19. Tedford, R. H., et al. in Late Cretaceous and Cenozoic Mammals of North America (ed. Woodburne,
M. O.) 169–231 (Columbia Univ. Press, New York, 2004).
20. Ginsburg, L., Maridet, O. & Mein, P. Un Ailurinae (Mammalia, Carnivora, Ailuridae) dans le Mioce
moyen de Four (Ise
re, France). Geodiversitas 23, 81–85 (2001).
t, M. New dental remains of an extinct lesser panda–morphotype or new species? J. Vert.
Paleontol. 17, 58A (1997).
22. Boyd-Dawkins, W. On Ailurus anglicus, a new carnivore from Red Crag. Q. J. Geol. Soc. Lond. 44,
23. Kormos, T. H. Beitrage Zur Kenntnis der Gattung Parailurus. Mitt. Jahrb. K. Ung. Geol. Anst. 30, 1–39
24. Schlosser, M. Parailurus anglicus and Ursus bo
ckhi, aus den Ligniten von Baro
k in Ungarn. Mitt. Jahrb. K. Ung. Geol. Anst. 13, 66–95 (1899).
25. Morlo, M. & Kundra
t, M. The ﬁrst carnivoran fauna from the Ruscinium (Early Pliocene, MN 15) of
ontol. Z. 75, 163–187 (2001).
26. Wang, X. New cranial material of Simocyon from China, and its implications for phylogenetic
relationship to the red panda (Ailurus). J. Vert. Paleontol. 17, 184–198 (1997).
27. Schmidt-Kittler, N. in The Vertebrate Locality Maramena (Macedonia, Greece) at the Turolian-
Ruscinian Boundary (Neogene) (ed. Schmidt-Kittler, N.) 75–86 (Mu
Abhandlungen, Munich, 1995).
28. Zdansky, O. Jungtertia
re carnivoren Chinas. Palaeont. Sin. C 2, 1–149 (1924).
29. Stach, J. Arctomeles pliocaenicus, nowy rodzaj i gatunek z podrodziny borsukowatych (Arctomeles
pliocaenicus nov. gen. & sp. from Weze). Acta Geol. Pol. 2, 129–157 (1951).
30. Parmalee, P. W., Klippel, W. E., Meylan, P. A. & Holman, J. A. A late Miocene–early Pliocene
population of Trachemys (Testudines: Emydidae) from east Tennessee. Ann. Carnegie Mus. 71,
Supplementary Information accompanies the paper on www.nature.com/nature.
Acknowledgements S.C.W. wishes to thank former Tennessee Governor Don Sundquist for
protecting the fossil deposit and awarding signiﬁcant funds for its preservation and study; the
Tennessee Department of Transportation, East Tennessee State University, Ofﬁce of Research and
Sponsored Programs, and the College of Arts and Sciences for their continuing support of this
project; and Larry Bristol for discovering the Pristinailurus M
and bringing it to my attention.
X.W. wishes to acknowledge the Chinese National Natural Science Foundation and the National
Geographic Society for support in comparative studies.
Competing interests statement The authors declare that they have no competing ﬁnancial
Correspondence and requests for materials should be addressed to X.W. (firstname.lastname@example.org).
New evidence on the earliest human
presence at high northern latitudes in
R. X. Zhu
, R. Potts
, F. Xie
, K. A. Hoffman
, C. L. Deng
, C. D. Shi
Y. X. Pan
, H. Q. Wang
, R. P. Shi
, Y. C. Wang
, G. H. Shi
Paleomagnetism Laboratory, Institute of Geology and Geophysics, Chinese
Academy of Sciences, Beijing 100029, China
Human Origins Program, National Museum of Natural Histor y, Smithsonian
Institution, Washington, DC 20560-0112, USA
Hebei Province Institute of Cultural Relics, Shijiazhuang 050000, China
Physics Department, California Polytechnic State University, San Luis Obispo,
California 93410, USA
The timing of early human dispersal to Asia is a central issue in
the study of human evolution. Excavations in predominantly
lacustrine sediments at Majuangou, Nihewan basin, north China,
uncovered four layers of indisputable hominin stone tools. Here
we report magnetostratigraphic results that constrain the age of
the four artefact layers to an interval of nearly 340,000 yr between
the Olduvai subchron and the Cobb Mountain event. The lowest
layer, about 1.66 million years old (Myr), provides the oldest
record of stone-tool processing of animal tissues in east Asia. The
highest layer, at about 1.32 Myr, correlates with the stone tool
layer at Xiaochangliang
, previously considered the oldest
archaeological site in this region. The ﬁndings at Majuangou
indicate that the oldest known human presence in northeast Asia
at 408 N is only slightly younger than that in western Asia
result implies that a long yet rapid migration from Afr i ca,
possibly initiated during a phase of warm climate, enabled
early human populations to inhabit northern latitudes of east
Asia over a prolonged period.
