Research in Action South African Journal of Science 104, September/October 2008 337

sublingual fossa is shallow and low down. First hominoid from the Late This mandible fragment is compatible in size and proportions to medium-sized of Niger hominoids with slender mandibular rami. It is too gracile to belong to an , even a small individual a,b* b c Martin Pickford , Brigitte Senut , Jorge Morales and such as AL 288-1 ()13 from Hadar, José Bragad Ethiopia. Comparisons with several hominoids, both extinct and extant, reveal overall similarities to both troglodytes ince the discovery of Orrorin tugenensis cene taxa are poorly known, and a (Fig. 1) and rather than to any other in the Late Miocene Lukeino Formation, considerable amount of debate has taken genera. The jaw is slightly less deep and STugenHills, , it is generally admit- place, even concerning some relatively more gracile than that of Orrorin28 and the ted that the origin of bipedal hominids occurred well-preserved specimens.3 Each addi- sublingual fossa is not as deeply excavated. earlier than 6 Myr ago and that the adaptation to bipedal stance and locomotion initially tion to the African Late Miocene We consider that the best match for the occurred in a forested or well-wooded setting. hominoid fossil sample is therefore pre- Niger hominoid fossil is with Pan, but In , eight localities aged between 13 cious. because the fossil is fragmentary, we hesi- and 5.5 Myr have yielded hominoid fossils In 1964, a small collection of vertebrate tate to attribute it to this or to any other belonging to nine species. We here report the fossils from Niger, which includes a frag- hominoid genus. occurrence of a Late Miocene hominoid in ment of hominoid mandible, was sent to Anthracotheriidae: Libycosaurus is repre- Niger, associated with a restricted fauna the Muséum National d’Histoire Naturelle sented at the Niger site by a distal which indicates an age of c. 11–8 Myr. The in Paris by the Bataafse Internationale Niger fossil locality is 940 km north of the metapodial attributed to a medium-sized nearest known extant hominoids, and 1000 km Petroleum Maatshappij (now Shell), species, larger than L. petrocchii and west of the nearest recorded fossil hominoid where it is curated under register No. smaller than the large species from Sahabi from Chad. The scientific value of the Niger 1964-27.885. The precise point of discovery and Chad.16,23 The section of the bone at specimen resides in its discovery locus far is not known, but it is probably close to the break reveals a marked degree of from any other known fossil hominoids, its 5°43’E, 15°32’N, where some Mesozoic dorso-plantar compression which distin- Late Miocene age and the attention that it will Chelonians2 with the register No. 840 guishes this fossil from hippopotamid focus on the Neogene fossil record of West were collected by a Mr Nieuwenhuys, metapodials24 with which it might other- Africa, currently almost unknown. who found the . wise be confused. Introduction Bovidae: The family Bovidae is repre- Palaeontology sented at point N 885 by a left frontal It is now widely accepted that the family The assemblage of fossils from N 885, with horn core, a distal left radius and a originated in Africa during Niger, comprises well-mineralized skeletal damaged left calcaneum. The horn core the Late Miocene. However, the African remains of lacustrine and terrestrial ani- belongs to a reduncine antelope, morpho- fossil record for this time period is not mals stained dark brown to black. The logically close to but smaller than Redunca well endowed with hominoid fossils, fauna is restricted, but contains a Nile darti.8,10 making it difficult to trace the transition perch (Lates niloticus), a crocodile (Croco- from pre-hominid to hominid status. dylus cf. niloticus), a chimp-sized homi- Age of the Niger fossils Since 1997, nine hominoid taxa have been noid, a medium-sized species of anthra- Two specimens from the Niger locality reported from eight localities in Africa, cothere (Libycosaurus sp.), and a bovid. are useful for biochronology. The most ranging in age between 13 and 5 Myr: one Pisces: Lates niloticus (Nile perch) is rep- telling evidence is the presence of an species in Namibia [ (12–13 14,25,28 resented at the Niger site by the posterior anthracothere, Libycosaurus sp., equiva- Myr)]; five taxa in Kenya [a chimp- part of the skull. lent to the medium-sized species from like hominoid (12.5 Myr), Crocodylia: Crocodylus cf. niloticus (Nile