CT-based study of internal structure of the anterior
pillar in extinct hominins and its implications for
the phylogeny of robust Australopithecus
Brian A. Villmoarea,1and William H. Kimbelb
aDepartment of Anthropology, University College London, London WC1H 0BW, United Kingdom; andbInstitute of Human Origins, School of Human
Evolution and Social Change, Arizona State University, Tempe, AZ 85287
Edited by Erik Trinkaus, Washington University, St. Louis, MO, and approved August 15, 2011 (received for review April 12, 2011)
The phylogeny of the early African hominins has long been
confounded by contrasting interpretations of midfacial structure.
In particular, the anterior pillar, an externally prominent bony
column running vertically alongside the nasal aperture, has been
identified as a homology of South African species Australopithe-
cus africanus and Australopithecus robustus. If the anterior pillar is
a true synapomorphy of these two species, the evidence for
a southern African clade of Australopithecus would be strength-
ened, and support would be given to the phylogenetic hypothesis
of an independent origin for eastern and southern African “ro-
bust” australopith clades. Analyses of CT data, however, show
that the internal structure of the circumnasal region is strikingly
different in the two South African australopith species. In A. afri-
canus the anterior pillar is a hollow column of cortical bone,
whereas in A. robustus it is a column of dense trabecular bone.
Although Australopithecus boisei usually lacks an external pillar, it
has internal morphology identical to that seen in A. robustus. This
result supports the monophyly of the “robust” australopiths and
suggests that the external similarities seen in the South African
species are the result of parallel evolution.
extinct group of hominins that flourished in Africa between ca.
4.2 and 1.4 Mya (reviewed in ref. 1; for the purposes of this
paper, we treat the genera Paranthropus and Kenyanthropus as
synonymous with Australopithecus). Taxonomic variation in the
morphology of the midface, including the subnasal plate, nasal
aperture margins, and position of the maxillary zygomatic pro-
cess, as well as the expanded, morphologically derived post-
canine dental battery, have supported the inference of dietary
specialization in the so-called “robust” species, Australopithecus
boisei of eastern Africa and Australopithecus robustus of southern
Africa (2–4). The presence of “anterior pillars,” distinctive bony
columns bordering the nasal aperture in A. robustus, at least one
specimen of A. boisei, and Australopithecus africanus, the oldest
and otherwise most symplesiomorphic of the three, was thought
by Rak (3) to unite these species in a monophyletic group
characterized by progressive specialization of the masticatory
system. Rak (3) hypothesized that the anterior pillars structurally
buttressed the midface to counter high-magnitude loading of the
molarized premolars in these species. Results of a recent finite-
element analysis of a model A. africanus face have been inter-
preted as support for Rak’s functional hypothesis (5), although
this study has been challenged (6).
Implicit in the phylogenetic and functional evaluation of the
anterior pillar is that it is compositionally identical among the
taxa in which it appears. Here we report the results of a CT-
based study of the australopith midface and show that, on the
contrary, the internal composition of the anterior pillar varies
taxonomically among the australopiths. This finding raises ques-
tions about the homology and functional significance of a structure
that has been central to debates about early hominin phylogeny
acial anatomy has played a prominent role in the elucidation
of the systematics and adaptations of the australopiths, the
In one of his first surveys of australopith discoveries made after
the Taung child, Robert Broom (7) noted the “curious bony ridge
[that] runs down from the inner border of the large infraorbital
foramen” on the face of the type specimen (TM 1517) of A. ro-
bustus from Kromdraai (ca. 1.5–2.0 Mya). Shortly thereafter,
Gregory and Hellman (8) drew attention to the similarity in
morphology of the area bordering the nasal aperture in TM 1517
and the A. africanus (ca. 2.5–3.0 Mya) maxilla TM 1512: “A
slightly defined ridge posterior to the elevation of the canine
socket extends upward, ending just below the infraorbital; this
seems to be homologous with the sharp ridge that bounds the
flattened lower nasal plate in the type of A. robustus.” Broom and
Robinson (9) characterized the facial plate lateral to the nasal
cavity in A. africanus (Sts 5) as a “rounded angle which runs up
towards the side of the nostril and divides the premaxillary an-
terior plane from the lateral part of the maxilla,” and on A.
robustus specimen SK 12 they observed “two slightly raised bony
ridges, which pass downwards and slightly outwards to below the
level of the infraorbital foramina and then almost straight down
over the region of the canine roots. . These ridges form the
lateral walls of the nasal opening” (10).