The Majuangou (MJG; 408 13.517
N, 1148 39.844
E) section lies
in the eastern margin of the Nihewan basin (Fig. 1). It is a lacustrine
sequence with brief intervals of wetland and lake-margin sediments,
and consists mainly of greyish-yellow and greyish-green clay, silty
clay and silt. It is underlain by red Jurassic volcanic breccia. Loess
sediments at the top of the section have been subjected to erosion.
The four artefact layers found in the MJG section are, from top
to bottom, Banshan
(44.3–45.0 m), MJG-I (ref. 5; 65.0–65.5 m),
MJG-II (73.2–73.56 m) and MJG-III (75.0–75.5 m) (Fig. 2).
The Banshan artefact layer, discovered and excavated in 1990
area, 70 cm thick), contained 95 stone artefacts in gravelly
. Excavation of MJG-I in 1993 (20 m
, 50 cm) yielded 111
stone tools in clayey silt
. Renewed excavation at Majuangou in 2001
and 2002 uncovered 226 artefacts in brown clayey silt of MJG-II
, 36 cm) and 443 artefacts in greyish-black silty clay of MJG-
III (85 m
, 50 cm). The sediments, numerous molluscan shells
(Gyraulus chihliensis and Planorbis youngi), and leaves and fruits
of aquatic plants (for example Trapa sp.) in MJG-III indicate a low-
energy lakeshore or marsh environment rich in organic materials.
The in situ artefact density in this layer was low overall (10.4
artefacts per m
), but artefacts and fauna in some 5-cm-thick
Figure 1 Location of the Majuangou and Haojiatai sections in the Nihewan basin. Some
sites mentioned in the text, Xiaochangliang, Donggutuo, Gongwangling and Xihoudu, are
indicated. The Qinling Mountains (bottom left) are the traditional dividing line between
north and south China. The Yellow River and Yangtze River are the major river systems in
north and south China, respectively.
letters to nature
NATURE | VOL 431 | 30 SEPTEMBER 2004 | www.nature.com/nature 559
units were as high as 170 specimens per m
over the entire
excavation and 620 specimens per m
in a single square metre.
These concentrations are comparable to those in African Plio-
Pleistocene archaeological sites
. MJG-III exhibits remarkable
preservation demonstrated by animal-trampled sedimentary sur-
faces, very fresh condition of the artefacts, and fossil bone surface
details that include tool percussion marks and numerous ﬁne
scratches attributed to trampling.
The four Majuangou layers preserve indisputable stone tools
indicative of repetitive stone-on-stone percussion ﬂaking (Fig. 3a–e).
The assemblages are dominated by core fragments that exhibit
truncated negative scars and by ﬂakes with percussion platforms
and bulbs. Each artefact layer also contained ﬂaked cores that show
striking platforms and multiple overlapping negative scars. The
cores can be placed in artefact categories of chopper, scraper and
polyhedron also known in African Plio-Pleistocene stone tool
assemblages. The MJG cores were chipped from angular fragments
of chert, sandstone, quartz and andesite, and thus differ from typical
East African Oldowan artefacts made on rounded lava cobbles. The
artefacts of MJG-I to MJG-III are signiﬁcantly outsized clasts in very
ﬁne-grain depositional contexts, which indicates the hominin
transport of rocks from outcrop sources over an unknown distance.