Toros-Ménalla, Chad and Sahabi, (9.5–10 Myr), (9.5 Myr), a crocodile) is represented at locality N 885 Libya.16,23 Secondly, the bovid horn core -like hominoid (6 Myr), and Orrorin 29 by a brevirostrine mandible with buccal has some resemblances to, but is smaller (6 Myr)]; two in Ethiopia [Chorora- depressions between the teeth, two der- than, material from Sahabi15 (Libya), pithecus (10–10.5 Myr) and 30 30 11 3 mal scutes, two articulated vertebrae, the Lukeino, Mpesida and Lothagam (5.7 Myr)]; and one in Chad [Sahelan- 9 23 occipital part of a skull and diverse frag- (Kenya) and Langebaanweg (South thropus (6–10 Myr) ]. Despite the draw- ments of skull. Africa). An age of between 11 and 8 Myr is backs of its Late Miocene fossil record, possible for the deposits that yielded Africa has now yielded a respectable Hominoidea is represented at site N 885 these fossils. diversity of hominoids comparable with by a right mandible fragment (Fig. 1) that of Europe and Asia, making it more containing the roots of the first . The Palaeoenvironment and likely to be the cradle of hominid evolution mandible is slender and moderately palaeoclimatology than Eurasia.1 All these African Late Mio- deep. The jaw beneath m/1 is 13.2 mm thick, and its depth from the alveolar The presence of Nile perch, Lates niloticus, aCollège de France, 11 place M. Berthelot, 75005, Paris, process to the ventral margin is 31.6 mm. in the Niger deposits indicates the former France. bUMR 5143 du CNRS, Case postale 38, 8 rue Buffon, The preserved part of the tooth measured presence of a freshwater lake or large 75005, Paris, France. at cervix is 9.4 × 9.7 mm, but judging from river in the country. The anthracothere cPaleobiologia, Museo Nacional de Ciencias Naturales, CSIC, José Gutierrez Abascal 2, 28006, Madrid, Spain. the position of the alveoli mesial and Libycosaurus was likely a denizen of dUniversité Paul Sabatier, Toulouse, FRE 2960 CNRS, distal to the preserved roots, the crown shallow swampy parts of a palaeolake, Anthropobiologie et Imagerie Anatomique, 39 Allées Jules Guesde, 31000, Toulouse, France length would originally have been c. 11 mm and the reduncine probably lived close to *Author for correspondence. E-mail: [email protected] and the breadth about 9.9 mm. The water. This evidence, combined with the 338 South African Journal of Science 104, September/October 2008 Research in Action

Fig. 1. Hominoid right mandible fragment from Niger (A) compared with a mandible (B): A1) buccal view; A2) sagittal scan, mesial to the left; A3) frontal scan, buccal to the left (arrow points to the mandibular canal). (Scale bars: 10 mm) presence of a hominoid , indicates the view that the Sahara is a relatively chimp-like form). In the circumstances, that the region was appreciably more young desert, post-dating the Late Mio- it seems superfluous to invoke a Late humid (at least 750 mm mean annual cene. Recent work in Egypt has revealed Miocene reintroduction of hominoids into rainfall) during the Late Miocene than it is that central Egypt was covered in wood- Africa from Eurasia in order to give rise to today (ranging from sahel to total desert land to forest some 11–10 Myr ago26,27 and hominids,1 a view that was attractive only over large areas of the country). Chad is long known to have been more while the Late Miocene hominoid fossil humid in the past than it is today5,6,20 as record of Africa was poorly known. Discussion were Libya7 and Tunisia.21,22 Apart from , hominoids are today At present, the living Conclusions confined to forest and woodland, but the closest to the Niger fossil site occur 940 km A restricted vertebrate fauna of Late common chimpanzee can survive in sub- to the south in Ghana and Nigeria and Miocene age from Niger is of interest on humid environments as long as there are 1700 km to the west in Mali (Fig. 2).4,32 account of its age and the presence within enough trees to supply adequate food and Niger thus joins an increasing list of it of a hominoid primate and an anthra- security.17 This means that the longest dry African countries that have yielded re- cothere. The discovery helps to fill what period should not exceed two months, mains of Late Miocene (c. 11–5.5 Myr ago) was a vast gap in the geographic coverage and that the area should be well endowed hominoids including: Kenya (Lukeino, of fossil hominoids; the nearest known with riparian forest so that there is always 6 Myr, two taxa);19,25,28 (Samburu Hills, fossil specimen of comparable age