Chinese Science Bulletin **** Vol. ** No. ** **** 2782 nately, Asia has been the focus of renewed interest after
having been largely ignored as an area for human origins
Early Pleistocene hominid research for many years. Further more, in opposition to
the hypothesis of an Africa center, the multiregional hy-
teeth recovered in Mohui cave pothesis advocated by Franz Weidenreich who had made
in Bubing Basin, Guangxi, great contributions to the research on Homo erectus in
Zhoukoudian has been revived[2]. Recently, some finds
South China from Asia have been dated to ages nearly equivalent to
early sites in Africa. Here we report the newest discovery
of early Pleistocene hominid fossils from Mohui cave in
WANG Wei1,2, Richard Potts3, HOU Yamei4,
the Bubing Basin, Guangxi Zhuang Autonomous Region,
CHEN Yunfa2, WU Huaying2 YUAN Baoyin5
South China.
& HUANG Weiwen4
1 Geological and archaeological background
1. Faculty of Earth Science, China University of Geoscience, Wuhan
The Bubing Basin is a small basin associated with the
430074, China;
2. Natural History Museum of Guangxi Zhuang Autonomous Region, larger Bose Basin and lies 12 km to the west of Tiandong
Nanning 530012, China;
County. The basin is formed on limestone of Devonian to
3. Human Origins Program, National Museum of Natural History,
Tertiary age, with a length of 16 km by 2 km wide. There
Smithsonian Institution, Washington DC 20560-0112, USA;
is a long and narrow horst that divides the two basins from
4. Institute of Vertebrate Paleontology and Paleoanthropology, Chinese
Academy of Sciences, Beijing 100044, China; north-west to south-east. Two small rivers flow out of the
5. Institute of Geology and Geophysics, Chinese Academy of Sciences,
mountains to the south, crossing the Bubing Basin and
Beijing 100029, China
downcutting the horst, to finally join the Youjiang River
Correspondence should be addressed to Wang Wei (email: wang-
(Fig. 1).
abplir@r.postjobfree.com)
Bottom of The Bubing Basin is relatively flat, with an
Abstract Two hominid teeth recovered in Mohui cave are
average elevation of 130 m above sea level (ASL), which
morphologically distinguished from Australopithecus in Af-
is a little higher than the Bose Basin. Karst mountains
rica, but close to Homo erectus in China. These teeth are
develop in the south-west margin of the basin, where there
therefore provisionally assigned to Homo erectus. The associ-
are also some isolated limestone hills within the bounda-
ated mammalian fauna include Gigantopithecus blacki, Nes-
ries. At least seven layers of level caves develop in these
toritherium sp., Sus xiaozhu, Sus peii and Ailuropoda microta,
which are typical early Pleistocene taxa in South China. The mountains and hills. The lowest layer of caves is just de-
general characteristics of the Mohui faunal assemblage are veloping at the elevation of 120 m ASL; the highest layer
similar to the Longgupo site, which is dated to 2 Ma, imply- of caves is about 250 m ASL (130 m above the level of
ing a contemporaneity for the two sites. To date, compared the local river). Mohui cave, at the sixth layer, is situated
with the discoveries in Africa, far fewer early Pleistocene
in a limestone half-mountainside in the south-east margin
hominid fossils have been recovered in Asia, and there are
of the basin (107 00.13 E, 23 34.891 N). The chamber is
intensive controversies concerning their stratigraphic prove-
50 m in length, 2 6 m in breadth and 5 6 m in height
nience and typological and temporal positions. The hominid
(Fig. 2). Large wavy pits on the cave walls indicate an
fossils from Mohui cave, with their reliable biostratigraphic
positions and distinct typological features, provide important ancient underground river that flowed from west to east.
evidence regarding the issue of early human origins and evo- Stalactites are scattered inside. The eastward facing cave
lution. entrance is 215 m ASL and 65 m above the nearby 1st
terrace of the basin (Fig. 3).
