Implications of New Fossil Material Attributed to Plio-Pleistocene Asian Hominidae

Dennis A. Etler*

Introduction

The recent description of fossil material attributed to an early form of Homo from Longgupo cave in Sichuan, China (Huang et al. 1995) raises new questions about the role Asia played in the evolution of the modern Hominoidea. On the one hand it refocuses attention on the fossil record of Asian Mio-Pliocene hominoids such as Sivapithecus, Ramapithecus and Lufengpithecus and the role they played, if any, in human phylogenesis. On the other hand it reopens questions about early hominid dispersal patterns and the evolutionary emergence of particular hominid taxa.

Longgupo, Wushan, Sichuan

The Longgupo remains, which consist of a fragmentary mandible and an isolated incisor, are dated between 1.8-2.0 MYA and regarded by their Chinese discoverers and American colleagues as too primitive to be considered H. erectus (Huang et al. 1995). Wood and Turner (1995) concur, suggesting they most likely represent an early form of Homo, such as H. habilis, or perhaps H. ergaster. If accepted as a full- fledged hominid, the small size and primitive morphology of the preserved mandibular dentition (P4 and M1) would probably warrant erecting an entirely new species for the specimen.

Based on the Longgupo material some have argued that a dispersal of early Homo into Asia prior to the advent of more advanced forms in Africa supports the claim that Asian H. erectus, was a side- branch on the human evolutionary tree, uninvolved in the African origins of later humans (Culotto 1995). A number of Chinese researchers, however, have an entirely different perspective. They have long argued that dentally progressive hominoids, such as Lufengpithecus known from the Miocene and early Pliocene of Yunnan province in southwestern China, are direct human ancestors and that this area of southern Asia, not eastern or southern Africa, was the true cradle of humankind. Early Pliocene hominoids from the Yuanmou basin of Yunnan, for instance, have been described as having small distolingually elongated P4's, and low crowned molars, lacking in cingula, with broad talonids and squarish outlines (Lu 1991), all features similar to those seen in the Longgupo specimen. Chinese advocates of an Asian origin for the Hominidae (humans and their direct ancestors) emphasize these hominid-like dental features of Late Tertiary southern Chinese apes.

Some Chinese paleontologists dismiss the African australopithecines as specialized off-shoots of the human family tree not directly involved in the origins of the genus Homo, much as some in the West dismiss H. erectus as not being directly involved in the later descent of modern humans. Remains such as those recovered at Longgupo, which bear a striking resemblance to the Yunnanese ape Lufengpithecus mentioned above, are seen as the "missing link" between these southern Chinese apes and later Chinese hominids. If the Longgupo specimens are accepted as representing a true hominid then it follows that the hypothesis of a southern Chinese origin for the Hominidae merits serious consideration. Many in the West, will categorically reject such a notion as absolutely unwarranted and contrary to all known facts. Chinese paleoanthropologists, however, find the proposition that Asian H. erectus was an evolutionary dead-end unrelated to modern Asians just as untenable as many Western scientists find claims for human descent from Chinese apes.

It can also be argued, however, that the Longgupo specimens are not true hominids after all, that Lufengpithecus and its descendents were not human ancestors and that Longgupo represents a heretofore unknown continuation of the Lufengpithecus ape lineage into the Plio-Pleistocene of southern China. This is the position taken here.

The Significance of Lufengpithecus in Human Evolutionary Studies in China

What sort of ape was Lufengpithecus? When first discovered in the late 1970s and early 1980s two morphs were recognized, a relatively gracile form initially attributed to Ramapithecus lufengensis, and a larger more robust form attributed to Sivapithecus yunnanensis. At the time the discoveries were made Ramapithecus was still regarded by many paleanthropologists as the earliest and most primitive hominid, based primarily on dental features thought to be derived in the human direction. Sivapithecus was considered to be an early pongid only indirectly related to the smaller hominid. It wasn't long, however, before these two fossil taxa were re-evaluated based on new fossil discoveries in the Siwaliks of Pakistan and molecular evidence which indicated that Ramapithecus occurred too early in the fossil record to be a direct human ancestor. The similarities between the two forms were quickly recognized and most researchers eventually came around to viewing Ramapithecus merely as the female of Sivapithecus. Further research showed that the Chinese specimens were distinct from those recovered in the Siwaliks and in 1987 the genus and species Lufengpithecus lufengensis was erected to accommodate them. Prior to this reassessment, however, Chinese researchers were convinced that Ramapithecus lufengensis was a direct human ancestor and some still are today. In this regard Wu and Oxnard published a series of papers in the early 1980s that focused on the hominid- like features of the smaller Lufeng morph.

