, and hammerstones, representing a total of 2500 artefacts associated with nearly 400 faunal remains. The exceptionnal preservation of the lithic assemblage permitted the reconstitution of 59 sets of refits [13]. Refitting is certainly not an aim in itself. However what refits can tell us about pliocene hominid technological skill is extremly interesting. At LA2C, the dominant reduction sequence of the cores (78%) consists of unidirectional or multidirectional removals flaked on a single preferential flat surface, from natural platforms. This knapping scheme implies the selection of cobbles with a specific morphology -with a triangular or quadrangular section and at least one flat surface- and a natural, adequate platform. It allows the production of series of removals (up to 11 removals per series), with the condition that the volumetric structure of the core is maintained. The repeated application by the knappers of the same, well mastered technical principles during the whole knapping sequence permitted the production of large numbers of flakes (up to 50 for one core). This may be an indication that the notion of production was already understood by a group of hominids in this particular area. However, the same core reduction scheme is not seen at the nearby and contemporaneous site of Lokalalei 1, where others clearly less productive technical options were chosen. Nor it is in the contemporaneous Omo sites, and it does not seem to be present in the Kada Gona sites. Thus, if there are such observable technological differences between those lithic productions, how can we assess them, and what do they reflect? More generally, lithic technical productions are materialized by three main features (tab.1): A) the production mode of artefacts B) the technique(s) used to produce them C) the "tools" which are produced In lithic technology, only four modes of production are known [14] :

ROCHE: Variability of Pliocene Lithic Productions in East Africa

101

1) undifferenciated flaking: intentional fracture of hard rock with no specific or systematic operative procedure; 2) debitage process: intentional fracture of hard rock with specific and sometimes highly systematic operative procedures, and with the aim directly using, further shaping or retouching the flakes produced; 3) shaping: sequence of knapping actions which imposes a particular form on a piece of raw material; 4) retouching: modification of a blank, whether natural or knapped. Table 1 TECHNICAL PRODUCTION IS MATERIALIZED BY A - one or several production modes B - one or several techniques C - the produced "tools" A - PRODUCTION MODE OF LITHIC ARTEFACTS

1 - undifferenciated flaking 2 - debitage process 3 - shaping 4 - retouching B - POSSIBLE TECHNIQUES USED DURING PLIO-PLEISTOCENE TIME INTERVAL

1 - direct percussion with stonehammer 2 - direct percussion on anvil and block 3 - bipolar percussion on anvi/block C - MANY POSSIBLE COMBINATIONS BETWEEN A AND B TO OBTAIN THE DESIRED PRODUCT

All these different production modes are implemented by techniques, and some of them by methods [14]. A method implies carrying out an orderly sequence of actions following a elaborate plan, according to one or more techniques (ex.: Kombewa method, Levallois method). Thus method does not apply to pliocene production modes. Techniques refers to the basic physical actions employed by the prehistoric stone knappers (direct percussion with a hard or a soft hammer, indirect percussion with the interposition of a punch, etc.). During prehistoric times, many different techniques were invented and used, most of which have been identified through observation and experimentation. However, for the time period we are considering, only three main techniques seem to have been used : 1) direct percussion with a stone hammer 2) direct percussion on stone anvil/block 3) bipolar percussion on a stone anvil/block and with a stone hammer Many combinations are possible between the production mode and the techniques used, and diagnostic traces resulting from direct percussion with a stone hammer or direct percussion on stone anvil/block are not always easy to differenciate. However, Tabl. 2 shows which products can be obtained with the different production modes.

ACTA ANTHROPOLOGICA SINICA

102

Supplement to Vol. 19, 2000

a) Direct undifferenciated flaking will produces cores or chopper-cores, and flakes. In the case of chopper-cores, it is difficult to ascertain what is the desired product and what is the waste. b) Controlled debitage produces flakes and cores, and in this case we know that flakes are the desired products, and cores the waste. c) Shaping produces all the shaped tools, such as polyhedrons and spheroids, handaxes and cleavers. Flakes are considered as waste, although they can be used as cutting tools, whether unmodified or retouched. d) Retouching produces sundry tools, such as scrapers, notches, denticulates, etc. Table 2 PRODUCTION MODE

TECHNIQUE

DESIRED PRODUCTS

WASTE

1 - indifferenciated flaking

1,2,3?

classical chopper cores?

Flakes?

flakes?

Cores?

1,2,3?

flakes

Core

1,2?

polyhedral and spheroidal shaped tools

Ordinary or shaping flakes

3b - bifacial shaping

1,2?

bifacial shaped tools

Ordinary or shaping flakes

3c - trihedral shaping

1,2?

trihedral shaped tools

Ordinary or shaping flakes

1,2?,3?

scrappers, notches, denticulates, etc.