Vertebrate fossil remains were best represented at MJG-III
(N ¼ 1,014), most of which are attributable to Elephas sp. Other
taxa include horse Equus sanmeniensis, hyena Pachycrocuta sp.,
rhinoceros Coelodonta antiquitatis, deer Cervus sp., bovid Gazella
Figure 2 Lithostratigraphy and magnetostratigraphy and correlation with the
geomagnetic polarity timescale (GPTS)
. a, Haojiatai; b, Majuangou. The four artefact
levels are shown. To conﬁrm the palaeomagnetic results, two sets of parallel samples
(black and blue circles in b) with independent orientation were measured on the
Majuangou outcrop and well samples. Inc., inclination; Dec., declination; VGP, virtual
letters to nature
NATURE | VOL 431 | 30 SEPTEMBER 2004 | www.nature.com/nature560
sp., ostrich Struthio sp., and Carnivora gen. et sp. indet. The
mammals from MJG-III and the Banshan layer are typical of the
taxa recorded in the Xiaochangliang site
. Evidence of sedimentary
abrasion due to trampling hinders an unambiguous identiﬁcation
of purposeful tool butchery marks at MJG; however, several
diaphysis fragments of deer- and horse-sized mammalian long
bones show tool percussion damage indicative of marrow extraction
(Fig. 3f). Although there was accumulation of tools and fossil bones
during depositional hiatuses, there is no evidence of deﬂation
surfaces that might have associated objects from separate strata.
Accumulation of both artefacts and fossil animals was therefore
contemporaneous, and the presence of tool-modiﬁed bones implies
that hominins acquired food from the animal remains preserved in
the MJG layers.
We examined the 95.6-m-thick MJG section palaeomagnetically
and compared it with a 128.8-m-thick parallel section 1.5 km away
named Haojiatai (HJT; 408 13.240
N, 1148 38.938
E) (Fig. 1). The
HJT section, which consists of ﬂat-lying beds exposed in deep
gullies, preserves the entire upper part of the Nihewan sequence,
including the Holocene soil and the last glacial loess (Fig. 2). All HJT
samples were collected from natural outcrops. Samples from the top
of the MJG section to a depth of 75.2 m came from natural outcrops;
two wells were dug to extend the MJG palaeomagnetic record. The
ﬁrst well, about 20 m southeast of the MJG-III site, recorded a depth
interval from 75.2 to 86.2 m; the second well, about 50 m northwest
of the site, recorded a depth interval from 75.2 to 95.6 m. The
sedimentary sequences at MJG and HJT are well correlated by two
distinctive marker layers: a conglomerate layer (found at the 45-m
depth at MJG and the 105-m depth at HJT) and a greyish-yellow
clay layer with molluscan fossils (found at the 66-m depth at MJG
and the 122.4-m depth at HJT) (Fig. 2).
Rock magnetic methods, which included anisotropy of magnetic
susceptibility, thermomagnetic analysis and hysteresis measure-
ments, showed that magnetite of pseudo-single-domain grain size
is the principal carrier of the magnetic remanence and that the
sedimentary magnetic fabric had been unperturbed since depo-
sition. After this check on the reliability of the two palaeomagnetic
records, we established the polarity stratigraphy through stepwise
demagnetization of the natural remanent magnetization. Complete
information on rock magnetic methods and demagnetization of
the natural remanent magnetization used here is given in Sup-
After the removal of secondary remanent magnetization com-
ponents from each sample through thermal and/or alternating ﬁeld
demagnetization procedures, virtual geomagnetic pole latitudes
were determined from the characteristic remanent magnetization
vector directions. These virtual geomagnetic poles were sub-
sequently used to deﬁne the succession of magnetostratigraphic
polarity in the two sections (Fig. 2).
