erratum: A new hominid from the Upper Miocene of Chad, Central Africa

Full text

A new hominid from the Upper Miocene
of Chad, Central Africa
Michel Brunet*, Franck Guy*†, David Pilbeam†, Hassane Taisso Mackaye‡, Andossa Likius*‡, Djimdoumalbaye Ahounta§, Alain Beauvilain§,
Ce
´
cile Blondel*, Herve
´
Bocherensk, Jean-Renaud Boisserie*, Louis De Bonis*, Yves Coppens{, Jean Dejax#, Christiane Denys#,
Philippe Duringerq,Ve
´
ra Eisenmann#, Gongdibe
´
Fanone§, Pierre Fronty*, Denis Geraads**, Thomas Lehmann*, Fabrice Lihoreau*,
Antoine Louchart††, Adoum Mahamat§, Gildas Merceron*, Guy Mouchelin*, Olga Otero*, Pablo Pelaez Campomanes‡‡,
Marcia Ponce De Leon§§, Jean-Claude Rage#, Michel Sapanetkk, Mathieu Schusterq, Jean Sudrek, Pascal Tassy#, Xavier Valentin*,
Patrick Vignaud*, Laurent Viriot*, Antoine Zazzo{{ & Christoph Zollikofer§§
* Faculte
´
des Sciences et CNRS UMR 6046, Universite
´
de Poitiers, 40 Avenue du Recteur Pineau, 86022 Poitiers Cedex, France
† Peabody Museum, Harvard University, 11 Divinity Avenue, Cambridge, Massachusetts 02138, USA
‡ Universite
´
de N’Djame
´
na, BP 1117, N’Djame
´
na, Tchad
§ Centre National d’Appui a
`
la Recherche, BP 1228, N’Djame
´
na, Tchad
k Institut des Sciences de l’Evolution, CNRS UMR 5554, Universite
´
de Montpellier II, Place E. Bataillon, 34095 Montpellier Cedex 5, France
{ Colle
`
ge de France, 3 rue d’Ulm, and # Muse
´
um National d’Histoire Naturelle et CNRS UMR 8569, rue Cuvier, 75005 Paris, France
q Centre de Ge
´
ochimie de la Surface, CNRS UMR 7517, Universite
´
Louis Pasteur, 1 rue Blessig, 67084 Strasbourg, France
** Centre National de Recherche Scientifique UPR 2147, 44 rue de l’Amiral Mouchez, 75014 Paris, France
†† Centres des Sciences de la Terre, CNRS UMR 5125, Universite
´
Claude Bernard, 27-43 Bd du 11 novembre 1918, 69622 Villeurbanne, France
‡‡ Museo de Ciencias Naturales, C/Guttierez Abascal 2, 28006 Madrid, Espan
˜
a
§§ Anthropologisches Institut/Multimedia Laboratorium, Universita
¨
tZu
¨
rich-Irchel, Winterthurer Str. 190, 8057 Zu
¨
rich, Switzerland
kk Centre Hospitalier Universitaire, Universite
´
de Poitiers, rue de la Mile
´
trie, 86021 Poitiers Cedex, France
{{ Centre National de Recherche Scientifique UMR 162, et Institut National de la Recherche Agronomique, Universite
´
Pierre et Marie Curie, 4 place Jussieu,
75252 Paris Cedex 05, France
...........................................................................................................................................................................................................................
The search for the earliest fossil evidence of the human lineage has been concentrated in East Africa. Here we report the discovery
of six hominid specimens from Chad, central Africa, 2,500 km from the East African Rift Valley. The fossils include a nearly
complete cranium and fragmentary lower jaws. The associated fauna suggest the fossils are between 6 and 7 million years old. The
fossils display a unique mosaic of primitive and derived characters, and constitute a new genus and species of hominid. The
distance from the Rift Valley, and the great antiquity of the fossils, suggest that the earliest members of the hominid clade were
more widely distributed than has been thought, and that the divergence between the human and chimpanzee lineages was earlier
than indicated by most molecular studies.
From their initial description in 1925
1
until 1995, hominids from
the Pliocene (5.3–1.6 million years, Myr) and late Upper Miocene
(,7.5–5.3 Myr) were known only from southern and eastern Africa.
This distribution led some authors to postulate an East African
origin for the hominid clade (where the term ‘hominid’ refers to any
member of that group more closely related to extant humans than to
the extant chimpanzee, Pan)
2,3
. The focus on East Africa has been
especially strong in the past decade, with the description of several
new forms from Kenya and Ethiopia, including Kenyanthropus
platyops (3.5 Myr; ref. 4); Australopithecus anamensis (3.9–4.1 Myr;
ref. 5); Ardipithecus ramidus ramidus (4.4 Myr; ref. 6); Ardipithecus
ramidus kadabba (5.2–5.8 Myr; ref. 7) and Orrorin tugenensis
(,6 Myr; refs 8, 9). The discoveries of A. ramidus ramidus, A. r.
kadabba and O. tugenensis have extended the human lineage well
back into the Miocene. However, the discovery of Australopithecus
bahrelghazali in Chad, central Africa
10,11
, demonstrated a consider-
ably wider geographic range for early hominids than conventionally
expected.