Retouch flakes

2 - debitage 3 - shaping 3a - polyhedral and spheroidal shaping

4 - retouching (flaked or non flaked blanks)

If these different production modes are considered from a chronological perspective, we can see that: 1) undifferenciated flaking is present anywhere and at all times and from the beginning at 2.6 Myr; 2) controlled debitage is only present in West Turkana at 2.34 Myr and solely in one site; 3) retouching is certainly more present than recorded in the publications, and is certainly present as early as 2.34 Myr in West Turkana (at both pliocene sites of LA1 and LA2). 4) and shaping does not appear before 1.9-1.8 Myr with polyhedral and spheroï dal shapping (at Olduvai), then at 1.7-1.6 Myr with bifacial shaping (in West Turkana). There is agreement on the fact that the first major technological change in the pliopleistocene record corresponds to the beginning of bifacial technology. However, before this innovation took place, the lithic productions did not reflect the same hominid cognitive capacities and motor skill, whether in Gona, in the Omo, or in Nachukui. Indeed, environmental constraints may have played a role, in particular as far as raw material is concerned. In the Shungura Formation (Omo), it seems that the only available raw materials were small blocks of quartz, which has poor flaking properties. However, there is an equivalent availibility of good lava raw materials in Hadar (Gona) and in Nachukui. On the other hand, the discontinuity of the pliopleistocene archaeological data mentionned in the introduction certainly amplified the contrast between the different assemblages.

ROCHE: Variability of Pliocene Lithic Productions in East Africa

103

In conclusion, given the variability observed in the chaines operatoires of lithic productions, it is not possible to confine pliocene and early pleistocene productions to a single vast technocomplex [4], within which stone-knapping did not vary and did not evolve. Is this so surprising, considering the period of time in question (more than one million years), the distance between the concerned areas (hundreds miles), and the increasing diversity of hominids forms present during this period of time in these areas? Acknowlegments : We thank the Government of Kenya for permission to carry out this research and the National Museums of Kenya, Meave Leakey and the Kalokol Project for logistical support. The field work was funded by the French Foreign Affairs Ministry (Sous-Direction des Sciences Sociales, Humaines et de l'Archéologie). We thank Total (Kenya) for vehicles and fuel donation, and Crédit Agricole-Indosuez Bank (Kenya) for financial support. We thank Jacques Pelegrin and Megan O'Farell for reviewing this paper. References: [1] ROCHE H. Remarques sur les plus anciennes industries en Afrique et en Europe [A]. XIII° UISPP Congress, Colloque VIII : Lithic industries, language and social behaviour in the first Human forms. Forlí, Abaco, 1995, 55-68. [2] ROCHE H, Tiercelin JJ. Découverte d'une industrie lithique ancienne in situ dans la formation dHadar, Afar central, Ethiopie [J]. C R Acad Sc Paris, Série D, 1977:1871-1874. [3] ROCHE H, TIERCELIN JJ. Industries lithiques de la formation plio-pléistocène d'Hadar : campagne 1976 [A]. Proceedings, VIIth Panafrican Congress of Prehistory and Quaternary Studies (Nairobi 1977), 1980, 194-199. [4] SEMAW S, RENNE P, HARRIS JWK et al. 2.5-million-year-old stone tools from Gona, Ethiopia [J]. Nature, 1997, 385:333-336. [5] FEIBEL CS, BROWN FH, MCDOUGALL I. Stratigraphic context of hominids from the Omo group deposits : northern Turkana Basin, Kenya and Ethiopia [J]. Am J phys Anthrop, 1989, 78:595-622. [6] FEIBEL CS, HARRIS JM, BROWN FH. Palaeoenvironmental context for the late Neogene of the Turkana Basin [A]. Koobi Fora Research Project, Volume 3, Stratigraphy, Artiodactyls and Paleoenvironments. Oxford : Clarendon Press, 1991, 321-346. [7] HOWELL FC, HAESAERTS P, DE HEINZELIN J. Depositional environments, archaeological occurences and hominids from members E and F of Shungura Formation (Omo Basin, Ethiopia) [J]. J Hum Evol, 1987, 16:643-664. [8]CHAVAILLON J. Evidence for technical practices of early Pleistocene hominids [A]. Earliest Man and Environments in the East Rudolf Basin. Chicago : Chicago University Press, 1976, 565-573. [9]MERRICK HV, MERRICK JPS. Archaeological occurences of earlier Pleistocene Age, from the Shungura Formation [A]. Earliest Man and Environments in the East Rudolf Basin. Chicago : Chicago University Press, 1976, 574-584. [10] KIBUNJIA M, ROCHE H, BROWN FH et al. Pliocene and pleistocene archaeological sites west of lake Turkana, Kenya [J]. J Hum Evol, 1992, 23:431-438. [11] ROCHE H, KIBUNJIA M. Les sites archéologiques plio-pléistocènes de la Formation de Nachukui, West Turkana, Kenya [J]. C R Acad Sc Paris, 1994, 318 ( série II): 1145-1151. [12] KIBUNJIA M. Pliocene archaeological occurences in the lake Turkana Basin [J]. J Hum Evol, 1994, 27:157-171. [13] ROCHE H, DELAGNES A, BRUGAL JP et al. Early hominid stone tool production and technical skill 2.34 Myr ago in West Turkana, Kenya [J]. Nature, 1999, 399:57-60. [14] INIZAN M-L, BALLINGER M, ROCHE H et al. Technology and terminology of knapped stone [M]. Préhistoire de la Pierre Taillée, 5. Paris, CREP, 1999.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