With the demise of Ramapithecus, the Lufeng material was quickly reevaluated. More recently Lufengpithecus has been characterized as a highly sexually dimorphic large-bodied hominoid with uncertain affinities to other living and extinct great apes (Wu and Wang 1987, Kelley and Etler 1989, Kelley and Xu 1991, Wood and Xu 1991). It must still be recognized, however, that the smaller Lufeng morph, now regarded as the female of Lufengpithecus lufengensis, does have certain dental traits that closely approximate those of early hominids (Wu and Oxford 1983a, 1983b). These include relatively low-crowned canines; incipiently bicuspid lower P3s with tall protoconids and low-lying metaconids; a foreshortened, lozenge-shaped lower P4 with protoconid and metaconid of approximately equal size; and low-crowned, thickly enameled molars. There are other features of the female Lufengpithecus dentition, however, that can be characterized as typical for Late Miocene apes with thickly enameled teeth. The long axis of lower P3 is oriented from the mesiobuccal to the distolingual; lower P4 is double rooted with a slender mesially positioned slit-like anterior fovea, an elevated trigonid and a depressed, elongate talonid that is skewed slightly distolingually; and lower M1 has cusps arrayed round the periphery of a broad talonid basin, with salient buccal protoconid and hypoconid that wear quickly, producing enamel windows, an endoconid that projects slightly distolingually and a small buccally displaced hypoconulid separated from the large endoconid by a short longitudinal groove.

In comparison to the above, the Longgupo mandible, which contains a well-worn P4 and M1, has virtually identical morphological features. The lower P4 is primitively double-rooted and shaped and configured like that from Lufeng with an elevated trigonid basin and a depressed and elongated talonid. Moreover, the mesial face of the P4 has an interproximal facet for the P3 that is displaced markedly lingually. This shows that the P3 crown was sectorial, with its long axis oriented from the mesiobuccal to the distolingual, the normal orientation for the P3 in apes with teeth of this size and the condition noted above in female Lufengpithecus. The lower M1 at Longgupo is a non-descript hominoid molar that does not differ from the description of the female Lufengpithecus M1 given above in any meaningful way. In all of the above features which relate to the preserved dentition at Longgupo the two forms are virtually identical.

Dimensionally the Longgupo dentition is also nearly identical to female Lufengpithecus. The dimensions of the Lufeng female lower P4 are: average MD length 7.89 mm, sd=0.65, n=35; average BL breadth 9.08 mm, sd=0.52, n=34. The dimensions of the Longgupo specimen are: MD length 7.4 mm, BL breadth 9.1 mm; well within the 95% confidence limits of the Lufengpithecus female range of variation. The dimensions of lower M1 of the female Lufengpithecus are: average MD length 10.57 mm, sd= 0.48, n=35; average BL breadth 9.08 mm, sd=0.37, n=35. The dimensions of the Longgupo lower M1 are MD length 11.1 mm, BL breadth 10.1 mm, again well within the 95% confidence limits of the size range of female Lufengpithecus. Moreover, female Lufengpithecus M1s are relatively longer than broad with a L/B ratio of 1.10, the L/B ratio of the Longgupo lower M1 is identical. Early hominid M1s vary in this regard but tend to be shorter and broader, either close to square in occlussal outline or broader than wide. Based on the above comparisons it can be seen that dimensionally as well as morphologically the dentitions of female Lufengpithecus and the Longgupo mandibular specimen are nearly indistinguishable. In comparing the Longgupo P4 to those known from Lufeng Huang and Fang (1991) made only one observation, that the buccal face of the Longgupo specimen is more rounded than that seen in Lufengpithecus, which is more convexly triangular. The Lufeng site, however, is 5 My older than Longgupo, and it is rather remarkable that there has been so little change in the preserved dental morphology. Similar hominoid material from Yuanmou, also in Yunnan, dated to the Early-Middle Pliocene, approximately intermediate in age between Lufeng and Longgupo, has moreover been described as somewhat more progressive than Lufeng (Lu 1991), meaning apparently that it has an even greater resemblance to the dentition of early hominids. Hence the rounding of the buccal face of lower P4 in the Longgupo specimen relative to that seen in female Lufengpithecus can be attributed to progressive evolutionary change during the 5 My that separate them.