Since 2001, the Mission Pale
´
oanthropologique Franco–Tchadi-
enne (MPFT), a scientific collaboration between Poitiers University,
Ndjamena University and Centre National d’Appui a
`
la Recherche
(CNAR) (Ndjame
´
na), has recovered hominid specimens, including
a nearly complete cranium, from a single locality (TM 266) in the
Toros-Menalla fossiliferous area of the Djurab Desert of northern
Chad (Table 1). The constitution of the associated fauna suggests
that the fossils are older than material dated at 6 Myr from Lukeino,
Kenya
8,9
. Preliminary comparison with the fauna from the Nawata
formation at Lothagam, Kenya
12,13
, suggests that the fossils are from
the Late Miocene, between 6 and 7 Myr old. All six recovered
specimens are assigned to a new taxon that is, at present, the oldest
known member of the hominid clade.
Systematic palaeontology
Order Primates L., 1758
Suborder Anthropoidea Mivart, 1864
Superfamily Hominoidea Gray, 1825
Family Hominidae Gray, 1825
Sahelanthropus gen. nov.
Etymology. The generic name refers to the Sahel, the region of Africa
bordering the southern Sahara in which the fossils were found.
Generic description. Cranium (probably male) with an orthog-
nathic face showing weak subnasal prognathism, a small ape-size
braincase, a long and narrow basicranium, and characterized by the
following morphology: the upper part of the face wide relative to a
mediolaterally narrow and anteroposteriorly short lower face; a
large canine fossa; a small and narrow U-shaped dental arch; orbits
separated by a very wide interorbital pillar and crowned with a large,
thick and continuous supraorbital torus; a flat frontal squama with
no supratoral sulcus but with a marked postorbital constriction; a
small, posteriorly located sagittal crest and a large nuchal crest (at
least, in presumed males); a flat and relatively long nuchal plane
with a large external occipital crest; a large mastoid process; small
occipital condyles; a short, anteriorly narrow basioccipital; the long
axis of the petrous temporal bone oriented roughly 308 relative to
articles
NATURE | VOL 418 | 11 JULY 2002 | www.nature.com/nature 145
© 2002
Nature
Publishing
Group

the sagittal plane; the biporion line touching the basion; a round
external auditory porus; a broad glenoid cavity with a large post-
glenoid process; a robust and superoinferiorly short mandibular
corpus associated with a wide extramolar sulcus; a large, anteriorly
opening mental foramen centred beneath lower teeth P
4
–M
1
, below
midcorpus height; relatively small incisors; distinct marginal ridges
and multiple tubercles on the lingual fossa of upper I
1
;small
(presumed male) upper canines longer mesiodistally than bucco-
lingually; upper and lower canines with extensive apical wear; no
lower c–P
3
diastema; upper and lower premolars with two roots;
molars with low rounded cusps and bulbous lingual faces, M
3
triangular and M
3
rounded distally; enamel thickness of cheek
teeth intermediate between Pan and Australopithecus.
Differential diagnosis. Sahelanthropus is distinct from all living
great apes in the following respects: relatively smaller canines with
apical wear, the lower showing a full occlusion above the well-
developed distal tubercle, probably correlated with a non-honing
C–P
3
complex (P
3
still unknown).
Sahelanthropus is distinguished as a hominid from large living
and known fossil hominoid genera in the following respects: from
Pongo by a non-concave lateral facial profile, a wider interorbital
pillar, superoinferiorly short subnasal height, an anteroposteriorly
short face, robust supraorbital morphology, and many dental
characters (described below); from Gorilla by smaller size, a nar-
rower and less prognathic lower face, no supratoral sulcus, and
smaller canines and lower-cusped cheek teeth; from Pan by an
anteroposteriorly shorter face, a thicker and more continuous
supraorbital torus with no supratoral sulcus, a relatively longer
braincase and narrower basicranium with a flat nuchal plane and a
large external occipital crest, and cheek teeth with thicker enamel;
from Samburupithecus
14
by a more anteriorly and higher-placed
zygomatic process of the maxilla, smaller cheek teeth with lower
cusps and without lingual cingula, and smaller upper premolars and
M
3
; from Ouranopithecus
15
by smaller size, a superoinferiorly,
anteroposteriorly and mediolaterally shorter face, relatively thicker
continuous supraorbital torus, markedly smaller but mesiodistally
longer canines, apical wear and large distal tubercle in lower canines,
and thinner postcanine enamel; from Sivapithecus
16
by a supero-
inferiorly and anteroposteriorly shorter face with non-concave
lateral profile, a wider interorbital pillar, smaller canines with apical
wear, and thinner cheek-teeth enamel; from Dryopithecus
17
by a less
prognathic lower face with a wider interorbital pillar, larger
supraorbital torus, and thicker postcanine enamel.
Sahelanthropus is also distinct from all known hominid genera in
the following respects: from Homo by a small endocranial capacity
(preliminary estimated range 320–380 cm
3
) associated with a long
flat nuchal plane, a longer truncated triangle-shaped basioccipital, a
flat frontal squama behind a robust continuous and undivided
supraorbital torus, a large central upper incisor, and non-incisiform
canines; from Paranthropus
18
by a convex facial profile that is less
mediolaterally wide with a much smaller malar region, no frontal
trigone, the frontal squama with no hollow posterior to glabella, a
smaller, longer and narrower braincase, the zygomatic process of the
maxilla positioned more posterior relative to the tooth row, and
markedly smaller cheek teeth; from Australopithecus
19–21
by a less
prognathic lower face (nasospinale–prosthion length shorter at least
in presumed males) with a smaller malar (infraorbital) region and a
larger, more continuous supraorbital torus, a relatively more
elongate braincase, a relatively long, flat nuchal plane with a large
external occipital crest, non-incisiform and mesiodistally long
canines, and thinner cheek-teeth enamel; from Kenyanthropus
4
by
a narrower, more convex face, and a narrower braincase with more
marked postorbital constriction and a larger nuchal crest; from
Ardipithecus
6,7
by upper I
1
with distinctive lingual topography
characterized by extensive development of the crests and cingulum;
less incisiform upper canines not diamond shaped with a low distal
shoulder and a mesiodistal long axis, bucco-lingually narrower
lower canines with stronger distal tubercle, and P
4
with two roots;
from Orrorin
8
by upper I
1
with multiple tubercles on the lingual
fossa, and non-chimp-like upper canines with extensive apical wear.