Four magnetozones are recognized in the HJT section: two with
normal polarity, N1 (0–49.0 m) and N2 (75.8–80.2 m); and two
with reverse polarity, R1 (49.0–75.8 m) and R2 (80.2–128.8 m). In
the MJG section there are ﬁve magnetozones: two normal and three
reverse. These magnetozones correlate to the polarity sequence
at HJT as follows: N2 (17.2–22.0 m) and N3 (85.0–90.5 m); R1
(0–17.2 m), R2 (22.0–85.0 m) and R3 (90.5–95.6 m). The sediment
layers containing stone artefacts all occur within magnetozone R2 at
Because the Holocene soil, the last glacial loess, and soil associ-
ated with the last interglacial overlay the HJT lacustrine sequence,
the magnetozones determined for HJT can readily be correlated to
the geomagnetic polarity timescale
. HJTmagnetozones N1 and N2
correspond to the Brunhes chron and the Jaramillo subchron,
respectively; thus, magnetozones N2 and N3 in the MJG section
correspond to the Jaramillo subchron and the Olduvai subchron,
respectively. Hence, these Nihewan basin sediments were deposited
from just before the onset of the Olduvai subchron into the Brunhes
The magnetostratigraphic correlation is strengthened in that the
mammalian fauna from the Banshan and MJG-III layers is late
Pliocene to early Pleistocene in age
. In addition, two short
intervals of possible transitional ﬁeld behaviour, labelled e1 and e2
in Fig. 2, are recorded within magnetozone R2 at both MJG and HJT
(e1, 29.5–30.5 m at MJG and 88.7–89.9 m at HJT; e2, 36.5–37.3 m at
MJG and 94.1–94.7 m at HJT). Given that the duration of magne-
tozone R2 is about 0.70 Myr
between the termination of the
Olduvai subchron (1.77 Myr) and the onset of the Jaramillo sub-
chron (1.07 Myr)
the interpolated ages for e1 and e2 are 1.16 Myr
and 1.24 Myr, respectively, based on an averaged rate of sediment
deposition. These values are remarkably similar to the
determinations of 1.10–1.11 Myr (ref. 13) and 1.21–1.24 Myr (ref. 9)
for the Punaruu and Cobb Mountain geomagnetic events. It is
therefore possible that the sediments in both sections record not
only the coarse magnetostratigraphy of the Matuyama chron (that
is, the Jaramillo and Olduvai normal polarity subchrons) but also
some of its ﬁne structure.
The Banshan, MJG-I, MJG-II and MJG-III artefact layers within
magnetozone R2, reﬂecting brief episodes of wetlands or lake-
margin deposition within a largely lacustrine sequence, have mid-
way depths of 44.65, 65.25, 73.38 and 75.25 m, respectively (Fig. 2).
Again with the use of an averaged sediment accumulation rate for
magnetozone R2 at MJG, the ages of these four artefact layers can be
estimated at 1.32, 1.55, 1.64 and 1.66 Myr, respectively.
Figure 3 Stone artefacts and modiﬁed bones from Majuangou. a, Notch made on a ﬂake
(MJG-III). b, Chopper made on an angular fragment (MJG-II). c, Multi-platform polyhedron
made on an angular fragment (MJG-III). d, Scraper made on a ﬂake (MJG-III). e,
Hammerstone; arrow indicates the main battered end. f, Two mammalian long-bone shaft
fragments with impact notches and ﬂake scars (arrows) typical of tool percussion damage
(MJG-III). Scale bars, 1 cm.
letters to nature
NATURE | VOL 431 | 30 SEPTEMBER 2004 | www.nature.com/nature 561
The ages for MJG-I, MJG-II and MJG-III are considerably older
than previous age estimates of Palaeolithic sites in northern China
and indicate that humans might have reached northeast Asia earlier
than previously thought. Along with estimated ages for the sites of
Gongwangling (1.15 Myr)
and Xihoudu (1.27 Myr)
southern Loess Plateau and for Xiaochangliang (1.36 Myr)
Donggutuo (1.1 Myr)
sites in the Nihewan basin, our new results
imply an expansion and lengthy ﬂourishing of human groups from
northern to north-central China during the early Pleistocene.