Given the above comparisons there is no discernible reason why Longgupo cannot be seen as a descendent of Lufengpithecus that survived into the Late Pliocene and early Pleistocene of southern China. The most obvious affinities of the specimen lies therefore with Lufengpithecus and not early Homo. The Longgupo jaw is not that of the earliest Chinese hominid, it is, rather, that of a late occurring Chinese fossil ape.

Other Purported Plio-Pleistocene Hominids from China

Another claim for a Plio-Pleistocene hominid in China rests with two incisors known since the 1960s from another locality in Yuanmou, Yunnan (Hu 1973). Dated to 1.7 MYA by paleomagnetism (Qian 1985), the Yuanmou incisors are very large and distinctly H. erectus-like in their enamel patterning and over-all construction, quite at variance to the small gracile incisor recovered at Longgupo. Persistent questions about the original provenance of the Yuanmou specimens and their relative scantiness mitigate against over reliance on them as proof positive of early Pleistocene hominids in south China. Nevertheless, the Yuanmou specimens can be argued to represent a hominid close to the base of the H. erectus lineage in China.

There are other specimens from China, collected from both traditional Chinese apothecaries (i.e. "Hemanthrpopus") and early Pleistocene cave deposits, that have in the past been favorably compared to African australopithecines due to their large size and derived cuspal morphology (von Koenigswald 1957, Gao Jian [Xu et al.] 1975, Franzen 1985). The dating of these remains is problematical but most likely falls within the early Pleistocene. Assessment of the true affinities of these specimens has been controversial, with some investigators suggesting that they can be accommodated within the known variation of H. erectus (Zhang 1984), while others have suggested possible links to the still poorly understood fossil orang-utan of the southern Chinese and insular southeast Asian Pleistocene (Gu 1989) or the enigmatic Southeast Asian Meganthropus.

One sample of four lower molars from Jianshi, Hubei (Gao Jian [pseudonym for Xu et al.] 1975) has generated considerable interest (Pope 1977, Franzen 1985) because of their large size, archaic appearance, inferred antiquity and association with Gigantopithecus. The original investigators (Gao Jian 1975) suggested australopithecine affinities for the assemblage. More recent comment (Zhang 1984), however, discounts this likelihood and suggests that the material can be accommodated within the hypodigm of Asian H. erectus. Of the four molars one, PA 507, a right lower M1, was collected from an apothecary shop in Badong county, Hubei and lacks proveniance. The other three (PA 502-504) were found in situ at Longgudong (E 110o 04', N 30o 38'), a horizontal karst cave, in association with Gigantopithecus teeth and a typical southern Chinese Stegodon- Ailuropoda faunal assemblage. A few relatively archaic elements are represented in the fauna (i.e. Tetralophodon serridentoides, Gomphotheriidae, Machairodontidae) and a number of the taxa are intermediate in size between samples from sites attributed to the early Pleistocene (Gigantopithecus Cave at Liucheng) and middle Pleistocene (Yanjinggou) respectively. Based on the analysis by Zhang (1982) of the evolutionary increase in tooth size within Chinese Gigantopithecus, it can be inferred that the relatively large size of the Gigantopithecus teeth at Longgudong indicates at least an early middle Pleistocene, if not a late early Pleistocene age for the material. Neither age is contravened by other elements within the faunal assemblage.