Rak’s (3) formal designation of this morphology as the ante-
rior pillar underscored the restricted distribution of the character
among the australopiths. As with the extant great apes, Aus-
tralopithecus afarensis (ca. 3.7–3.0 Mya) does not display this
structure; here, the large maxillary canine root plays the major
role in shaping the facial plate alongside the nasal aperture (11).
The pillar also is absent from the Australopithecus aethiopicus
cranium KNM-WT 17000 (ca. 2.5 Mya) (12) and the Austral-
opithecus garhi type specimen ARA-VP 12/130 (ca. 2.5 Mya)
(13), which, given the strongly divergent maxillary anatomies of
these two specimens, would seem to reinforce the phylogeneti-
cally derived status of the pillar’s presence in A. africanus and A.
robustus. Although most A. boisei crania lack the anterior pillar,
Rak (3) asserted that this species actually was situated at the
most derived end of the feature’s morphocline of expression. The
basis for this conclusion was Rak’s linking taxonomically con-
gruent morphoclines in anterior pillar expression, maxillary
prognathism, and maxillary zygomatic position in a functional
complex designed to resist deformation of the variably prog-
nathic snout in the face of high-magnitude occlusal loads im-
posed on the relatively large, molarized premolars. For A. boisei,
the extreme anterior position of the zygomatics obviated the
need for anterior pillars, because the peripheral face itself served
Author contributions: B.A.V. and W.H.K. designed research; B.A.V. and W.H.K. performed
research; B.A.V. and W.H.K. analyzed data; and B.A.V. and W.H.K. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1To whom correspondence should be addressed. E-mail: email@example.com.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
| September 27, 2011
| vol. 108
| no. 39www.pnas.org/cgi/doi/10.1073/pnas.1105844108
as the buttress against loading of the extraordinarily large,
The absence of the pillar in the KNM-WT 17000 cranium of A.
aethiopicus is important, because this middle Pliocene species
usually is interpreted as cladistically basal to the later “robust”
australopiths and thus critical to the phylogenetic reconstruction
of this group. Different investigators have interpreted the ab-
sence of pillars in A. aethiopicus and A. boisei differently,
depending on whether they accept Rak’s (3) functional model.
For example, Skelton and McHenry (14) considered pillars in
these two taxa to be present (but “obscured by infilling”) and
thus derived a priori, whereas Strait et al. (15) identified them as
being absent. The phylogenetic consequences of these kinds of
distinctions are not trivial. A “robust” clade comprising A. ae-
thiopicus, A. robustus, and A. boisei (as in refs. 11 and 16) would
imply that the development of the anterior pillars in A. robustus
and A. africanus is homoplastic, whereas McCollum’s (17) pro-
posal of separate eastern and southern African australopith
clades, essentially an argument for a polyphyletic origin of robust
australopiths, would concede that anterior pillars in A. africanus
and A. robustus are synapomorphies and that their absence in A.
aethiopicus and A. boisei is symplesiomorphic.
All the research on the australopith circumnasal region has
assumed that the anterior pillar conforms to Rak’s (3) de-
scription of a solid column of bone. Both phylogenetic and
functional conclusions about the pillar are susceptible to the
finding, as we report here, that the internal structure of the pillar
in fact varies among taxa that manifest it topographically on the
external aspect of the facial skeleton (Table 1).
Australopithecus afarensis. We examined CT scans of A.L. 444–2
(11). The fragmentary nature of other maxillae attributed to this
species allowed direct examination of the maxillary sinus walls
(“non-CT” in Table 1). Externally, on all these specimens, the
lateral margin of the nasal aperture is sharp. Laterally, the ex-
ternal contour of the midface reflects the bulging contour of the
canine jugum which, at a level approximating one-third the
height of the nasal aperture, forms part of the lateral nasal
margin. Internally, the area lateral to the nasal aperture is
formed by the hollowed maxillary sinus and the internal contour
of the canine root and socket. On A.L. 922–1, damage exposes
the canine root, broken just below the tip, immediately above the
inferolateral corner of the nasal aperture. A. afarensis is the only
hominin species examined for this study in which the canine root
extends above the inferior nasal margin (clearly visible in CT
scans of A.L. 444–2) to influence the form of the external facial
contour [although this condition is reported in A. anamensis as
well (18)]. The canine jugum strongly influences the form of the
facial contour in extant apes, and this morphology is likely the
ancestral condition for early hominins.