The estimated age of 1.66 Myr for the MJG-III artefact layer pre-
dates the previous oldest age of unambiguous human presence at
408 N in East Asia by about 0.3 Myr. Our ﬁndings, particularly for
the MJG-III layer, document the oldest coexistence of stone tools
and man-made bone modiﬁcations in East Asia, indicating possible
continuity with the oldest stone tools and artiﬁcial bone modiﬁ-
cations reported in eastern Africa
. Archaeological evidence at
MJG indicates the oldest known use of animal tissues, especially
marrow processing, by early humans in Asia. The earliest archae-
ological level in the Nihewan basin is slightly younger than the
1.75 Myr estimated age for early humans at the Dmanisi site at 408 N
latitude in western Eurasia
. Our estimated ages also fall within the
1.66–1.51-Myr range for the earliest known human fossils in south-
. The combined evidence suggests that, near the start of
the Pleistocene, early human populations spread relatively rapidly
across Asia, presumably from an African origin, and reached at least
408 N latitude. Our ﬁndings further establish that the earliest
populations to reach northeast Asia were able to survive for at
least 500 kyr before the mid-Pleistocene onset of high-amplitude
Received 19 February; accepted 8 July 2004; doi:10.1038/nature02829.
1. Zhu, R. X. et al. Earliest presence of humans in northeast Asia. Nature 413, 413–417 (2001).
2. Gabunia, L. et al. Earliest Pleistocene hominid cranial remains from Dmanisi, Republic of Georgia:
Taxonomy, geological setting, and age. Science 288, 1019–1025 (2000).
3. Vekua, A. et al. A new skull of early Homo from Dmanisi, Georgia. Science 297, 85–89 (2002).
4. Wei, Q. Banshan Paleolithic site from the lower Pleistocene in the Nihewan Basin in northern China.
Acta Anthropol. Sinica 13, 223–238 (1994).
5. HPICR, Papers on Archaeology in Hebei Province 30–45 (East, Beijing, 1998).
6. Potts, R. Early Hominid Activities at Olduvai (Aldine de Gruyter, New York, 1988).
7. Potts, R., Behrensmeyer, A. K. & Ditchﬁeld, P. Paleolandscape variation and Early Pleistocene hominid
activities: Members 1 and 7, Olorgesailie Formation, Kenya. J. Hum. Evol. 37, 747–788 (1999).
8. Tang, Y. J., Li, Y. & Chen, W. Y. Mammalian fossils and the age of Xiaochangliang paleolithic site of
Yangyuan, Hebei. Vertebrata Palasiatica 33, 74–83 (1995).
9. Berggren, W. A., et al. in Geochronology, Timescales, and Stratigraphic Correlation (eds Berggren, W. A.,
Kent, D. V., Aubry, M. & Hardenbol, J.) 129–212 (SEPM Spec. Publ. 54, Tulsa, Oklahoma, 1995).
10. Wei, Q., et al. in Evidence for Evolution
Essays in Honor of Prof. Chungchien Yong on the Hundredth
Anniversary of His Birth (ed. Tong, Y.) 193–207 (Ocean, Beijing, 1997).
11. Huang, W. P. & Fang, Q. R. Wushan Hominid Site 105–109 (Ocean, Beijing, 1991).
12. Qiu, Z. X. Nihewan fauna and Q/N boundary in China. Quat. Sci. 20, 142–154 (2000).
13. Singer, B. S. et al. Dating transitionally magnetized lavas of the late Matuyama chron: Toward a new
Ar timescale of reversals and events. J. Geophys. Res. 104, 679–693 (1999).
14. An, Z. S. & Ho, C. K. New magnetostratigraphicdates of Lantian Homo erectus. Quat. Res. 32, 213–221
15. Zhu, R., An, Z., Potts, R. & Hoffman, K. A. Magnetostratigraphic dating of early humans in China.
Earth Sci. Rev. 61, 341–359 (2003).
16. Quaternary Research Association of China, Li, H. M. & Wang, J. D. Quaternary Geology and
Environment of China 33–38 (Ocean, Beijing, 1982).