Type species. Sahelanthropus tchadensis sp. nov.
Etymology. In recognition that all specimens were recovered in Chad.
Holotype. TM 266-01-060-1, a nearly complete cranium with the
following: on the right
—
I
2
alveolus, C (distal part), P
3
–P
4
roots,
fragmentary M
1
and M
2
,M
3
; and on the left
—
I
2
alveolus, C–P
4
roots, fragmentary M
1
–M
3
(Fig. 1 and Tables 1–5). Found by D.A.
on 19 July 2001.
After study, the holotype and paratype series will be housed in the
De
´
partement de Conservation des Collections, Centre National
d’Appui a
`
la Recherche (CNAR) in Ndjame
´
na, Chad. The holotype
has been dubbed ‘Toumaı
¨
’; in the Goran language spoken in the
Djurab Desert, this name is given to babies born just before the dry
season, and means ‘hope of life’.
Paratypes. See Table 1 for a list of paratypes, and Fig. 2 for
illustrations.
Locality. Toros-Menalla locality TM 266, a single quarry of about
5,000 m
2
,168 14
0
30
00
–168 15
0
30
00
N, 178 28
0
30
00
–178 30
0
00
00
E (wes-
tern Djurab Desert, northern Chad).
Horizon. All hominid specimens were found in the Toros-Menalla
anthracotheriid unit (AU) and come from a perilacustrine sand-
stone
12
. The associated fauna is biochronologically
12
older than
fossils from Lukeino, Kenya (,6 Myr; refs 8, 9), and more closely
resembles material from the Nawata formation at Lothagam, Kenya,
which is radio-isotopically dated to 5.2–7.4 Myr (ref. 13). Biochro-
nological studies are still underway and TM 266 fauna can be
tentatively dated between 6 and 7 Myr (ref. 12).
Diagnosis. Same as for genus.
Preservation
All specimens are relatively well preserved, but almost the entire
Table 1 Specimens of Sahelanthropus tchadensis gen. et sp. nov.
Specimen number Collected Element Discoverer Dental dimensions (mm)
...................................................................................................................................................................................................................................................................................................................................................................
TM 266-01-060-1 (holotype) (Fig. 1) 2001 Cranium D.A. RC, BL ¼ 10.2;
RM
1
,MD¼ 10.9;
RM
2
,MD¼ 13.0, BL ¼ (12.8);
RM
3
,MD¼ 10.8, BL ¼ 14.9;
LM
1
,MD¼ 11.5
TM 266-01-060-2 2001 Symphyseal fragment with I and C alveoli Group
TM 266-01-447 2001 Right M
3
Group RM
3
,MD¼ 10.7, BL ¼ 12.7
TM 266-01-448 (Fig. 2a) 2001 Right I
1
Group RI
1
,MD¼ (13.3), BL ¼ 8.9
TM 266-02-154-1 (Fig. 2b, c) 2002 Right mandible, (P
3
)P
4
–M
3
D.A. RP
4
,MD¼ 8.0;
RM
1
,MD¼ 11.0, BL ¼ 11.9;
RM
2
,MD¼ 12.5;
RM
3
,MD¼ 13.3, BL ¼ 12.2
TM 266-02-154-2 (Fig. 2d, e) 2002 Right c D.A. Rc, MD ¼ 11.0, BL ¼ 8.5
...................................................................................................................................................................................................................................................................................................................................................................
Fossil hominids recovered from Toros-Menalla between July 2001 and February 2002. BL, buccolingual; L, left; MD, mesiodistal; R, right. Parentheses indicate estimated measurements
—
(P
3
) indicates that
only the roots are known.
articles
NATURE | VOL 418 | 11 JULY 2002 | www.nature.com/nature146
© 2002
Nature
Publishing
Group

cranium has been flattened dorsoventrally and the entire right side
is depressed. The cranium exhibits broken but undistorted bone
units and matrix-filled cracks (Fig. 1). Estimated measurements are
given from a preliminary three-dimensional reconstruction. It will
soon be possible to use computer tomography (CT) scans to
generate a definitive three-dimensional reconstruction and stereo-
lithographic casts.
Comparative observations
Cranial comparisons of comparably aged hominids are limited to a
fragmentary basicranium and temporal bone: ARA-VP1/500 and
ARA-VP1/125 (ref. 6) from 4.4 Myr Ardipithecus ramidus ramidus.
The next-oldest maxillary and cranial specimens are younger:
KNM-KP-29283 (A. anamensis
5,20
), KNM-WT40000 (Kenya-
nthropus platyops
4
), and A. afarensis specimens, which include the
well-preserved AL 444-2 (ref. 21). The most notable anatomical
features of the S. tchadensis cranium for comparative purposes are
to be found in the face, which shows a mosaic of primitive and
derived features. The face (Fig. 1a) is tall with a massive brow ridge,
yet the mid-face is short (in the superoinferior dimension) and less
prognathic than in Pan or Australopithecus (Table 3). This unusual
combination of features is in turn associated with a relatively long
braincase, comparable in size to those of extant apes (Fig. 1b, c).