Keywords: Mohui cave, hominid teeth, stone artifact, mammalian
fauna, early Pleistocene. In March 2001, some mammalian fossils were first
recovered in the deposits and one piece of a large hominid
DOI:10.1360/982004-614
tooth was collected from original sediments in Mohui
cave when a research project in the Bubing Basin was
Since the 1950s, researchers who examine the issue of
carried out by the Natural History Museum of Guangxi.
human beginnings often turn to Africa where there is a
Although local villagers had dug the cave for fertilizer in
picture of human origins and evolution based on African
the mid 20th century, partial deposits remain, that retain
hominid fossils with ages that are constantly revised to be
the complete geological profile. From October to Decem-
older and older. However, there are many other unsolved
ber, 2002 and November, 2003, two seasons of excavation
problems about early human origins and evolution that
took place. As a result, abundant mammalian fossils and a
may be solved by looking outside Africa. Over seventy
few stone artifacts were recovered, including another
years ago, Asia was described as a dispersal center of the
hominid tooth reported in the present paper.
earliest human industry, and a key arena for human evolu-
From east to west inside the cave, three areas were de-
tion. It was deemed nearly unimaginable to unravel the
signed as A, B and C based on the composition of the de-
complex issue of human origins without Asia[1]. Fortu-
Chinese Science Bulletin Vol. 50 No. 23 December 2005 2777
sediment per layer to sieve only.
The stratigraphic sequence can be divided from upper
to lower as follows (Fig. 2):
(5) Grey-white flow-stone covered with stalagmites yield-
ing stone artifacts below 5-35 cm
(4) Brown sandy clay with breccia, containing no fossils
60 cm
(3) Brown sandy clay with rare breccia, yielding abundant
mammalian fossils and some human teeth 210 cm
(2) Light brown silt clay and sandy clay, bearing micro-
mammalian fossils 400 cm
(1) Dark brown fluvial sands with oblique bedding
20-40 cm
Fig. 1. Map showing the location and plan of Mohui cave.
posits. A square of 1.4 m 1 m was selected in Area B,
and excavated in 5 cm levels; four squares of 2 m 2 m
were chosen in Area A (A0 A3), and excavated in 10 cm
levels; and one square of 2 m 2 m was investigated in Fig. 2. Map showing stratigraphic sequence of Mohui cave.
Area C (C1) and excavated in 20 cm levels. In Areas A
and B, all of the sediments were then water-screened for
2 Hominid and Gigantopithecus teeth
micro-mammal collection using 35 mesh sieves. Consid-
A complete lower right second molar (M2)(No.
ering fossils were relatively rare in C1, we took 50 kg of
2778 Chinese Science Bulletin Vol. 50 No. 23 December 2005
ARTICLES
tal roots of the tooth are very strong, and exhibit slightly
MH0001 in Fig. 4(a) (e)) was collected from the lower
distal curvature. The mesial or anterior surface of the me-
part in square B (elevation 1.6 m). The second specimen
sial root has a broad, shallow fossa which divides the root
is a broken upper left molar (M1/M2) (No. MH0018 in Fig.
into two tapered tips. The buccal-lingual breadth of the
4(f)) collected from disturbed deposits
mesial root is 14.5 mm and the distal root is 10.8 mm.
MH0001 is milky white, with a large crown and very
MH0018 is brown and large in size, with only the buc-
strong roots It is well preserved except for the tips of the
cal half preserved. The lingual half is broken, and most of
roots that are absent due to porcupine gnawing. The five
the roots are absent except for a small portion on the buc-
cusps have only been slightly worn, so their relationship is
cal side. Based upon what is preserved, the crown has a
relatively clear. There is a larger oval interproximal facet
rhomboid shape with a mesial-distal length of 14.3 mm.
at the middle of the mesial surface, while on the distal
The paracone and metacone are integrated and are similar
surface there is a smaller oval interproximal facet near the
in height. The buccal surface is relatively convex, while
lingual side, adjacent to another, much smaller oval facet.
the distal and mesial surfaces are flat. The mesial inter-
According to these features, the shapes of the interproxi-
proximal facet is near the buccal side and large; the distal
mal facets and their relationships, MH0001 is probably a
facet is smaller and far from the buccal side. The meta-
lower second molar.