Of the four molars, PA 502 and 503 from Longgudong are most likely antimeres given their metric and morphological similarities. They were originally designated as left and right lower M2 by Xu et al. but if so they are rather anomalous. A more likely designation would be lower M3. PA 504, also from Longgudong, is considered to be a right lower M2. PA 507, collected at Badong, is thought to be a right lower M1. All four molars are distinguished by their large size. They have lengths and breadths that are near or well beyond the upper limits for these dimensions in the Zhoukoudian Locality 1 sample. Another peculiarity of these molars, particularly PA 504 and PA 507, is their low length-breadth index (BL width/MD length x 100) indicating the long and narrow contour of the teeth. The values of this index for PA 504 (88.9) and PA 507 (84.7) are substantially less then the lowest recorded values at Zhoukoudian (89.4 for lower M1 sp. #34, 89.7 for lower M2 sp. # 43). While the absolute dimensions of the Jianshi molars fall beyond the upper range of most samples of early Homo and their length-breadth indices are also uncommonly low for hominids in general, all the above dimensions and index values fall well within the range of variation of fossil orangs as described by Hooijer (1952).

Morphological features of the Jianshi molars that have been used to suggest affinities to H. erectus or gracile australopithecines such as presence of a protoconidal cingulum, development of lingual and distal accessory cusps, disposition and relative proportions of the major cusps, etc. are features also seen in fossil orang- utans. In addition Hooijer (1952) makes note of the fact that in nearly all cases the greatest breadth of lower M1 in his sample of fossil orangs falls across the talonid, while in lower M2 it generally falls across the trigonid. This same pattern can be seen in the Jianshi molars. Hooijer also comments that in his sample of fossil orang molars there is a great amount of variation in the type and extent of crenulation of the occlusal surface with some specimens being of the "coarsely wrinkled type". The crenulation of the Jianshi molars is relatively weak but the development of accessory ridges and grooves observable in the specimens could well fall within Hooijer's latter category. The possibility that some or all of the Jianshi teeth may actually represent a Pleistocene form of Pongo must therefore be given due consideration (see also Gu et al. 1988). In this respect Xu et al. make note of the fact that in the IVPP collection of fossil orang teeth from cave deposits in southern China (Pei 1935, Kalke 1972, Gu et al. 1987, Zhou 1988) there is at least one specimen rather like the Jianshi molars (Gao Jian 1975).

Von Koenigswald also procured a large collection of southern Chinese Pleistocene orang teeth from Chinese drugstores (von Koenigswald 1957). Of the more than 1500 specimens so obtained a number were noticed to differ from the others "in having less wrinkles and a double cusp on the middle of the upper molar" (von Koenigswald 1981). These were given the designation Hemanthropus peii (von Koenigswald 1957) and were regarded by him to show affinities to both Meganthropusas known in Java and Australopithecus as known in south Africa. Their large size clearly distinguished them from H. erectus. Von Koenigswald (1981) notes, moreover, that his fossil orang teeth are "much bigger than those of the modern form, some of them even surpassing the large teeth of the gorilla" (see also Pei 1935). Given Hooijer's observations about the large degree of morphological variation in the collection of fossil orang teeth he studied and the large size of the southern Chinese orangs noted by both von Koenigswald and Pei it seems reasonable to suggest that Hemanthropus can also be accommodated within Pongo.

Also to be considered in the above calculus is the possibility, mentioned above, that large-bodied hominoids other than Pongo and Gigantopithecus may be found in early Quaternary deposits of southern China. As previously discussed, hominoids from the late Miocene site of Lufeng and from early-middle Pliocene sites in the Yuanmou basin, both located in the southern Chinese province of Yunnan, represent a distinct lineage of southern Chinese ape named Lufengpithecus. The affinities of Lufengpithecus are still unclear but most investigators while noting similarities to Pongo have also noted many differences. Lufengpithecus is characterized by extreme sexual dimorphism and thick enameled molars that preserve a late Miocene ape morphology. Some of the large-sized fossil hominoid molars from southern China that lack derived features of modern orangs could possibly be attributable to large males of this genus that persisted into the early Pleistocene.1 Given Hooijer's observations about morphological variability in Pleistocene orang teeth from Java, von Koenigswald's remarks on the large size of orang teeth from southern China and the possibility of the persistence of forms such as Lufengpithecus into the early Pleistocene of southern China, it seems reasonable that the specimens under consideration can best be accommodated under the rubric of Pleistocene "large-bodied hominoid" rather than some form of Asian australopithecine or a widely size varying form of Asian H. erectus.