Preliminary comparisons with Pan suggest an endocranial volume
Figure 1 Cranium of Sahelanthropus tchadensis gen. et sp. nov. holotype (TM 266-01-060-1). a, Facial view. b, Lateral view. c, Dorsal view. d, Basal view.
articles
NATURE | VOL 418 | 11 JULY 2002 | www.nature.com/nature 147
© 2002
Nature
Publishing
Group

of 320–380 cm
3
.
Although the Sahelanthropus cranium is considerably smaller
than that of a modern male Gorilla, its supraorbital torus is
relatively and absolutely thicker. This is probably a sexually
dimorphic character (see Fig. 3), presumably reflecting strong
sexual selection. If this is a male, then the combination of a massive
brow ridge with small canines suggests that canine size was probably
not strongly sexually dimorphic. The interorbital pillar is wide
(Fig. 1a). The zygomatic process of the maxilla emerges above the
mesial margin of M
1
and is therefore more posterior relative to the
cheek teeth than in Australopithecus
19–21
and Paranthropus
18
. The
infraorbital plane (from the lower orbital margin to the inferior
Table 2 Comparative dental measurements
Mesiodistal Buccolingual
n Min. Max. Mean s.d. n Min. Max. Mean s.d.
...................................................................................................................................................................................................................................................................................................................................................................
Upper dentition
I
1
S. tchadensis 1 – – (13.3) – 1 – – 8.9 –
O. tugenensis
8
1 – – (10.0) – 1 – – 8.7 –
A. r. ramidus
6
1 – – (10.0) – 2 7.5 8.2 – –
A. anamensis
20
3 10.5 12.4 11.3 1.0 3 8.2 9.3 8.8 0.6
A. afarensis
6
3 10.8 11.8 11.2 0.6 5 7.1 8.6 8.2 0.6
P. t. troglodytes
26
14 – – 12.2 0.8 14 – – 9.4 0.8
C S. tchadensis – – – – – 1 – – 10.2 –
O. tugenensis
8
1 – – 11.0 – 1 – – 9.3 –
A. r. ramidus
6
2 (11.2) 11.5 – – 2 11.1 11.7 – –
A. anamensis
20
2 (10.6) 11.7 – – 2 10.2 11.2 – –
A. afarensis
6
9 8.9 11.6 10.0 0.8 10 9.3 12.5 10.9 1.1
P. t. troglodytes
26
15 – – 15.6 2.1 15 – – 11.3 1.37
M
1
S. tchadensis 2 (10.9) (11.5) – – – – – – –
A. r. kadabba
7
1 – – (10.6) – 1 – – 12.1 –
A. anamensis
20
7 10.3 12.9 11.7 0.8 6 11.7 14.1 13.0 0.9
A. afarensis
25
14 10.5 13.8 12.2 1.0 12 12.0 15.0 13.4 0.9
P. t. troglodytes
26
14 – – 10.5 0.5 14 – – 11.3 0.6
M
2
S. tchadensis 1 – – 13.0 – 1 – – (12.8) –
A. r. ramidus
6
2 (11.8) 11.8 – – 2 (14.1) (15.0) – –
A. anamensis
20
6 10.9 14.2 12.5 1.2 6 13.2 16.3 14.8 1.0
A. afarensis
6
5 12.1 13.5 12.8 0.5 6 13.4 15.1 14.7 0.6
P. t. troglodytes
26
16 – – 10.7 0.6 16 – – 11.7 0.8
M
3
S. tchadensis 2 10.7 10.8 – – 2 12.7 14.9 – –
O. tugenensis
8
2 10.2 10.3 – – 2 12.9 13.1 – –
A. r. kadabba
7
1 – – 10.9 – 1 – – 12.2 –
A. r. ramidus
6
1 – – 10.2 – 1 – – 12.3 –
A. anamensis
20
7 11.1 15.7 12.4 1.6 5 13.0 14.7 13.8 0.6
A. afarensis
6
8 10.5 14.3 11.9 1.4 8 13.0 15.5 13.8 1.0
P. t. troglodytes
26
16 – – 9.9 0.6 16 – – 10.8 1.0
Lower dentition
c S. tchadensis 1 – – 11.0 – 1 – – 8.5 –
A. r. kadabba
7
2 10.8 11.2 – – 2 7.8 7.8 – –
A. r. ramidus
6
– – – – – 1 – – 11.0 –
A. anamensis
20
7 6.6 10.4 9.0 1.3 6 9.2 11.4 10.2 1.0
A. afarensis
27
11 7.5 11.7 8.9 1.2 13 8.8 12.4 10.4 1.1
P. t. troglodytes
26
15 – – 14.0 1.5 15 – – 11.4 1.4
P
4
S. tchadensis 1– – 8.0––– – ––
O. tugenensis
8
1 – – (8.0) – 1 – – (9.0) –
A. r. kadabba
7
1 – – (8.1) – 1 – – 10.0 –
A. r. ramidus
6
2 7.5 8.9 – – 2 (9.9) (11.5) – –
A. anamensis
20
8 7.4 9.8 8.8 1.0 9 9.6 11.9 10.7 0.8
A. afarensis
6
15 7.7 11.1 9.7 1.0 14 9.8 12.8 10.9 0.8
P. t. troglodytes
26
15 – – 8.1 0.6 16 – – 8.8 0.8
M
1
S. tchadensis 1 – – 11.0 – 1 – – 11.9 –
A. r. ramidus
6
2 11.0 11.1 – – 2 (10.2) 10.3 – –
A. anamensis
20
11 11.5 13.8 12.6 0.9 12 10.5 14.8 12.1 1.3
A. afarensis
6
17 11.2 14.0 13.0 0.6 16 11.0 13.9 12.6 0.8
P. t. troglodytes
26
15 – – 10.7 0.4 15 – – 9.2 0.6
M
2
S. tchadensis 1 – – 12.5 – – – – – –
O. tugenensis
8
1 – – (11.5) – 1 – – (11.8) –
A. r. kadabba
7
1 – – (12.7) – 1 – – 11.8 –
A. r. ramidus
6
1 – – (13.0) – 1 – – 11.9 –
A. anamensis
20
8 13.0 15.9 14.1 1.4 11 12.3 15.1 13.5 0.9
A. afarensis
6
23 12.4 16.2 14.3 1.0 22 12.1 15.2 13.5 0.9
P. t. troglodytes
26
15 – – 11.3 0.5 15 – – 10.6 0.9
M
3
S. tchadensis 1 – – 13.3 – 1 – – 12.2 –
O. tugenensis
8
2 (12.3) (12.4) – – 2 10.4 11.2 – –
A. r. kadabba
7
1 – – 13.3 – – – – – –
A. r. ramidus
6
1 – – 12.7 – 1 – – 11.0 –
A. anamensis
20
6 13.7 17.0 14.6 1.2 6 11.9 13.4 12.8 0.7
A. afarensis
6
14 13.7 16.3 14.8 0.8 14 12.1 14.9 13.3 0.8
P. t. troglodytes
26
16 – – 11.0 0.6 16 – – 9.6 0.8
...................................................................................................................................................................................................................................................................................................................................................................