Australopithecus africanus. In many specimens, the anterior pillar
is the most prominent topographic feature of the A. africanus
midface, forming the sides of the “nasoalveolar triangular frame”
of Rak (3). The anterior pillar dulls the lateral margin of the
nasal aperture for all, or nearly all, of the aperture’s height. In
horizontal CT sections above the nasal cavity floor the external
aspect of the pillar is composed of a thin, curved plate. Internal
to the curved plate, the anterior pillar is hollow, rather than
solid, forming part of the anterior wall of maxillary sinus cavity.
This morphology can be seen clearly in a horizontal section of Sts
5 (Fig. 1) and Sts 71 (Fig. S1). The matrix infilling of the max-
illary sinus cavity in Sts 17 makes interpretation more challeng-
ing, but this specimen appears to match the pattern of Sts 5 and
Sts 71. In all cases the external contour surrounds the internal
hollow space, and in no case is the pillar formed by solid cortical
or cancellous bone (Figs. 2 and 3 and Fig. S2).
Maxillae Stw 73 and Stw 183 follow this pattern, but in these
specimens the alveoli for the canine roots reach further superi-
orly than in other specimens of A. africanus we examined for this
study, extending along the subnasal plate to terminate approxi-
mately at the level of the inferior nasal margin (see canine roots
of Sts 71 in Fig. S3). On other specimens the trabecular bone of
the alveolus does not extend superiorly as far as the nasal ap-
erture, whereas on these two specimens maxillary trabecular
bone extends slightly above the inferior nasal margin. It there-
fore appears that the presence of trabecular bone above the level
of the nasal cavity floor, protruding into the maxillary sinus, is in
A. africanus related to the height of the canine root (and so is not
similar to the condition found in A. robustus; see below).
Australopithecus robustus. In CT images of the anterior margin of
the nasal margin in a horizontal plane above the nasal floor, the
lateral margin of the nasal aperture can be seen readily in SK 12,
SK 13/14, SK 46, SK 48, SK 52, SK 79, SK 83, and SKW 11. As is
apparent in Fig. 1, on SK 12 the facial plate of the maxilla im-
mediately lateral to the nasal aperture is composed of a column of
trabecular bone, triangular in cross-section, immediately lateral to
which is the anterior compartment of the maxillary sinus. Even on
specimens where the maxillary sinus is filled with matrix (e.g., SK
48, SK 13/14, and SK 11) this trabecular column is clearly visible.
On specimens SK 12, SK 52, and SK 83 matrix does not fill the
maxillary sinus, and the CT scan provides more resolution. On all
these specimens, the triangle of bone is clear (Figs. S4 and S5).
Superiorly, the triangular column extends to about one-half
the height of the nasal aperture in SK 48 and SK 12 (Figs. 2 and
3). In other specimens breakage precludes determining the ver-
tical extent of the triangular column. Externally on SK 48 the
anterior pillar terminates midway up the nasal aperture, corre-
sponding to the superior extent of the trabecular column visible
in CT scans, but on SK 12 the anterior pillar is visible externally
running along the entire vertical extent of the nasal aperture.
Thus in A. robustus there is not a close relationship between the
external topographic manifestation of the anterior pillar and the
extent of the internal trabecular column. The column of trabec-
Australopithecus species and specimens examined for
A. africanus Sts 5
Stw 73 (cast)
Stw 183 (cast)
AL 200–1 (non-CT)
AL 417–1 (non-CT)
AL 922–1 (non-CT)
Villmoare and KimbelPNAS
| September 27, 2011
| vol. 108
| no. 39
ular bone forming the interior of the anterior pillar is contiguous
with the trabecular bone that makes up the canine alveolus and is
visible in a coronal section of SK 12 just posterolateral to the
nasal spine (Fig. 3).
Australopithecus boisei. As noted by Rak (3), externally there is no
anterior pillar in A. boisei (with the exception of KNM-ER 732).
On OH 5, there is no exterior column of bone along the nasal
margin topographically distinguishable from the infraorbital plate
immediately lateral to it. However, as observed in the horizontal
CT section above the nasal cavity floor (Fig. 1) of OH 5, the lateral
margin of the nasal aperture is formed by a thickened column of
cortical and trabecular bone that is very similar in form and lo-
cation to that seen in A. robustus. As in SK 12, the column extends
superiorly approximately half the height of the nasal aperture, as
seen in sagittal and coronal sections (Figs. 2 and 3).
5 mm above the nasospinale. The hollow anterior pillar is visible on both sides of the nasal margin in Sts 5 (arrow). In SK 12 and OH 5 a cross-section of the
column of trabecular bone is visible (arrows). See also Fig. S1.