On the other hand, if all the above specimens are recognized as valid hominids we are left with the possibility of considerable taxonomic diversity amongst hominids from the early Pleistocene of East Asia, with perhaps three distinct lineages represented, i.e. a small-bodied form of early Homo (Longgupo), a large-bodied form which shows phenetic similarities to Australopithecus (Jianshi, "Hemanthropus") and another form that is consistent with attribution to H. erectus (Yuanmou).

Implications of the Re-dating Of Indonesian Hominids

Turning to the situation in Indonesia, the re-dating of strata that have produced hominid remains at Sangiran and Modjokerto to 1.6-1.8 MYA (Swisher et al. 1994), opens up an entirely new Pandora's box of diversity. Some of the specimens dated to the earliest Pleistocene have long been recognized as well-evolved forms of H. erectus. The Indonesian paleoanthropologist Prof. T. Jacob (1981) has in the past allied one of them, Sangiran 17 (Pithecanthropus VIII), with H. erectus soloensis, long thought to be a late occurring archaic human that persisted into the upper Pleistocene (approximately 100-200 KYA) of Java, while some have seen similarities between the Modjokerto juvenile cranium (H. modjokertensis) and H. habilis (Tobias and von Koeningswald 1964, Cronin et al. 1981). Other sites, from what could be equally old stratigraphic horizons, have yielded remains of hyper-robust individuals attributed by some to an australopithecine-like taxon, Meganthropus (Robinson 1953, Franzen 1985, Tyler 1995). In Java, as in China, we have specimens from the same time and place possibly representing three very different grades or clades of hominid.

If All the Above Claims Are Accepted, What Are the Implications for Our Understanding of Human Evolution?

If, for the sake of argument, we accept all of the above evidence as representing valid Plio-Pleistocene Asian hominids, and there seems to be a much greater willingness to do so now than in the recent past, what do we get? Apparently at least three and possibly four distinct hominid lineages in Asia at or near the Plio-Pleistocene boundary: a small, gracile form of early Homo in China; a form, H. modjokertensis, in Java, thought by many until the early 1980s to be conspecific with H. habilis (Cronin et al 1981); a larger, fully evolved form of H. erectus in China and Java; and larger still, a form attributed to a kind of robust australopithecine, both in China and island Southeast Asia. Hominid diversity of this magnitude would be highly reminiscent of that encountered in East Africa at approximately the same time, where Australopithecus boise, H. habilis, H. rudolfensis and H. ergaster are said to occur more or less sympatrically and synchronically (Wood 1992).

What, if accepted, does this apparent diversity say about hominid dispersal patterns? It would seem, for one, that it would not have take much in the way of cultural accouterments for hominids to have made their way from Africa to the far reaches of east Asia and Sundaland. It could be argued, of course, that other higher primates have had little difficulty in dispersing throughout Africa, Europe and Asia, where their remains are found in deposits that date from the early Miocene to the middle Pliocene, so why should early hominids have been any different? Given the success of their progeny, it could be argued that they were very hardy creatures, perfectly capable of meeting whatever environmental challenges they may have encountered on their trek to the Far East. Moreover, it could be argued that various early hominid lineages dispersed soon after their first appearance, suggesting that some later lineages may have diverged in Asia rather than Africa. Thus the possibility has been raised that H. erectus actually evolved in Asia from a species of early Homo, perhaps returning to Africa fully formed and ready to smite its more primitive brethren, the australopithecines. Another line of reasoning, well-liked by proponents of the "Out of Africa" hypothesis, is that H. erectus evolved in Asia and stayed there, part of an evolutionary side-show that had no bearing on the origins of later H. sapiens (Culotta 1995). The prospect of multiple hominid species spread far and wide during the early Quaternary also fits in well with the perspective that sees hominids as no less prone to rapid and frequent speciation than any other mammalian group (Tattersall 1986, Foley 1991).

If we accept the above scenario, the early Quaternary landscape of the orient would have been replete with a dazzling array of hominoids. In addition to the purported hominid-types mentioned above, Plio-Pleistocene Asia was also inhabited by a variety of pongids (sensu stricto), including the immense ape Gigantopithecus, the gorilla-sized fossil orang-utan, and perhaps descendants of the chimpanzee-sized Mio-Pliocene southern Chinese ape Lufengpithecus. Add to this the whole panoply of gibbons known throughout the Pleistocene of southern China and southeast Asia and you get quite an interesting and diverse menagerie. Then again, if we accept the argument that hominids were able to disperse so quickly after their first appearance on the evolutionary scene we may have to cast our nets much further afield when looking for potential human ancestors than previously assumed.