Parentheses indicate estimated measurements.
articles
NATURE | VOL 418 | 11 JULY 2002 | www.nature.com/nature148
© 2002
Nature
Publishing
Group

malar margin) is similar to that of Pan and differs from both Gorilla
and larger Australopithecus, in which this region is absolutely and
relatively taller. The canine fossa is similar to that in AL 444-2 (ref.
21) and there is no diastema between the alveoli of I
2
and C in the
Chadian specimen. There is no supratoral sulcus. Between the
temporal lines, the frontal squama is slightly depressed but not
like the frontal trigone usually seen in P. boisei
18
. The temporal lines
converge behind the coronal suture as in crested Pan, whereas their
junction is more anterior than generally seen in some large A.
afarensis (for example, AL 444-2; ref. 21). The sagittal crest is a little
larger posteriorly than in either AL 444-2 (ref. 21) or KNM-WT
40000 (ref. 4). The compound temporal–nuchal crest is marked as
in KNM-ER 1805 (ref. 22) and much larger than in the male A.
afarensis AL 444-2 (ref. 21), suggesting a relatively large posterior
temporalis muscle. The postorbital breadth (Fig. 1c) is absolutely
smaller than the Pan–Gorilla average, and similar to that in AL 444-
2 and AL 417-1 (ref. 21) (Table 5).
The basicranium (Fig. 1d) has small occipital condyles associated
with an apparently large foramen magnum. Despite damage, the
foramen magnum seems to be longer than wide, and not like the
rounded shape typical of Pan.AsinA. ramidus, the basion is
intersected by the bicarotid chord; the basion is posterior in large
apes and anterior in some of the later hominids. The Sahelanthropus
basioccipital is correlatively short and shaped like a truncated
triangle as in Ardipithecus
6
; it is not as triangular as in other early
hominids.
Like Pan, Australopithecus
21
and Ardipithecus
6
, the orientation of
the petrous portion of the temporal is approximately 608 relative to
the bicarotid chord, instead of 458 as is typical of Paranthropus and
Homo. Unlike Gorilla, Pan and Ardipithecus
6
, the posterior margin
of the tympanic tube does not have a crest-like morphology.
Compared with known Ardipithecus (ARA-VP-1/500)
6
, the glenoid
cavity is larger and the postglenoid process mediolaterally wider, so
the pronounced squamotympanic fissure is situated more medially
to the postglenoid process. The height of the relatively flat temporo-
mandibular joint above the tooth row suggests a high ascending
Table 4 Cranio-facial measurements of S. tchadensis
Cranio-facial features Size (mm)
.............................................................................................................................................................................
Orbital height £ width (36 £ 35)
Maximum breadth
Bicarotid chord (45)
Bimastoid (at the nuchal crest level) (108)
Biporion (102)
Mastoid mesiodistal length 54.5
Nuchal plane length (opisthion–inion) (47)
.............................................................................................................................................................................
Measurements are from TM 266-01-060-1. Parentheses indicate estimated measurements from a
preliminary three-dimensional reconstruction.
Table 3 Hominid alveolar height measurements
Taxon Alveolar height (mm)
.............................................................................................................................................................................
S. tchadensis (TM 266-01-060-1) (22)
A. afarensis
21
(AL 417-1, AL 444-2) 30, 33
A. africanus
28
(mean and range, n ¼ 11) 25.7 (21.1–30.0)
P. boisei
28
42.2
H. habilis
28
(KNM-ER 1470, KNM-ER 1813) 31.0, 25.0
.............................................................................................................................................................................
Alveolar height from nasospinale to prosthion. Parentheses indicate estimated measurements from
a preliminary three-dimensional reconstruction.