104-114

Greeting Chinese Paleolithic Archaeology in the 21st Century (A Retrospective) HUANG Wei-wen (Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, P.O. Box 643, Beijing, 100044 China)

Abstract The systematic Paleolithic research in the early 20th Century in the northern part of China resulted in the discovery of several important Paleolithic sites. It initiated the hominid Asian origin theory which dominated later dacades. With the new and much earlier Paleolithic discoveries in East Africa in the middle of 20th Century, the hominid African origin idea took the leadership in the field. But with the recent new and encouraging Paleolithic discoveries dated more than 2 mya in China, the multi-regional origin theory seems more acceptable. Based on the environmental background to the hominid origin and evolution, China appears to be a promising location for the discovery of earlier Paleolithic sites in the 21th Century.

Key words:

Paleolithic; China; The 21th Century

1 In the early 20th Century, some western scholars seeking the homeland of humans began to turn their attention toward Asia, a vast, remote and mysterious land. They hypothesized that Asia was a critical region in the development of humans, both biologically and culturally. The complex question of human origins could not be answered without taking into account the antiquity of human occupation and the early stone industries of Asia [1]. Other than E. Dubois's important discovery of Homo erectus fossils in Java at the end of the 19th Century, little was known about the Asian fossil and archaeological records. This mystery inspired several western scientists to plan campaigns of exploration to this region. Among the first foreign scientists to investigate the Paleolithic localities of China were Emile Licent and Teilhard de Chardin from France. They discovered archaeological materials from Qingyang (Gansu), Shuidonggou (Ningxia), Salawusu (Inner Mongolia) and other sites in the upper-middle reaches of the Yellow River of North China. In 1921 and 1923, the historic discoveries of Swedish geologist, John Gunnar Andersson, at the cave of Zhoukoudian near Beijing initiated the first international cooperative excavation of the famous Peking Man site. This research program that began in 1927 was supported by the Rockerfeller Foundation. It reinforced the idea that Asia had an important place in the story of human origins and scholars worldwide were attracted to the region. Twenty years later, by the 1950's, the focus of human evolutionary studies shifted toward Africa as the newly established the People's Republic of China became inaccessible to western scientists. In the 1960's, important discoveries in Africa included human fossils from the Early Pleistocene strata of Olduvai Gorge. Fossils of robust Australopithecus and Homo habilis from these beds were dated to 1.8 mya, almost 1 million years earlier than the Javan fossils and Chinese Homo erectus from Zhoukoudian. Later discoveries of Homo habilis and stone tools dating as early as 2.6 mya from Ethiopia, Zaire, Kenya and other areas in Africa convinced most researchers that Africa was the cradle of humans. The antiquity of findings outside Africa, particularly in Asia was questioned and most assumed that evidence of human occupation earlier than 1 mya in East Asia, would not be found. American anthropologist, Hallam Movius, proposed a "Two-Culture" Theory in the 1940's that was used to reinforce the above conclusion [2]. According to this theory, there were two groups of Homo erectus in the Old World, each with different cultures. One group possessed the Acheulean or Mode II technology, making bifacial handaxes in a stepwise, standardized procedure. The other group made less refined chopper-chopping tools (Mode I technology), interpreted as Biography: HUANG Wei-wen, Reserrch Professor at the Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, specialised in Paleolithic research since 1960.

HUANG:

Greeting Chinese Paleolithic Archaeology in the 21th Century (A Retrospective)

105

unimaginative and monotonous. Concluding that the two populations were separated by different levels of cognitive capabilities, Movius divided the Old World into two regions by drawing an intangible technological barrier, later called the "Movius Line." East of that line (that is, East Asia) was the region that Movius and his supporters maintained was isolated and self-sufficient. Here, a Mode I technology persisted until the Late Pleistocene when technological advances like blade industries were imported from the West. Since no Mousterian or Mode III technological stage could be clearly identified in East Asia, the western tripartite classification of the Paleolithic into Lower, Middle and Upper phases are not applicable.

2 From a 21st Century perspective, how have these ideas progressed? In the last few decades, new discoveries and new interpretations from Asian localities have portrayed a very different picture. A brief presentation of some of these new ideas follows. The early human occupation of Asia is temporally and spatially much more extensive than previously thought. Russian archaeologists report stone artifacts from Ulalinka in the Altai Mts. of Siberia in northern Asia. These have been dated by paleomagnetics and TL dating to 0.73 mya and 1.5 mya, respectively. In addition, the Diring-luiakh Site on the upper-middle reaches of Lena River inside the Arctic Circle (paleomagnetic date of 1.8 mya and a TL date of 0.3- 0.4 mya) also supports this broad temporal and spatial range [4]. The prehistory of the Japan Archipelago has also undergone revisions with recent excavations in the northwestern Honshu. Human occupation of the Japan Islands may extend back to the lower Middle Pleistocene [5]. The locality of Chongokni, near Seoul in peninsular Korea, has yielded an Acheulean-like industry with a controversial range of dates from Middle to Late Pleistocene. Similar controversy surrounds the 1.8 mya Ar/Ar dates [6] of Mojokerto and Sangiran in Java that give Homo erectus in Asia an antiquity comparable to its African counterpart.