cone is larger than the paracone, which is separated from
The crown is nearly rectangular, and its width is longer
the hypocone but connected with the protocone. The buc-
than length. It measures 15.3 mm in mesial-distal length
cal furrow is shallow and straight, extending down 1/3 of
and 16.1 mm in the buccal-lingual width, yielding a
the crown. The two cusps appear simple; the occlusal sur-
length-breadth index ([(MD/BL) 100]) of 95. The crown
faces are flat and relatively smooth with no wrinkles,
is broader mesially than distally. The protoconid is convex
As MH0001 is relatively intact and can be compared
and the metaconid is the highest cusp. The mesial surface
with known taxa, although the size of the tooth is within
is flat while the distal, lingual and buccal surfaces are
the range of variation for Pongo, the morphology is dif-
relatively convex. The five cusps on the occlusal surface,
ferent. MH001 has simple furrows on the buccal surface
ordered from larger to smaller sequentially, are the proto-
and lacks the abundant enamel wrinkles that are charac-
conid, metaconid, hypoconid, entoconid and mesoconid.
teristic of Pongo. MH001 is square in overall shape,
These are separated by a Y pattern of fissures, indicat-
yetalmost all of the second lower molars of Pongo are
ing the Dryopithecus pattern . All of the furrows are
rectangular, so their breadth/length indices are distinctly
shallow and simple with no wrinkles. The mesial and dis-
Fig. 3. Geomorphological landscape of Bubing Basin in Guangxi.
Chinese Science Bulletin Vol. 50 No. 23 December 2005 2779
Fig. 4. Hominid, Gigantopithecus teeth and stone artifacts recovered from Mohui cave. (a) (f) Hominid teeth, MH0001, right M2. (a)
occlusal view; (b) mesial view; (c) distal view; (d) buccal view; (e) lingual view, (f) MH0018, left M1/M2, occlusal view. (g) (j)
Gigantopithecus blacki. (g) left P3; (h) left M1/M2; (i) left M1/M2; (j) left M3; (k), (l) stone artifacts.
different. Thus, MH0001 cannot be assigned to Pongo. MH0001 presents the Dryopithecus pattern, the grooves
The mesial-distal and buccal-lingual dimensions of between the cusps are very simple, and the lingual and
MH0001 are within the range of variation of Australo- distal surfaces are convex. These detailed characteristics
are commonly seen in H. erectus from ZKD[4]; the only
pithecus africanus (MD = 14.3 17.8 mm, BL = 12.7
difference is that the protoconid of MH0001 is large and
16.2 mm), A. robustus (MD = 14.8 18.2 mm, BL = 12.8
strongly buccally projecting. The reason why the size of
16.2 mm) and A. boisei (MD = 16.4 20.0 mm, BL =
the Mohui hominid tooth is markedly larger than Chinese
15.8 18.6 mm) [3]. However, its breath/length index is
H. erectus may be that it represents a much earlier repre-
outside the known variation of these three species (103 sentative of the taxon.
118, 102 117 and 104 118 respectively)[3]. Furthermore, Compared with hominid teeth recovered recently in Ji-
although the index of MH0001 is within the A. afarensis anshi, Hubei Province, MH0001 is larger with a relatively
range of variation (93 118), its dimensions are larger low crown, and it has a different crown shape as illus-
(MD = 12.1 15.4 mm, BL = 12.1 15.2 mm) [3]. More- trated by the difference in the breadth/length index (the
over, there is no evidence supporting the existence of Aus- first right lower molar, PA1277: MD = 14.6 mm BL =
13.6 mm)[7]. Thus, the teeth from Jianshi and Mohui are
tralopithecus outside Africa to date, it is difficult to attrib-
ute MH0001 to this genus. different, yet their sizes are all larger than those of teeth
In comparison with Homo erectus in China, MH0001 is from ZKD, and they therefore might be earlier than the
larger than the teeth from Zhoukoudian (MD = 11.3 13.2 ZKD site.
mm, BL = 11.1 13.0 mm)[4], Hexian (MD = 12.6 mm, As a result, MH0001 is viewed as most similar to H.
erectus in China.