17. Semaw, S. et al. 2.5-million-year-old stone tools from Gona, Ethiopia. Nature 385, 333–336 (1995).
18. de Heinzelin, J. et al. Environment and behavior of 2.5-million-year-old Bouri hominids. Science 284,
19. Swisher, C. C. III et al. Age of the earliest known hominids in Java, Indonesia. Science 263, 1118–1121
20. Larick, R. et al. Early Pleistocene
Ar ages for Bapang Formation hominins, Central Java,
Indonesia. Proc. Natl Acad. Sc i. USA 98, 4866–4871 (2001).
21. Potts, R. in Human Roots: Africa and Asia in the Middle Pleistocene (eds Barham, L. & Robson-Brown,
K.) 5–21 (Western Academic & Specialist Press, Bristol, 2001).
22. Clark, P. U., Alley, R. B. & Pollard, D. Northern Hemisphere ice-sheet inﬂuences on global climate
change. Science 286, 1104–1111 (1999).
23. Tian, J., Wang, P., Cheng, X. & Li, Q. Astronomically tuned Plio-Pleistocene benthic
O record from
South China Sea and Atlantic–Paciﬁc comparison. Earth Planet. Sci. Lett. 203, 1015–1029 (2002).
Supplementary Information accompanies the paper on www.nature.com/nature.
Acknowledgements We thank R. J. Enkin for providing palaeomagnetic software. This work was
supported by the National Natural Science Foundation of China and Chinese Academy of
Sciences. R.P. was supported by the US National Science Foundation and the Smithsonian
Human Origins Program. K.A.H. also received support from the US National Science
Competing interests statement The authors declare that they have no competing ﬁnancial
Correspondence and requests for materials should be addressed to R.X.Z. (email@example.com
and firstname.lastname@example.org) or R.P. (email@example.com).
Modelling the recent common
ancestry of all living humans
Douglas L. T. Rohde
, Steve Olson
& Joseph T. Chang
Department of Brain and Cognitive Sciences, Massachusetts Institute of
Technology, Cambridge, Massachusetts 02139, USA
7609 Sebago Road, Bethesda, Maryland 20817, USA
Department of Statistics, Yale University, New Haven, Connecticut 06520, USA
If a common ancestor of all living humans is deﬁned as an
individual who is a genealogical ancestor of all present-day
people, the most recent common ancestor (MRCA) for a ran-
domly mating population would have lived in the very recent
. However, the random mating model ignores essential
aspects of population sub struct ure, such a s the tendency of
individuals to choose mates from the same social group, and
the relative isolation of geographically separated groups. Here we
show that recent common ancestors also emerge from two
models incorporating substantial population substructure. One
model, designed for simplicity and theoretical insight, yields
explicit mathematical results through a probabilistic analysis. A
more elaborate second model, designed to capture his torical
population dynamics in a more realistic way, is analysed compu-
tationally throug h Monte Carlo simulations. These analyses
suggest that the genealogies of all living humans overlap in
remarkable ways in the recent past. In particular, the MRCA of
all present-day humans lived just a few thousand years ago in
these models. Moreover, among all individuals living more than
just a few thousand years earlier than the MRCA, each present-
day human has exactly the same set of genealogical ancestors.
In investigations of the common ancestors of all living humans,
much attention has focused on descent through either exclusively
maternal or exclusively paternal lines, as occurs with mitochondrial
DNA and most of the Y chromosome
. But according to the more
common genealogical usage of the term ‘ancestor’, ancestry encom-
passes all lines of descent through both males and females, so that
the ancestors of an individual include all of that person’s parents,
grandparents, and so on.
For a population of size n, assuming random mating (and so
ignoring population substructure), probabilistic analysis
proved that the number of generations back to the MRCA, T
has a distribution that is sharply concentrated around log
express this using the notation T
n, meaning that the
n converges in probability to 1 as n !1.In
contrast, the mean time to the MRCA along exclusively matrilineal
or patrilineal lines is approximately n generations
, and the distri-
bution is not sharply concentrated. For example, in a panmictic
population of one million people, the genealogical MRCA
would have lived about 20 generations ago, or around the year
AD 1400, assuming a generation time of 30 years. The MRCA along
letters to nature
NATURE | VOL 431 | 30 SEPTEMBER 2004 | www.nature.com/nature562