Figure 1

Pick made of rhinocero’s mandible from Renzidong (after Zhang et al. [3])

Additional evidence of ancient humans in Asia has been found on the Rawalpindi Plateau of Pakistan. Stone tools earlier than 1 mya have been recovered from Riwat, Pabbi Hills and other localities at the foot of the Himalayas. Paleomagnetic studies and a comparison with nearby

ACTA ANTHROPOLOGICA SINICA

106

Supplement to Vol. 19, 2000

volcanic strata estimate that humans may have been present as early as 2 mya on the Indian Subcontinent [7]. Three Lower Pleistocene sites in China discovered in 1960s, Xihoudu, Lantian of North China and Yuanmou of South China, have all yielded evidence older than 1 my. More recently, even greater antiquity (as early as ca. 2 mya) has been suggested for the Paleolithic localities of the Nihewan Basin of North China, together with the Wushan Site and the Renzidong Site of South China. At Renzidong, a mammalian fauna and associated stone and bone artifacts (Fig. 1) were found from cavern deposit and estimated to be 2 - 2.4 my old on the basis of mammal fossils that represent Pliocene species. Clearly, Movius’ ideas are outdated. New lithic Studies of Chinese localities identify more complexity and diversity in the technology. The Wushan stone assemblage contains pick, cleaver, chopper and Kombewa flake (Fig.2).

Figure 2

Pick made of cobble from Wushan (after Hou et al. [8])

Huang [9] identifies handaxes in both North and South China toolkits of the Lower to Middle Pleistocene. Despite its low frequeny of occurrence, it is found with other components of Acheulean toolkits such as picks, cleavers and spheroids. A bifacially flaked handaxe made on a heavy quartzite flake from the loess deposits near the Gongwanling Homo erectus ( “Lantian man” ) site, North China was in a comparable stratigraphic position to the human fossil that is dated to 1.15 mya by paleomagnetic and loess-paleosols sequence. Tattersall et al. [10] considers this an example of East Asian Acheulean tools. Excavations on the laterite terraces of Bose Basin, South China have uncovered an Acheulean-like industry of picks and handaxes (Fig. 3) made on large cobbles and flakes. More than 100 handaxes (ca. 6% of the total tools) that are technologically comparable to Acheulean Mode II have been studied. Tektites in association with these artifacts are dated by fission track and 40Ar/39Ar to 0.733 and 0.803 mya respectively (Guo et al. [11]; Hou et al. [12]), predating European Acheulean tools. The stone industry from the Locality 1 of Zhoukoudian (Peking man Site) has long been used to exemplify the differences between eastern and western Middle Pleistocene cultures. Its apparent ‘lack of handaxes’ claimed by some workers and high proportion of small tools made from vein quartz, was difficult to compare to contemporaneous western sites. Now, we know that assemblages from Olduwai, East Africa and Arago, southwestern France have similarities to Locality 1 and that bifacial retouch is present and well developed at Zhoukoudian (Fig.4).

HUANG:

107

Greeting Chinese Paleolithic Archaeology in the 21th Century (A Retrospective)

A Figure 3

B Handaxes from Gongwangling (A, after Dai [13]) and Bose (B, after Huang and Wang [14])

Figure 4

Stone artefacts from the Locality 1 of Zhoukoudian (after Black et al. [15])

ACTA ANTHROPOLOGICA SINICA

108

Supplement to Vol. 19, 2000

Several late Middle and Upper Pleistocene sites have been well-studied by both eastern and western scholars who have reflected on their technological characteristics. For example, Henri Breuil [1, 32] and L.G. Freeman [16] have commented one after another on the Acheulean-like tools of the Dingcun industry (160-210 kya, U-series) from North China. Breuil characterizes the Late Pleistocene Shuidonggou assemblage (32-40 kya, U-series ) "seemingly halfway between a very evolved Mousterian and a nascent Aurignacian, or a combination of the two"[1] (Fig. 5). Bordes [17] also recognized technological attributes of Levallois in these tools. In fact, developed blade industry associated with microliths also are known from other sites of North China, for example Xiachuan (Fig. 6). Sysmatic excavations at Panxian Dadong, and important new cave site (ca.130-260 kya by U-series and ESR dating ) on the Yunnan-Guizhou Plateau of SW China, is yielding a stone industry associated with a late Middle Pleistocene Ailuropoda-Stegodon fauna and several fossil human teeth. The artifacts, made of limestone, chert and basalt include prepared cores and flakes with faceted striking platforms. The human teeth show a combination of H. erectus and H. sapiens traits [18].