Figure 2 Sahelanthropus tchadensis gen. et sp. nov. paratypes. a, Right upper I
1
, lingual
view (TM 266-01-448). b, c, Right lower jaw (TM 266-02-154-1), occlusal view (b) and
axial CT scan (c). d, e, Right lower canine (TM 266-02-154-2), distal view (d) and buccal
view (e).
articles
NATURE | VOL 418 | 11 JULY 2002 | www.nature.com/nature 149
© 2002
Nature
Publishing
Group

ramus of the mandible more similar in absolute size to Gorilla than
to Pan. The large pneumatized mastoid process is anteroposteriorly
long, more so than in Ardipithecus
6
and Australopithecus (AL 444-2,
AL 333)
19,21
. The flat nuchal plane is relatively longer (Fig. 1b, d and
Table 4) than in Pan, Gorilla, AL 444-2 and AL 333, and with crests
as marked as those of Gorilla, implying the presence of relatively
large superficial neck muscles. The horizontally oriented nuchal
plane is much flatter than the convex nuchal plane of Pan. There is
not yet sufficient information to infer reliably whether Sahelan-
thropus was a habitual biped. However, such an inference would not
be unreasonable given the skull’s other basicranial and facial
similarities to later fossil hominids that were clearly bipedal.
Little is preserved of the mandibular corpus in TM-266-02-154-1
or the symphysis in TM-266-01-060-2 to permit much discussion of
mandibular anatomy. However, the former, presumably a male, has
a thick corpus with a wide extramolar sulcus (Fig. 2a).
In dentition, both lower incisor roots are small. The lingual
surface of upper I
1
displays distinct marginal ridges converging
basally onto a narrow gingival eminence and with several tubercles
extending incisally into the lingual fossa (Fig. 2a). The smaller upper
I
2
alveolus is situated, in frontal view, lateral to the lateral edge of the
nasal aperture (Fig. 1a), a probable symplesiomorphy with Pan. The
upper and lower canines are small (Tables 1, 2). Given the absolutely
and relatively massive supraorbital torus of the cranium (Figs 1a
and 3), possibly reflecting strong sexual selection, and the thick
corpus of the mandible (Fig. 2b, c), which probably indicate male
status, we infer that Sahelanthropus canines were probably weakly
sexually dimorphic. The upper canine (Fig. 1d) has both distal and
apical wear facets whereas M
3
is unworn. The lower canine (Fig. 2d,
e) has a strong distal tubercle that is separated from a distolingual
crest by a fovea-like groove; the large apical wear zone at a level
above this distal tubercle implies a non-honing C–P
3
complex (the
P
3
is still unknown). The upper canine, judging from the steep,
narrow distal wear strip reaching basally, we believe had a somewhat
lower distal shoulder than Ardipithecus
6,7
, suggesting an earlier
evolutionary stage. Moreover, a small, elliptical contact facet for
P
3
on the distobuccal face of the distal tubercle indicates the absence
of a lower c–P
3
diastema. Sahelanthropus thus probably represents
an early stage in the evolution of the non-honing C–P
3
complex
characteristic of the later hominids
7
.
The cheek teeth are small (Tables 1, 2), within the size range of A.
ramidus and the lower end of A. afarensis. Lower c and the lower and
upper premolars each have three pulp canals and two roots (Fig. 2c).
The P
3
has a 2R:MB þ D pattern (terminology following ref. 22: R,
root; M, mesial; MB, mesiobuccal; D, distal) with a small mesio-
buccal root and two partially fused, larger oblique distal roots. The
P
4
has two transverse roots (2R:M þ D), with the mesial one
smaller than the distal; in both premolars the distal root has two
pulp canals (Fig. 2c). The P
4
teeth of Ardipithecus have more-
derived root patterns, with either a Tome’s root (A. r. kadabba)
7
or a
single root (A. r. ramidus)
6
. S. tchadensis has a P
4
with a large talonid
and well-marked grooves on the buccal face. The molar size gradient
is M2 . M3 . M1. The upper molars have three roots, two buccal
and one lingual, which is large and mesiodistally elongated. The M
3
crowns are triangular whereas the M
3
teeth are rectangular and
rounded distally. The lower molar root pattern is 2R:M þ D: a
mesial root with two pulp canals and a distal root with only one
canal (Fig. 2c).
The radial enamel thickness is 1.71 mm at paracone LM
3
and
1.79 mm at hypocone LM
2
. The molar enamel is therefore thicker
than in Pan, possibly thicker than in Ardipithecus ramidus ramidus,
but thinner than in Australopithecus
6
. Given individual and intra-
specific variation, further study using high-resolution imaging (CT
scanning) is needed for useful comparisons
7
.
Figure 3 Vertical (superoinferior) thickness of the supraorbital torus in extant hominoids,
A. africanus and S. tchadensis gen. et sp. nov. Measurements are in millimetres at
the highest point of the superior orbital margin. Data for apes and A. africanus are from
ref. 28.
Table 5 Measurements and indices of hominoid frontal bones
Post orbital breadth* (a) (mm) Superior facial breadth† (b) (mm) Fronto-facial breadth index (a/b £ 100)
...................................................................................................................................................................................................................................................................................................................................................................
P. troglodytes (10 males)
19
71.0 (62.0–76.5) 110.1 (99.4–129.0) 64.4 (58.1–68.5)
P. troglodytes (10 females)
19
70.1 (65.0–76.0) 102.2 (87.3–112.0) 68.8 (66.6–74.5)
G. gorilla (10 males)
19
69.5 (60.5–77.0) 135.1 (127.7–144.0) 51.5 (47.4–59.4)
G. gorilla (10 females)
19
68.5 (65.5–73.5) 115.5 (108.0–127.0) 59.5 (57.0–63.4)
P. pygmaeus (5 males)
19
63.8 (59.0–69.5) 105.7 (88.3–116.2) 61.2 (51.7–72.5)
P. pygmaeus (4 females)
19
64.0 (61.5–65.5) 90.1 (87.8–94.0) 71.1 (68.1–74.0)
S. tchadensis gen. et sp. nov. (TM 266-01-060-1) (60) (102) (59)
A. afarensis (AL 444-2)
21
77.0 118.6 64.9
...................................................................................................................................................................................................................................................................................................................................................................