BL = 13.0 mm)[5] and Lantian (MD = 13.3 mm, BL = 13.3
mm)[6], though the breadth/length index is very similar (99 Gigantopithecus teeth (Fig. 3(g) (j)): Thirteen in total,
including one right lower canine, three left upper premo-
115 for ZKD, 97 for Hexian and 100 for Lantian).
lars, one right first or second upper molar, one left third
Morphologically, as noted above, the cusp arrangement of
2780 Chinese Science Bulletin Vol. 50 No. 23 December 2005
ARTICLES
upper molar, three left first or second lower molars, one tocene. Based on these characters, the age of the Mohui
right first or second lower molar and two left third lower fauna should be close to the above two sites.
molars. These teeth have similar morphological features to
5 Preliminary conclusions
other Gigantopithecus from sites in China, thus they were
assigned to G. blacki. The dimensions of the teeth sizes are In Mohui cave, a mammalian fauna with typical early
similar to those from early Pleistocene sites in Liucheng[8], Pleistocene features of the Gigantopithecus fauna of South
Guangxi and Longgupo, Chongqing[9], but smaller than China provides a strong biostratigraphic foundation for
those from middle Pleistocene sites in China. interpreting the age of the deposits as early Pleistocene .
3 Stone artifacts Considering that the general aspect of the fauna is similar
to the aspects of Liucheng and the Longgupo sites, which
Eight stone artifacts have been recovered in Mohui
has been dated to about 2 M by paleomagnetism and ESR
cave. Two of them were separately collected from a ce-
analyses[9], Mohui cave may represent one of the earliest
mented breccia bed at the cave entrance and excavated
hominid sites in Asia. However, this preliminary conclu-
from the cover flowstone in square A2, while the other six
sion should be substantiated with further chronological
were collected from disturbed deposits in Area C. The raw
studies.
materials are quartz in two cases; the other six are made
on sandstone. Presently, considering there is no evidence of Australo-
Three cleavers were made from flakes, and two picks pithecus found outside Africa, and the hominid teeth in
were made from a cobble and a flake. These five tools are Mohui cave have many features of Chinese H. erectus,
retouched by hard-hammer technique with alternate per- we temporarily assign these teeth to Homo erectus. Up to
cussion. One of the cleavers is retouched from the dorsal now, Modjokerto[12] in Java, Yuanmou[13], Longgupo[9],
surface of the flake toward the right and distal edges of the and Jianshi[7] in China are the rare early Pleistocene sites
ventral surface. The other three artifacts are flakes (Fig. 4). yielding hominid fossils in East and Southeast Asia.
The stone artifacts recovered in Mohui cave indicate the Moreover, there are intensive controversies concerning
existence of early human activities in the cave, but it is their stratigraphic proveniences and temporal positions.
impossible that humans came into the cave when it was The hominid fossils from Mohui cave, with their reliable
underground. Therefore, early humans could only have
biostratigraphic position and distinct typological features,
used the cave after it had been uplifted above the ground
provide important evidence for addressing the role of Asia
level. So, the age of human activities in Mohui cave
in early hominid evolution.
should be later than the deposition of the sediments with
Recently, in the Renzidong site located at the lower
the hominid and Gigantopithecus teeth. There is a similar
Yangzi River, artifacts were recovered which might date to
technology of tool manufacturing between Mohui cave
the beginning of the early Pleistocene based on the faunal
and the Bose Basin Paleolithic sites, which implies a pos-
assemblage[14]; and in the Nihewan Basin the evidence of
sibility of affinity among the sites. Thus, the Mohui cave
early human activities dates prior to 1.66 Ma[15]. So the
site provides a clue for solving the scientific issues sur-
developing picture of early hominid evolution is presently
rounding the Bose Paleolithic, which is dated to 0.803
undergoing major revisions as we learn more about the
Ma[10].