Figure 5

Stone tools from Dingcun (upper, after Pei et al. (below, after Jia et al. [20])

[19]

) and Shuidonggou

HUANG:

Greeting Chinese Paleolithic Archaeology in the 21th Century (A Retrospective)

Figure 6

109

Stone tools from Xiachuan (after Wang et al. [21])

By the Upper Pleistocene, burial and symbolic expression is a feature of several Asian localities. In the 1930's at Zhoukoudian, Upper Cave ornamental objects such as perforated animal teeth , stone beads, bone pendants, shells, hematite and polished antler, were associated with burials [22]. At Xiaogushan Cave in Northeast China, perforated animal teeth, bone needles, a javelin bone point, a bone harpoon and a bone disc with a sun engraving design accompany a rich stone assemblage and mammalian fauna (Fig. 7). These two localities predate the European Magdalenian, with Upper Cave yielding radiocarbon dates (AMS) of 32-24 kya and Xiaogushan dating to 40-30 kya [23]. A fragment of decorated antler of Cervus elephus was found from a cave site (AMS 14C 13 065 +/- 270 B.P.) of Xinglong county, Hebei [24]. The pattern on its surface can be compared with those of European Magdalenian (Fig. 8).

ACTA ANTHROPOLOGICA SINICA

110

Figure 7

Supplement to Vol. 19, 2000

Bone artefacts and ornaments from Xiaogushan (after Huang et al. [25])

HUANG:

Greeting Chinese Paleolithic Archaeology in the 21th Century (A Retrospective)

Figure 8

111

The pattern of decorated antler from Xinglong (after You and Wang [24])

3 The study of Chinese Paleolithic Archaeology, initiated in 1921, has matured through eighty years of scholarly endeavors. Much progress has been made in the quantity and quality of research projects in the past three decades. The new century necessitates a thorough review of the past in order to assure a productive future. The author of this paper suggests that the following two points in summary. 3.1

Suggestions for the theoretical orientation of new research Recently, Chinese geologist Liu Tongshen wrote a postscript for the fourth issue of Quaternary Sciences [26] in which he insightfully summarized a century of research on Chinese loess. He commented that "there are generally two perspectives when doing loess research. One is a ‘regionally oriented perspective’ where the researcher forms hypotheses based on the regional characteristics of loess in China and deductively applies these to loess deposits globally (eg. the aeolian origins of loess deposits). The other is a ‘comparative perspective,’ where data from loess research worldwide can then be applied to the Chinese case by comparison (eg. paleoclimatic information from loess deposits). The two approaches are compatible but each has been more or less popular depending on where and when the research was done. In this author's opinion, the above approaches are relevant to Chinese Paleolithic research. Regional Paleolithic studies must be put into global context. As early as the 1920's Boule, Breuil, Licent and Teilhard recognized that there is not much difference between the Pleistocene in China and that in other countries from the perspective of basic attributes. The difference lies in "quantity rather than quality;" "Chinese loess is an extension of North and West Asian loess, Russian loess, Middle European loess or even that in the various parts and plateaus of North France . From one end to the other, in different regions of the globe, these deposits share common origins." They continued to point that fauna from Chinese loess deposits are similar to those from the Paris Basin and the types of Paleolithic industries in China are fairly similar to those in Europe [1]. These early conclusions about Paleolithic culture in China have stood the test of time. The theoretical construct of the "Movius Line," however, has been detrimental to an understanding of the Paleolithic in East

112

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

and Southeast Asia [27]. This author urges that new discoveries and analytical techniques of the new century be interpreted with an open mind allowing for the formulation of new hypotheses regarding the contributions of the Chinese Paleolithic in global context.

Figure 9

Map showing the paleogeographic outline of marginal seas of the West Pacific during the LGM (after Wang [28]) A. Sea of Okhotsk; B. Sea of Japan; C. Yellow Sea & Bohai Sea; D East China Sea; E. South China Sea; F. Sulu Sea; G. Celebes Sea; H. Banda Sea; I. Java Sea; J. Timor Sea; K. Arafura Sea; L. Gulf of Carpentaria

3.2

Reconsidering the Quaternary Asian paleoenvironment The Quaternary global climate was dynamic with fluctuations in sea level, advance and retreat of continental ice sheets and expansions and contractions of savanna and rainforests. Some researchers maintained that by contrast, Asian Quaternary climates remained relatively stable. Unfortunately, this viewpoint is not conform to the facts showed by the Quaternary research during the past decades, and it has been detrimental an understanding of the geological background of Asian Paleolithic. For example, the pedostratigraphcal, geophysiccal and geochemical research on Chinese loess

HUANG:

Greeting Chinese Paleolithic Archaeology in the 21th Century (A Retrospective)

113

and paleosol sequences document the environmental and climatic changes of the last 2.5 my. These data indicate that there were at least 37 glacial-interglacial cycles [29]. Dramatic fluctuations in sea level are supported by research on the marginal seas of the West Pacific. During the LGM (last glacial maximum ) and the low sea-level period (ca. 20-15 kya) , the China Seas was reduced about one third of its current area (Fig.9). The vegetable cover also accompanied these climatic and sealand changes of the LGM. At the same time, the exposed shelf of the northern South China Sea which is localited in the tropics at present was covered by temperate zone’s grassland with Artemisia based on pollen record from deep sea [30]. The Qinghai-Xizang (Tibet) Plateau and high mountains of East China had developed glacials during the Pleistocene. Even at present, there is an ice bank which covers an area of more than 400 sq. km on 6 000-6 500 m above sea level of the central Xizang, and it is the remains of the Last Glaciation only. The research suggests that four glaciations at least have been recognized from Xizang Plateau since the MIS (marine isotope stage) 24. Among them the 2nd Glaciation which is equal to MIS 16 (ca. 0.7 mya) is the largest one [31]. Clearly, this evidence above mentioned indicates that East Asia was subject to the global climatic fluctuations of the Quaternary. Acknowledgents: I would like to express my most sincere gratitude to Dr. Sari Miller-Antonio (Department of Anthropology and Geography, California State University, Stanislaus) for her kindness in revising the manuscript of the present paper. I also thank my daughter Reiping Huang for her help in translation from Chinese to English of the paper, who is studying in the Department of Sociology, the Minnesota University.