The horizontal line in A indicates the alveolar plane. Sts 5 (B), SK 12 (C), and OH 5 (D) are shown in horizontal section parallel with the alveolar plane,
aperture in Sts 71 (B) and the trabecular column in SK 12 (C) and OH 5 (D). See Fig. S2 for Sts 5 and Fig. S5 for SK 52.
The vertical line in A shows the plane of the coronal section posterior to the nasospinale. Note the lack of trabeculae at inferior corner of nasal
| www.pnas.org/cgi/doi/10.1073/pnas.1105844108Villmoare and Kimbel
Specimens KNM-ER 406 and KNM-ER 732 are not as well
preserved as OH 5, and infilling matrix is not as distinct from the
fossilized bone in CT scans. In KNM-ER 406 the facial surface
below the orbits is eroded, so we cannot make any interpretation
of the external morphology, and in coronal CT sections the
distribution of trabecular bone cannot be identified in the eroded
region. However, in a horizontal section slightly below the level
of the nasal cavity floor, it is possible to identify the base of the
trabecular column (Fig. 4), which separates the maxillary sinus
cavity into anterior and posterior chambers in OH 5 and KNM-
ER 406, exactly as in A. robustus. Preservation around the nasal
aperture in KNM-ER 732 is poor also. However, the same pat-
tern seen in the maxillary sinus cavity of KNM-ER 406 is visible
in CT scans of the internal structure of the maxillary sinus of
KNM-ER 732, with trabecular bone separating the maxillary
sinus into two chambers. Because this trabecular formation is
clearly the base of the trabecular column bordering the nasal
aperture in OH 5 (and in A. robustus), we infer that KNM-ER
406 and KNM-ER 732 possessed the same internal bony column.
Australopithecus aethiopicus. The KNM-WT17000craniumsuffers
damage to the face, including breakage at the left margin of the
nasal aperture (12); however, it is possible to determine the
morphology in this region both by studying the CT scans and ex-
amining the interior directly through breaks in the bone of the
midface. Although much of the maxillary sinus cavity remains fil-
led with matrix, the CT scan clearly discriminates between the
fossilized bone and the matrix, so determining the internal mor-
phology is straightforward. There is no external manifestation of
an anterior pillar on the face of KNM-WT 17000 (12, 19), and it
lacks the broadly rounded lateral margins found low on the nasal
the interior via CT scans shows the region bordering the nasal
aperture is formed by the maxillary sinus antrum. There is no in-
ternal column of trabecular bone as seen in A. robustus and A.
boisei, and there is no separate hollow column as in A. africanus
(Fig. S6). This anatomy appears to be most similar to that of A.
afarensis, except that thecaninesarereduced, and thecanine roots
have less influence on the external and internal morphology.
Rak’s (3) recognition of the anterior pillar in South African
Australopithecus focused attention on the phylogenetic and
functional significance of the australopith midface. The anterior
pillar was incorporated into the body of evidence supporting a
phylogenetic hypothesis that linked A. africanus exclusively to the
“robust” australopith clade, with A. afarensis as the generalized
mid-Pliocene ancestor to both the “robust” and Homo clades
(20), and bolstered the explanatory scheme that linked these
species, via postcanine megadontia, premolar molarization, and
derived facial morphology, to an adaptation for specialized
feeding (21–23). In this scheme, the “absence” of pillars in A.
boisei actually was interpreted as a “presence,” although sub-
sequent demonstration of their absence in the more symplesio-
morphic cranium of A. aethiopicus created ambiguity regarding
at inferior corner of nasal aperture in Sts 71 (arrow) and the trabecular column in SK 12 and OH 5 (arrows). See Fig. S2 for Sts 5 and Fig. S5 for SK 52.
The vertical line in A shows the plane of the coronal section posterior to the nasospinale in Sts 71 (B), SK 12 (C), and OH 5 (D). Note lack of trabeculae
A. boisei (KNM-ER 406) (not to scale). Although KNM-ER 406 is infilled by
matrix, a circular sinus cavity (marked by arrow) is visible on right of the
specimen and also is visible in SK 12. Similar morphology is present on OH 5
and KNM-ER 732. This sinus is formed at the base of the trabecular column
bordering the nasal aperture in A. robustus and OH 5 and suggests that the
trabecular column is present in all specimens of A. boisei, even though
damage to Koobi Fora specimens makes identification of the column im-
possible above the nasal floor. Section is horizontal at the level of the
nasospinale (5 mm below plane shown in Fig. 1). The sinus also is visible
a few millimeters higher on the left side of SK 12.