contributions of China, and Asia, in general, to paleoan-
4 Mammalian fauna thropological research.
Mammalian fossils recovered in Mohui cave include: Acknowledgements The authors would like to thank colleagues and
Macaca sp., Leopoldamys sp., Niviventer sp., Hapalomys professors from the Natural History Museum of Guangxi, the Tiandong
sp., Typhlomys sp., Hystrix subcristata, Hystrix magna, Museum and the Institute of Vertebrate Paleontology and Paleoanthro-
pology, Chinese Academy of Sciences for their help in field excavations,
Cuon (Cyon) sp., Nestoritherium sp., Ailuropoda microta,
fossil identification and for access to reference collections. We thank
Ursus thibetanus, Mustelidae, Felis teilhardi, Stegodon sp., Prof. Wu Xinzhi for his valuable suggestions. We thank Profs. Zheng
Sus xiaozhu, Sus peii, Cervidae, and Bovidae. In addition, Liang and Ji Xueping in the Institute of Archeology of Yunnan Province
for their gracious help with fossil comparisons. We also thank Prof. Guo
sixteen primate teeth are identified as hominoid according
Zhengtang from the Institute of Earth Environment, Chinese Academy of
to their morphological characteristics. Sciences and Prof. Zhu Rixiang and Dr. Deng Chenglong from the Insti-
Among the fauna, Ailuropoda microta, Nestoritherium tute of Geology and Geophysics, Chinese Academy of Sciences for their
support and encouragement on this project. This work is supported by the
sp., and Sus peii are typical members of the early Pleisto-
National Natural Science Foundation of China (Grant No. 40163001),
cene Liucheng fauna[11]. These taxa are also found in the the Emergency Excavation Fund from the National Culture Relics Bu-
Longgupo faunal assemblage[9]. The size of the Ailuro- reau of China, special funds of Talents Project in the New Century of
poda microta in Mohui cave is relatively smaller, which the Guangxi Government (Grant No. 2001216), and the National Science
Foundation, USA.
should represent an early form of the species in the Pleis-
Chinese Science Bulletin Vol. 50 No. 23 December 2005 2781
References 9. Huang, W. P., Ciochon, R. L., Gu, Y. M. et al., Early Homo and as-
sociated artifacts from Asia, Nature, 1995, 378(16): 275 278.
1. Boule, M., Breuil, H., Licent, E. et al., Le Paleolithique de la Chine,
10. Hou, Y. M., Potts, R., Yuan, B. Y. et al., Mid-Pleistocene
Archives de L Institut de paleontologie humaine, 1928, (4): 1
Acheulean-like stone technology of the Bose Basin, south China,
138.
Science, 2000, 287: 1622 1626.
2. Wu, R. K., On Weidenreich s work of Sinanthropus pekinensis and
11. Pei, W. Z., Carnivora, Proboscidea and Rodentia from Liucheng
his theories of human origins, Acta Anthropologica Sinica (in Chi-
Gigantopithecus Cave and other caves in Guangxi, Memoirs of In-
nese), 1999, 18 (3): 161 164.
stitute of Vertebrate Paleontology and Paleoanthropology, Acade-
3. White, T. D., Australopithecus africanus: its phyletic position re-
mia Sinica, No. 18 (in Chinese), Beijing: Science Press, 1987, 1
considered, South African Journal of Science, 1981, 77(5): 445
119.
470.
12. Larick, R., Ciochon, R. L., The African emergence and early Asian
4. Weidenreich, F., The dentition of Sinanthropus pekinensis: a com-
dispersals of the genus Homo, American Scientist, 1996, 84: 538
parative odontography of the hominids, Palaeontologia Sinica,
551.
1936, New Series D. No. 1: 1 162.
13. Hu, C. Z., Ape-man teeth from Yuanmou, Yunnan, Acta Geologica
5. Wu, R. K., Dong, X. R., Preliminary study of Homo erectus re-
Sinica (in Chinese), 1973, 47 (1): 65 69.
mains from Hexian, Anhui, Acta Anthropologica Sinica (in Chi-
14. Jin, C. Z., Han, L. G., Wei, G. B., New discovery of early Pleisto-
nese), 1982, 1 (1): 2 11.
cene Paleoliths from Renzidong, Fanchang County, Anhui Province,
6. Wu, R. K., Mandible of the Sinanthropus-type discovered at
Acta Anthropologica Sinica (in Chinese), 1999, 18(1): 70 71.
Lantian, Shanxi, Vertebrata PalAsiatica (in Chinese), 1964, 8(1): 1
15. Zhu, R. X., Potts, R., Xie, F. et al., New evidence on the earliest
17.
human presence at high northern latitudes in northeast Asia, Nature,
7. Zheng, S. H., Jianshi Hominid Site (in Chinese), Beijing: Science
2004, 431: 559 562.
Press, 2004, 1 412.
8. Wu, R. K., The mandibles and dentition of Gigantopithecus, Pale- (Received February 6, 2005; accepted June 10, 2005)
ontol. Sinica (in Chinese), 1962, New Series D, No. 11: 1 62.
2782 Chinese Science Bulletin Vol. 50 No. 23 December 2005