References: [1] BOULE M, BREUIL H, LICENT E et al. Le Paleolithique de la Chine [M]. Archives de Le Institut de Paleotologie Humaine, 1928, Mem 4. [2] MOVIUS HL. The Lower Paleolithic Cultures of Southern and Eastern Asia [M]. Trans Am Philosoph Soc, 1948, N Ser, 33(4):329-420. [3] ZHANG SS, JIN CZ, WEI GB et al. On the artifacts unearthed from the Renzidong Paleolithic site in 1998 [J]. Acta Anthropol Sin, 2000, 19 (3):169-183. [4] DEREV’ANKO A. Introduction. In: Derev’anko A et al. ed. The Paleolithic of Siberia [M]. Translated to English by Inna P Laricheva. Urbana and Chicago: University of Illinois Press, 1998. [5] SAGAWA M. Recent progress in studies on the Early and Middle Paleolithic period of the Japanese Archipelago, and their possible relations with the northern and eastern Asia [J]. Acta Anthropol Sin, 1998, 17(1):1-21. [6] SWISHER III C, CURTIS GH, JACOB T et al. Age of the earliest known hominids in Java, Indonesia [J]. Science, 1994, 263:1118-1121. [7] DENNELL RW, RENDELL H, HURCOMBE L et al. Archaeological evidence for homonoids in northern Pakistan before one million years ago [J]. Courier Forschunges-Institut Senckenberg, 1994, 171:151-155. [8] HOU Y, XU Z and HUANG W. Some new stone artifacts discovered in 1997 at Longgupo, southern China [J]. Longgupo Prehist Culture, 1999, 1:69-80. [9] HUANG W. Bifaces in China [J]. Acta Anthropol Sin, 1987, 6 (1):61-68. [10] TATTERSALL I, DELSON E, COUVERING JV eds. Encyclopedia of Human Evolution and Prehistory [M]. New York and London: Garland Publishing, 1988. [11] GUO S, HUANG W, HAO X et al Fission track dating of ancient man site in Baise, China, and its significances in space research, paleomagnetism and stratigraphy [J]. Radiation Measurements, 1997, 28(1-6):565-570. [12] HOU Y, POTTS R, YUAN B et al. Mid-Pleistocene Acheulean-like stone technology of Bose basin, South China [J]. Science, 2000, 287(5458):1622-1626. [13] DAI E. The Paleoliths found at Lantian man locality of Gongwangling and its vicinity [J]. Vertebr PaleAsiatica, 1966, 10(1):30-32. [14] HUANG W, WANG D. La recherche recente sur le paleolithique ancien en Chine [J]. L’Anthropologie (Paris), 1995, 99(4):637-651. [15] BLACK D, TEILHARD de CHARDIN P, YOUNG CC et al. Fossil Man in China [M]. Geol Mem, 1933, Ser A (11).

114

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

[16] FREEMAN LG. Paleolithic archaeology and paleoanthropology in the People’s Republic of China [A]. In: HOWELLS WW, TSUCHITANI PJ eds. Paleoanthropology in the People’s Republic of China, CSRPC Report. 1977, (4):79-113. [17] BORDES F. The Old Stone Age [M]. New York and Toronto: McGraw-Hill Book Company, 1968. [18] HUANG W, HOU YM, SI X. Stone industry from Panxian Dadong, a cave site of Southeastern China [J]. Acta Anthropol Sin, 1997, 16(3):171-192. [19] PEI WZ, CHIA (JIA) LP. Study of Tingtsun (Dingcun) Palaeoliths [A]. In: PEI W ed. Report on the excavation of Paleolithic sites at Tingsun, Hsiangfenhsien, Shansi province, China. Beijing: Sciences Press, 1958, 97-111. [20] JIA LP, GAI P, LI Y. New materials from the Paleolithic site of Shuidonggou [J]. Vertebr PalAsiatica, 1964, 8(10):7583 (in Chinese with Russian abstract). [21] WANG J, WANG X, CHEN Z. Xiachuanian culture, an investigation at Xiachuan site of Shanxi [J]. Acta Archaeol Sin, 1978, (3). [22] PEI WC. The Upper Cave of Choukoutien (Zhoukoudian) [M]. Palaeontol Sin, 1939, N Ser D (9). [23] HUANG WP, HOU YM. A perspective on the archaeology of the Pleistocene-Holocene transition in North China and the Qinghai-Tibetan Plateau [J]. Quat Internat, 1998, 49/50:117-127. [24] YOU Y and WANG F. An decorated antler from Xinglong county, Hebei province [A]. In: ZHOU G et al. eds. The 60th Anniversary Essays for Discovery of the first Peking man’s skull. Beijing: Science and Technique Press of Beijing, 1992, 38-41. [25] HUANG W, ZHANG Z, FU R et al. Bone artifacts and ornaments from Xiaogushan site of Haicheng, Liaoning province [J]. Acta Anthropol Sin, 1986, 5 (3):259-266. [26] LIU T. Postscipt [J]. Quat Sci, 1999, (4). [27] HUANG W. On the typology of heavy-duty tools of the Lower Paleolithic from East and Southeast Asia (Comment on the Movius’ system) [J]. Acta Anthropol Sin, 1993, 12(4):297-304. [28] WANG P. The role of West Pacific marginal seas in glacial aridification of China: a preliminary study [J]. Quat Sci, 1995, (1):33-42. [29] LIU T, DING Z. Progresses of loess research in China (Part 2): Paleoclimatology and global change [J]. Quat Sci, 1990, (1):1-9. [30] SUN X, LI X, LUO Y. Environment change from pollen record in deep sea core from northern South China Sea [J]. Quat Sci, 1999, (1):18-26. [31] LIU T, SI Y, WANG R et al. Table of Chinese Quaternary stratigraphic correlation remarked with climate change [J]. Quat, Sci, 2000, 20(2):108-128. [32] PEI WC. Professor Heri Breuil: Poineer of Chinese Paleolithic archaeology and its progreess after him. Miscelanea in homenaje al Abate Heri Breuil [M]. Vol. 2(E Ripoll ed.), Barcelona, 1965, 251-271.