being a supply of fresh vegetation in the form of 9.5 Myr);12 (Nakali, c. 10 Myr);14 Ethiopia from Chad, over 1000 km to the east and fruit, leaves, pith and bark, even during (Ch’orora, c. 10.5 Myr, Middle Awash, 3500 km from sites in Ethiopia and Kenya the driest months of the year. In general, c. 5.7 Myr)29 and Chad (Toros-Menalla, that have yielded Late Miocene hominoids. however, chimpanzees and flour- c. 10–6 Myr)3,23 and it indicates that further The site is closer to the Spanish Late Mio- ish best in tropical rainforest of Guineo- research in West Africa will undoubtedly cene hominoid sites (3000 km)18 than it is Congolian affinities.31 At present, the add fuel to the debate about hominid to the East African ones. The find site, the most arid environment inhabited by origins. It is clear that during the Late position of which is not precisely known chimpanzees is wooded savanna in Miocene, hominoids were widespread in at present, is about 940 km north of the Liberia and Miombo woodland in Tanza- Africa and, although the quantity of closest extant hominoids (Pan troglodytes) nia.17,32 The presence of a chimp-sized fossils is limited, the diversity is large, with in Ghana and Nigeria. The fauna as a hominoid in the Late Miocene of Niger at least seven lineages already recognized whole indicates the former presence of a therefore provides evidence that the in the time range 11–5.5 Myr ago (in the freshwater lake or large river in Niger,and country was considerably more humid order in which they were reported: a Late Miocene palaeoclimate consider- then than it is today, and was at least Samburupithecus, Orrorin, Ardipithecus, ably more humid than that of today. The covered in wooded savanna if not denser Sahelanthropus, an unnamed protogorilla- scientific value of the Niger hominoid vegetation categories. like form, , and Nakali- specimen resides in the attention that it The discovery of a Late Miocene aquatic pithecus) and two between 13 and 11 Myr will focus on a palaeontologically poorly and terrestrial fauna in Niger confirms ago (Otavipithecus and an unnamed known part of the continent, which Research in Action South African Journal of Science 104, September/October 2008 339

Fig. 2. Extant distribution of species of Pan, and the position of two Late Miocene hominoid localities (stars) (Pan distribution based on data in Butynski4). nevertheless probably has a great deal to from Langebaanweg, South Africa. Ann. S. Afr. the Miocene. In Wood to Survive, Liber Amicorum Mus. 79, 213–337. Roger Dechamps, eds F. Maes and H. Beeckman. contribute to our knowledge about the 10. Gentry A.W.and Gentry A. (1978). Fossil Bovidae Ann. Sci. Econ. Mus. Roy. Afr. Cent. Tervuren 25, origin of hominids. (Mammalia) of , Tanzania. Bull. Br. 195–212. Mus. (Nat. Hist.) Geol. 29, 289–446 and 30, 1–83. 22. Pickford M. (2000). Crocodiles from the Beglia We thank Christine Argot and Claire Sagne for their 11. Harris J. (2003) Bovidae from the Lothagam Suc- Formation, Middle/Late Miocene Boundary, assistance with the collections and registers of fossils cession. In Lothagam: The Dawn of Humanity in Tunisia, and their significance for Saharan at the MNHN. Thanks also to Philippe Richir for mak- Eastern Africa, eds M. Leakey and J. Harris, pp. palaeoclimatology. Ann. Paléont. 86, 59–67. 531-579. Columbia University Press, New York. 23. Pickford M. (2008). Libycosaurus petrocchii ing casts and Jacques Treil (MEDES, Toulouse) for CT 12. Ishida H. and Pickford M. (1997). A new late Bonarelli, 1947, and Libycosaurus anisae (Black, scans. Miocene hominoid from Kenya: Samburupithecus 1972) (Anthracotheriidae, Mammalia): nomencla- kiptalami gen. et sp. nov. C.R. Acad. Sci. Paris 325, tural and geochronological implications. Ann. 1. Begun D. (2001). European hominoids. In The 823–829. Paléont. 94, 39–55. Primate Fossil Record, ed. W.C. Hartwig, pp. 13. Johanson D.C., White T.D.and Coppens Y. (1978). 24. Pickford M. (2008). The myth of the hippo-like 339–368. Cambridge University Press, Cam- A new species of the genus anthracothere: The eternal problem of homology bridge. (, Hominidae) from the of East and convergence. Revista Española de Paleontologia Africa. Kirtlandia 28, 1–14. 23, 31–90. 