Values are mean and range. Values for S. tchadensis were estimated from a preliminary three-dimensional reconstruction.
*The least-frontal breadth.
†Left frontomalare temporale to right frontomalare temporale (outer biorbital breadth).
articles
NATURE | VOL 418 | 11 JULY 2002 | www.nature.com/nature150
© 2002
Nature
Publishing
Group

Discussion
Sahelanthropus has several derived hominid features, including
small, apically worn canines
—
which indicate a probable non-hon-
ing C–P
3
complex
—
and intermediate postcanine enamel thickness.
Several aspects of the basicranium (length, horizontal orientation,
anterior position of the foramen magnum) and face (markedly
reduced subnasal prognathism with no canine diastema, large
continuous supraorbital torus) are similar to later hominids includ-
ing Kenyanthropus and Homo. All these anatomical features indicate
that Sahelanthropus belongs to the hominid clade.
In many other respects, however, Sahelanthropus exhibits a suite
of primitive features including small brain size, a truncated tri-
angular basioccipital bone, and the petrous portion of the temporal
bone oriented 608 to the bicarotid chord. The observed mosaic of
primitive and derived characters evident in Sahelanthropus indi-
cates its phylogenetic position as a hominid close to the last
common ancestor of humans and chimpanzees. Given the biochro-
nological age of Sahelanthropus, the divergence of the chimpanzee
and human lineages must have occurred before 6 Myr, which is
earlier than suggested by some authors
23,24
. It is not yet possible to
discern phylogenetic relationships between Sahelanthropus and
Upper Miocene hominoids outside the hominid clade. Ourano-
pithecus
15
(about 2 Myr older) is substantially larger, with quadrate
orbits, a very prognathic and wide lower face, large male canines
with a long buccolingual axis, and cheek teeth with very thick
enamel. Samburupithecus
14
(about 2.5 Myr older) has a low, poster-
iorly positioned (above M
2
) zygomatic process of the maxilla, cheek
teeth with high cusps (similar to Gorilla), lingual cingula, large
premolars and a large M
3
.
Sahelanthropus is the oldest and most primitive known member
of the hominid clade, close to the divergence of hominids and
chimpanzees. Further analysis will be necessary to make reliable
inferences about the phylogenetic position of Sahelanthropus rela-
tive to known hominids. One possibility is that Sahelanthropus is a
sister group of more recent hominids, including Ardipithecus.For
the moment, productive comparisons of Sahelanthropus with
Orrorin are difficult because described craniodental material of
the latter is fragmentary and no Sahelanthropus postcrania are
available. However, we note that in Orrorin, the upper canine
resembles that of a female chimpanzee. The discoveries of Sahelan-
thropus along with Ardipithecus
6,7
and Orrorin
8
indicate that early
hominids in the late Miocene were geographically more widespread
than previously thought.
Finally, we note that S. tchadensis, the most primitive hominid, is
from Chad, 2,500 km west of the East African Rift Valley. This
suggests that an exclusively East African origin of the hominid clade
is unlikely to be correct (contrary to ref. 8). It will never be possible
to know precisely where or when the first hominid species origi-
nated, but we do know that hominids had dispersed throughout the
Sahel and East Africa
10
by 6 Myr. The recent acquisitions of Late
Miocene hominid remains from three localities, as well as func-
tional, phylogenetic and palaeoenvironmental studies now under-
way, promise to illuminate the earliest chapter in human
evolutionary history. Sahelanthropus will be central in this endea-
vour, but more surprises can be expected. A
Received 13 March; accepted 27 May 2002; doi:10.1038/nature00879.
1. Dart, R. A. Australopithecus africanus, the man-ape of South Africa. Nature 115, 195–199 (1925).
2. Kortlandt, A. New Perspectives on Ape and Human Evolution 9–100 (Stichting voor Psychobiologie,
Amsterdam, 1972).
3. Coppens, Y. Le singe, l’Afrique et l’Homme (Fayard, Paris, 1983).
4. Leakey, M. G. et al. New hominin genus from eastern Africa shows diverse middle Pliocene lineages.
Nature 410, 433–440 (2001).
5. Leakey, M. G., Feibel, C. S., McDougall, I. & Walker, A. C. New four-million-year-old hominid species
from Kanapoi and Allia Bay, Kenya. Nature 376, 565–571 (1995).
6. White, T. D., Suwa, G. & Asfaw, B. Australopithecus ramidus, a new species of hominid from Aramis,
Ethiopia. Nature 371, 306–312 (1994).
7. Haile-Selassie, Y. Late Miocene hominids from the Middle Awash, Ethiopia. Nature 412, 178–181
(2001).
8. Senut, B. et al. First hominid from the Miocene (Lukeino Formation, Kenya). C. R. Acad. Sci. Paris
332, 137–144 (2001).
9. Deino, A. L., Tauxe, L., Monaghan, M. & Hill, A.
40
Ar/
39
Ar geochronology and paleomagnetic
stratigraphy of the Lukeino and lower Chemeron Formations at Tabarin and Kapcheberek, Tugen
Hills, Kenya. J. Hum. Evol. 42, 117–140 (2002).
10. Brunet, M. et al. The first australopithecine 2500 kilometres west of the Rift Valley (Chad). Nature 378,
273–275 (1995).