Supplement to Vol. 19, 2000

ACTA ANTHROPOLOGICA SINICA

115-118

Trends Peculiar to the Chinese Palaeolithic Marcel OTTE (Prehistory Department, University of Liège, 7 place du XX Août A1, B-4000 Liège, Belgium.)

Abstract The Palaeolithic of China is very special in nature through its autonomy and its enormous antiquity. It appears to have been isolated very early from the African continent, and underwent its own original development. Its study enables one to perceive phenomena of convergence, and trends that are peculiar to the human mind, reflected in morphology and in industries. This tradition culminates in an autonomous origin of agriculture that is rooted in the local Palaeolithic.

Key words:

Palaeolithic; China; Evolution

The successive forms of human skulls discovered in China from different phases of the Palaeolithic display some constant and coherent trends (Fig. 1). They can be classed in accordance with the development of cranial capacity, the reduction of bony protuberances, the retreat of the face, and the general gracilisation of structures. All these elements are responses to global mechanical modifications linked both to the manipulation of objects and to the development of the cerebral functions. This vast, profound and complex phenomenon seems to have been perpetually active, following the same orientation, from the very start and without any major interruptions. It seems to function in parallel with other regions of the world in the form of a convergence. Seen from this viewpoint, the Far East is both isolated and very extensive. The evolutionary processes seem to be deployed continually through accentuating the original trend more emphatically than in other places where interregional exchange interferes with our global understanding. In China, we thus have an example of autonomous human evolution, free of the constraints that result from successive acculturations, and contrasting with the Near East which was used as a corridor, and with South-East Asia which functioned as a cul-de-sac, analogous to Europe. Hence, the forces that came into action from the very start of humankind had a tendency to modify morphology towards a "modern" aspect. This trend is especially clear and harmonious in China, but is not exclusive to this region: in all places it traverses human evolution, sometimes with fits and starts which were interpreted as traces of migrations. These modifications in bone are in fact the indirect material reflection of development and behaviour. This development gets underway through the successive acquisition of different techniques and of language. The bony forms thus express far more fundamental modifications of a spiritual nature. They themselves seem to evolve in parallel on different continents. When one turns to the production of tools, a similar trend is revealed, both continuous and coherent (Fig. 2 and 3). The reduction of a stone block, the extraction of controlled flakes and the ever-increasing calibration of blades and then bladelets suggest the same domination of mind over matter, in the same way as in Africa or Europe, but here without the jolts caused by invasions. China thus presents a model of the relationship between physical and technical evolution, towards the constitution of present-day humankind. This autonomy is displayed until the Neolithic, when agriculture appears in isolation and very early. The study of the Chinese Palaeolithic enables one to analyse some global phenomena, peculiar to the human mind, in terms of convergence and tendencies, that are reflected both in techniques and in bony remains [1]. Biography: Born in Liège, Belgium, in 1948, Marcel Otte obtained two doctorates in the History of Art and Archaeology at the University of Liège concerning the Upper Palaeolithic of Europe. He has been the Professor of Prehistory at the Univeristy of Liège since 1990, an instructor since 1982.