2. Bergounioux F. and Crouzel F. (1968). Deux 14. Kunimatsu Y., Nakatsukasa M., Sawada Y., Sakai 25. Pickford M. and Senut B. (2005). Hominoid teeth tortues fossiles d’Afrique. Bull. Soc. Hist. Nat. T., Hyodo M., Hyodo H., Itaya T., Nakaya H., with chimpanzee- and gorilla-like features from Toulouse 104, 179–186. Saegusa H., Mazurier A., Saneyoshi M., Tsujikawa the Miocene of Kenya: implications for the 3. Brunet M., Guy F.Pilbeam D., Taisso Mackaye H., H., Yamamoto A. and Mbua E. (2007). A new Late chronology of the - divergence and Likius A., Ahounta D., Beauvilain A., Blondel C., Miocene great ape from Kenya and its implica- biogeography of Miocene hominoids. Anthropol. Bocherens H., Boisserie J-R., De Bonis L., Coppens tions for the origins of African great and Sci. 113, 95–102. Y., Dejax J., Denys C., Duringer P., Eisenmann V., humans. Proc. Natl Acad. Sci. USA 104, 19220– 26. Pickford M., Wanas H., Mein P. and Soliman H. Fanone G., Fronty P., Geraads D., Lehmann T. 19225. (2006). Indications for a humid climate in the Lihoreau F., Louchart A., Mahamat A., Merceron 15. Lehmann U. and Thomas H. (1987). Fossil Western Desert of Egypt 11–10 million years ago: G., Mouchelin G., Otero O., Pelaez Campomanes Bovidae (Mammalia) from the Mio-Pliocene of evidence from Galagidae (Primates, Mammalia). P., Ponce de Leon M., Rage J-C., Sapanet M., Sahabi, Libya. In Neogene and Geology C. R. Palévol 5, 935–943. Schuster M., Sudre J., Tassy P., Valentin X., of Sahabi, ed. N. Boaz et al., pp. 323–335. Alan Liss, 27. Pickford M., Wanas H., Mein P. and Soliman H. Vignaud P., Viriot L., Zazzo A. and Zollikofer C. New York. (2008). Humid conditions in the Western Desert of (2002). A new hominid from the Upper Miocene 16. Lihoreau F. (2003). Systématique et Paléoécologie des Egypt during the Vallesian (Late Miocene). Proc. of Chad, Central Africa. 418, 145–151. Anthracotheriidae (Artiodactyla: Suiformes) du Mio- Third International Conference on the Geology of the 4. Butynski T.M.(2004). Le chimpanzé commun, Pan Pliocène de l’Ancien Monde: Implications paléo- Tethys, South Valley University, Aswan, January, troglodytes, taxinomie, distribution, effectif et biogéographiques. Thèse, University of Poitiers, 2008, pp. 1–17. statut de la conservation. In Chimpanzés d’Afrique France. 28. Senut B., Pickford M., Gommery D., Mein P., de l’Ouest, pp. 15–22. Union Mondial pour la 17. Moore J. (1992). ‘Savanna’ Chimpanzees. In Topics Cheboi K. and Coppens Y. (2001). First hominid Nature, Gland and Cambridge. in Primatology, 1: Human Origins, eds T.Nishida, W. from the Miocene (Lukeino Formation, Kenya). C. 5. Coppens Y. and Koeniguer J-C. (1967). Sur les McGrew, P.Marler,M. Pickford and F.de Waal, pp. R. Acad. Sci. Paris 332, 137–144. flores ligneuses disparues plio-quaternaires du 99–118. University of Tokyo Press, Tokyo. 29. Suwa G., Kono R.T., Katoh S., Asfaw B. and Tchad et du Niger. C. R. Acad. Sci. Paris 265, 18. Moyà-Solà S. and Köhler M. (1996). A Beyene Y. (2007). A new species of great ape from 1282–1285. skeleton and the origins of great ape locomotion. the late Miocene epoch in Ethiopia. Nature 448, 6. Coppens Y. and Koeniguer J-C. (1976). Significa- Nature 379, 156–159. 921–924. tion climatique des paléoflores ligneuses du 19. Nakatsukasa M., Pickford M., Egi N. and Senut B. 30. Thomas H. (1980). Les Bovidés du Miocène Tertiaire et du Quaternaire du Tchad. Bull. Soc. (2007). Body weight, femoral length and stature of supérieur des couches de Mpesida et de la géol. Fr. 18, 1009–1015. Orrorin tugenensis, a 6 Ma hominid from Kenya. formation de Lukeino (District de Baringo, 7. Dechamps R. (1987). Xylotomy of fossil wood Primates 48, 171–178. Kenya). Actes 8ème Congr. Panafr. Préhist. Etudes from the Sahabi Formation. In Neogene Palaeontol- 20. Otero O., Pinton A., Vignaud P. and Brunet M. Quat., Nairobi, pp. 82–91. ogy and Geology of Sahabi, ed. N. Boaz, pp. 37–41. (2006). The fish fauna from TM 266 (Late Miocene 31. White F. (1986). La végétation de l’Afrique. Mémoir Alan Liss, New York. from Toros-Menalla, Western Djurab, Chad), accompagnant la carte de végétation de l’Afrique, 8. Gentry A.W. (1978). The fossil Bovidae of the environment and biogeography. Annual Sympo- Unesco/AETFAT/UNSO. Orstom–Unesco, Paris. Baringo Area, Kenya. In Geological Background to sium of Vertebrate Palaeontology and Comparative 32. Wolfheim J.H. (1983). Primates of the World. Fossil Man, ed. W.W.Bishop, pp. 294–308. Scottish Anatomy, Paris, Abstracts, p. 10. Distribution, Abundance and Conservation. New Academic Press, Edinburgh. 21. Pickford M. (1999). Aubréville’s hypothesis of a York Zoological Society, University of Washing- 9. Gentry A.W. (1980). Fossil Bovidae (Mammalia) southwards shift of Africa’s vegetation belts since ton Press, Seattle and London.