A More Realistic Appraisal

The above embroidery is, however, contingent on the correct dating, identification and interpretation of the specimens considered. And herein lies the rub. As discussed above many of the specimens under consideration are very fragmentary and/or very poorly preserved and subject to varying and at times contradictory interpretations. The dating of other specimens is still subject to doubt. It will therefore be difficult to accept the intriguing picture of hominid diversity painted above, or the prospects for early human ancestry in Asia, until far more substantial discoveries are forthcoming.

Literature Cited

  1. Cronin, J.E., Boaz, N.T., Stringer, C.B. and Rak, Y. 1981. Tempo and mode in hominid evolution. Nature 292:113-122.
  2. Culotta, E. 1995. Asian hominids grow older. Science 270:1116-1117.
  3. Etler, D. A., Crummett, T. L. and Wolpoff, M. N. 1995. Earliest Chinese hominid mandible may be an ape. Submitted.
  4. Foley, R. 1991. How many species of hominid should there be? J. Hum. Evol. 20:413-27.
  5. Franzen, J. L. 1985. Asian australopithecines? In, Hominid Evolution: Past, Present and Future. Ed. E. Delson, pp. 253-263. New York: Alan R. Liss.
  6. Gao Jian (Xu, C., Wang, L. and Han, K.). 1975. Australopithecine teeth associated with Gigantopithecus. Vert. PalAsiat. 13:81-87.
  7. Hu, C. Z. 1973. Ape-man teeth from Yuanmou, Yunnan. Act. Geol. Sin. 1:65-71.
  8. Huang, W., Ciochon, R., Gu, Y., Larick, R., Fang, Q., Schwarcz, H., Yonge, C., de Vos, J. and Rink, W., et al. 1995. Early Homo and associated artifacts from Asia. Nature 378:275-78.
  9. Jacob, T. 1981. Solo Man and Peking Man. In, Homo erectus: Papers in Honor of Davidson Black. Eds., B. A. Signon and J. S. Cybulski. pp. 87-103. Toronto: University of Toronto press.
  10. Lu, Q.W. 1991. Book review: Other Origins. The Search for the Great Ape in Human Prehistory. Internat. J. Primatol. 12:537-538.
  11. von Koenigswald, G. H. R. 1957. Remarks onGigantopithecusand other hominoid remains from southern China. Proc. Kon. Nederl. Akad. Wet. (B) 60: 153.
  12. Qian, F. 1985. On the age of "Yuanmou Man"-a discussion with Liu Tung- sheng et al. Act. Anthropol. Sin. 4:324-31.
  13. Robinson, J.T. 1953. Meganthropus, australopithecines and hominids. Amer. J. Phys. Anthropol. 11:1-38.
  14. Swisher, C. C., Curtis, G. H., Jacob, T., Getty, A.G., Suprijo, A. and Widiasmoro. 1994. Age of the earliest known hominids in Java, Indonesia. Science 263:1118-21.
  15. Tattersall, I. 1986. Species recognition in human paleontology. J. Hum. Evol. 15:165-175.
  16. Tobias, P.V. and von Koeningswald, G. 1964. A comparison between the Olduvai hominines and those of Java, and some implications for hominid phylogeny. Nature 204:515-18.
  17. Tyler, D. E. 1995. The current picture of hominid evolution in Java. Act. Anthropol. Sin. 14:313-323.
  18. Wood, B. 1992. Origin and evoultion of the genus Homo. Nature 355:783-90.
  19. Wood, B. and Turner, A. 1995. Out of Africa and into Asia. Nature 378:239-240.
  20. Zhang, Y.Y. 1984. The "Australopithecus" of west Hubei and some early Pleistocene hominids of Indonesia. Act. Anthropol. Sin. 3: 85-92.

* Research Associate, Laboratory for Human Evolutionary Studies, Museum of Vertebrate Zoology, University of California, Berkeley, CA 95720 USA

Back to the top | Back to home page