ACTA ANTHROPOLOGICA SINICA

116

Supplement to Vol. 19, 2000

Liujiang

135129,000

Maba

209,000

Dali

418,000

Locality 1, Zhoukoudian

>780,000

Gongwangling, Lantian

Figure 1

Successive forms of human skulls discovered in China (after Wu & Poirier Olsen [3], Tattersall [4])

[2]

, Wu and

117

OTTE: Trends Peculiar to the Chinese Palaeolithic

3

4

2

1

Figure 2

1. Gongwangling (Lantian), > 780,000 BP; 2. Zhoukoudian, locality 15, early Upper Pleistocene; 3. Shuidonggou, early Upper Palaeolithic; 4. Xiachuan, 19,600-21,700 BP. (after Wu and Olsen [3])

118

Figure 3

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

Number of technical phases through time necessary to produce the lithics from Figure 2

References: [1] LEROI-GOURHAN A. Le Geste et la Parole [M]. Paris: Albin Michel, 1964. [2] WU XZ, POIRIER FE. Human Evolution in China. A Metric Description of the Fossils and A Review of the Sites [M]. Oxford: Oxford University Press, 1995. [3] WU RK, OLSEN JW. Palaeoanthropology and Palaeolithic Archaeology in the People's Republic of China [M]. New York: Academic Press, 1985. [4] TATTERSALL I. The Fossil Trail. How We Know What We Think We Know About Human Evolution [M]. Oxford: Oxford University Press, 1995.

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

119-125

A Use-Wear Study of Lithic Artifacts from Xiaochangliang and Hominid Activities in Nihewan Basin SHEN Chen 1 , CHEN Chun 2 (1. Royal Ontario Museum, Toronto, Canada, M5S 2C6; 2. Department of Cultural Relics and Museology, Fudan University, Shanghai, China, 200433)

Abstract Xiaochangliang is a well-known Paleolithic site dated to the early Pleistocene in the Nihewan basin, North China. The 1998 excavation yielded 901 lithic artifacts along with many mammalian fauna. A total 126 specimens were selected for use-wear analysis. The samples were examined under a Nikon SMZ800 stereoscopic microscope with magnification from 10x to 180x. A total 17 pieces were identified with 18 employed units and 10 pieces showing microfracture use-wear. The activities inferred from wear types suggest that most specimens were related to meat or hide procession. Due to hydrodynamic process, "activity clusters" of early hominids at the site can not be identified.

Key words:

Use-wear analysis; Low-power technique; Xiaochangliang; Paleolithic

1

Introduction

In this study we presents the results of a use-wear analysis of lithic artifacts recovered from 1998 excavation at the Xiaochangliang site. We will first briefly introduce methods of the fieldwork and the context of the lithic assemblages. Application of the low-power use-wear technique in this study will be briefly discussed, followed by the presentation of analytic results from the microscopic examination. Based on the new evidence, hominids activities in the Nihewan basin during the early Pleistocene will be explored.

2

The Xiaochangliang site

Xiaochangliang is one of well-known lower Palaeolithic sites of the early Pleistocene, located in the Nihewan basin, Hebei Province, North China. The site, first identified in1978, has been regarded as one of the earliest sites in China [1-2]. It has been dated somewhere between 1.67 to 1 million years ago on the grounds of a series of palaeomagnetical dating [3-6]. Over the past 20 years, many excavations have been conducted in the Nihewan Basin, including a Sino-American geological and archaeological expedition in the early 1990s. Among these investigations, a large number of lithic artifacts and faunal remains have been recovered from the Xiaochangliang site. Unfortunately, precise proveniences of these materials were not duly recorded, making difficult any further investigations of hominids activities in the region. In 1998 we re-visited the site and conducted a systematic excavation. A preliminary excavation report has been published [7-8].

3

Excavation methods

One of our goals in the 1998 excavation is to explore lithic technology. We are particularly interested in finding more evidence, which could shed new light on hominid behavior during the early Pleistocene in North China. We concentrated our excavation at original Location A where promising density of artifacts was assured as known from previous excavations. The fieldwork exposed 16 contiguous square meters of cultural deposit from the early Pleistocene Nihewan Formation about two and half meters deep below the present surface [7]. Excavation was carried out vertically in each arbitrary level of 10 cm. A total of 8 levels were revealed before the cultural deposit ends on the top of a sterile sediment of reddish clay. More than four thousands lithic artifacts and bone remains were recovered from the excavation, most of which

120

ACTA ANTHROPOLOGICA SINICA

Supplement to Vol. 19, 2000

are from level 3 to level 5 [7]. All pieces larger than 10 mm in size were remained in situ before being recorded three-dimensionally. Small bones or bone fragments and flake debris were also recovered through a ¼ inch screen. The spatial distribution of artifact was later reconstructed through computer database designed for the project.

4

Lithic artifacts

A total of 901 lithic artifacts were unearthed from the 1998 field season; small flakes and irregular chunks are predominant. Lithic artifacts from Xiaochangliang were classified into five categories: nodule, core, formal type, debitage, and debris. In this study, we define formal type that replaces a conventional “tool” category, in order to distinguish typological “tools” from functional “tools.” Formal types are artifacts that show clear evidence of intentional retouches so that the pieces are usually classified on the basis of typological terms, such as scrapers, burins, and points, etc. On the contrary, debitage are flakes without modification at all, but somehow remains useable in function. In this functional study, we realize that both modified flakes (formal types) and unmodified flakes (debitage) have equal potentials to have been used as tools at the Xiaochangliang site. However, debris refers to completely waste flakes given their small size (usually