11. Brunet, M. et al. Australopithecus bahrelghazali, une nouvelle espe
`
ce d’Hominide
´
ancien de la re
´
gion
de Koro Toro (Tchad). C. R. Acad. Sci. Paris 322, 907–913 (1996).
12. Vignaud, P. et al. Geology and palaeontology of the Upper Miocene Toros-Menalla hominid locality,
Chad. Nature 418, 152–155 (2002).
13. McDougall, I. & Feibel, C. S. Numerical age control for the Miocene–Pliocene succession at Lothagam,
a hominoid-bearing sequence in the northern Kenya Rift. J. Geol. Soc. Lond. 156, 731–745 (1999).
14. Ishida, H. & Pickford, M. A new Late Miocene hominoid from Kenya: Samburupithecus kiptalami gen.
et sp. nov. C. R. Acad. Sci. Paris 325, 823–829 (1998).
15. Bonis, L. de & Koufos, G. D. The face and the mandible of Ouranopithecus macedoniensis: description
of new specimens and comparisons. J. Hum. Evol. 24, 469–491 (1993).
16. Pilbeam, D. New hominoid skull material from the Miocene of Pakistan. Nature 295, 232–234 (1982).
17. Kordos, L. & Begun, D. R. A new cranium of Dryopithecus from Rudabanya, Hungary. J. Hum. Evol. 41,
689–700 (2001).
18. Tobias, P. V. The Cranium and Maxillary Dentition of Australopithecus (Zinjanthropus) boisei, Olduvai
Gorge (Cambridge Univ. Press, London, 1967).
19. Kimbel, W. H., White, T. D. & Johanson, D. C. Cranial morphology of Australopithecus afarensis;A
comparative study based on a composite reconstruction of the adult skull. Am. J. Phys. Anthropol. 64,
337–388 (1984).
20. Ward, C. V., Leakey, M. G. & Walker, A. Morphology of Australopithecus anamensis from Kanapoi and
Allia Bay, Kenya. J. Hum. Evol. 41, 235–368 (2001).
21. Kimbel, W. H., Johanson, D. C. & Rak, Y. The first skull and other new discoveries of Australopithecus
afarensis at Hadar, Ethiopia. Nature 368, 449–451 (1994).
22. Wood, B. Koobi Fora Research Project: Hominid Cranial Remains Vol. 4 (Clarendon, Oxford, 1991).
23. Kumar, S. & Hedges, B. A molecular time scale for vertebrate evolution. Nature 392, 917–920 (1998).
24. Pilbeam, D. in The Primate Fossil Record (ed. Hartwig, W.) 303–310 (Columbia Univ. Press, New York,
2002).
25. Kimbel, W. H. Systematic assessment of a maxilla of Homo from Hadar, Ethiopia. Am. J. Phys. Anthop.
103, 235–262 (1997).
26. Uchida, A. Intra-species Variation among the Great-Apes: Implications for Taxonomy of Fossil
Hominoids Thesis, Harvard Univ. (1992).
27. Lockwood, C. A., Kimbel, W. H. & Johanson, D. C. Temporal trends and metric variation in the
mandibles and dentition of A. afarensis. J. Hum. Evol. 39, 23–55 (2000).
28. Lockwood, C. A. Sexual dimorphism in the face of Australopithecus africanus. Am. J. Phys. Anthropol.
108, 97–127 (1999).
Acknowledgements
We thank the Chadian Authorities (Ministe
`
re de l’Education Nationale de l’Enseignement
Supe
´
rieur et de la Recherche, Universite
´
de N’djame
´
na, CNAR). We extend gratitude for
their support to the French Ministries, Ministe
`
re Franc¸ais de l’Education Nationale
(Faculte
´
des Sciences, Universite
´
de Poitiers), Ministe
`
re de la Recherche (CNRS),
Ministe
`
re des Affaires Etrange
`
res (Direction de la Coope
´
ration Scientifique, Universitaire
et de Recherche, Paris, and SCAC Ambassade de France a
`
N’djame
´
na), to the Re
´
gion
Poitou-Charentes, the De
´
partement de la Vienne, the Association pour le Prix scientifique
Philip Morris, and also to the Arme
´
e Franc¸aise (MAM and Epervier) for logistic support.
For giving us the opportunity to work with their collections, we are grateful to the National
Museum of Ethiopia, the National Museum of Kenya, the Peabody Museum and Harvard
University, the Institute of Human Origins and the University of California. Special thanks
to Scanner-IRM Poitou Charentes (P. Chartier and F. Perrin), for industrial scanner to
EMPA (A. Flisch Ing. HTL) and to Multimedia Laboratorium-Computer Department,
University of Zurich-Irchel (P. Stucki). Many thanks to all our colleagues and friends for
their help and discussion, and particularly to F. Clark Howell, A. Garaudel, Y. Haile-
Selassie, D. Johanson, W. Kimbel, M. G. Leakey, D. Lieberman, R. Macchiarelli,
M. Pickford, B. Senut, G. Suwa, T. White and Lubaka. We especially thank all the other
MPFT members who joined us for the field missions, and V. Bellefet, S. Riffaut and J.-C.
Bertrand for technical support. We are most grateful to G. Florent for administrative
guidance. We dedicate this article to J. D. Clark. All authors are members of the MPFT.
Competing interests statement
The authors declare that they have no competing financial interests.
Correspondence and requests for materials should be addressed to M.B.
(e-mail: michel.brunet@univ-poitiers.fr).
articles
NATURE | VOL 418 | 11 JULY 2002 | www.nature.com/nature 151
© 2002
Nature
Publishing
Group