Base of trabecular column in (Left) A. robustus (SK 12a) and (Right)
Villmoare and KimbelPNAS
| September 27, 2011
| vol. 108
| no. 39
the state of the structure in the east African ”robust” austral-
opiths (14, 19). Our examination of the internal structure of the
anterior pillar clarifies the role of this feature in phylogenetic
research on early hominins.
A pattern of internal circumnasal structure is shared uniquely
by A. robustus and A. boisei. In crania of both species the facial
plate lateral to the nasal aperture is composed of dense trabec-
ular bone. However, this underlying structural similarity belies
differences in external morphology, because the largest A. boisei
specimens (OH 5 and KNM-ER 406) lack a distinct pillar,
whereas A. robustus invariably has it. Rak’s (3) observation of
a rudimentary external pillar in the female A. boisei cranium
KNM-ER 732 now attains greater significance as a sign of
common descent: The “layer” of internal structural similarity in
the faces of these species is consistent with their derivation from
a taxon with the same internal morphology (contra ref. 17). A.
aethiopicus is symplesiomorphic in the internal structure of the
region surrounding the nasal aperture and in this respect is more
similar to A. afarensis than to A. robustus and A. boisei. Our CT-
based findings support the monophyly of the two late “robust”
australopiths, A. robustus and A. boisei, and highlight the mor-
phological gap, noted by others (12, 14, 19), between the cra-
niofacial anatomy of A. aethiopicus and that expected of the last
common ancestor of A. robustus and A. boisei.
Unexpectedly, the internal morphology of the anterior pillar in
A. africanus, the species that initially drew Rak’s (3) attention to
this feature, is singular among the australopiths. Despite the rel-
ative prominence of the external pillar in many crania of this
species, and in contrast to Rak’s (3) description of it as a solid
column of bone, CT reveals it to be a hollow tube, at least in the
specimens examined in this study, notably including Sts 5, which
has the most prominent external pillars of any A. africanus spec-
imen. This finding suggests that the anterior pillars of A. africanus
and the anterior pillars of A. robustus are not homologous, as was
implied already by phylogenetic hypotheses that position a species
with the facial structure of A. aethiopicus as the sister taxon to A.
robustus and A. boisei (e.g., 11, 16, 19) (Fig. 5).
Rak’s (3) functional hypothesis assumed that anterior pillars
comprise solid bone. Our finding that uniquely in A. africanus—
the speciesinwhichthey are bestdevelopedexternally—the pillars
are hollow tubes rather than solid columns means that disparate
functional causes could underlie the development of the struc-
results do not address directly the functional basis of the anterior
pillars, but recent debate about the functional interpretation of
finite element analysis of the model A. africanus midface (5, 6) will
need to incorporate questions of homology raised by the discovery
of diversity in the internal composition of the anterior pillar.
Patterns of similarity in midfacial morphology among extinct
Plio-Pleistocene hominins have drawn paleoanthropologists’ at-
tention since the 1930s. Rak’s (3) identification of the anterior
pillar in mid-Pliocene A. africanus constituted evidence of an
exclusive phylogenetic relationship between this species and the
late hypermegadont “robust” australopiths. His functional hy-
pothesis, linking the pillars to high-magnitude biting on the pre-
molars, attempted to account for the persistence of this structure
in A. robustus and its near-disappearance in A. boisei, species
with much more derived masticatory configurations than A.
africanus. Rak’s interpretation and most subsequent analyses of
the distribution (14–16, 24–26) and function (5, 6, 27) of the
anterior pillar have assumed that the structure is composed of
solid bone. Here we have shown that this assumption is false.
Our CT-based study demonstrates that the pillar in A. africanus
is hollow, whereas in A. robustus it comprises dense trabecular
bone; in A. boisei the comparable midfacial zone also is com-
posed of trabecular bone, even though usually there is no dis-
cernible trace of the anterior pillar externally. The primitive
hominin condition is exemplified by A. afarensis, in which the
presence of the long maxillary canine root adjacent to the nasal
aperture shapes the midface, as in extant great apes.
Our results indicate that the anterior pillar, a frequently cited
synapomorphy of the midface in A. africanus and A. robustus,
actually may be a homoplasy, which cedes support to hypotheses
of a monophyletic rather than a polyphyletic origin of the geo-
logically late “robust” australopith species (Fig. 5). The anterior
pillar in A. africanus is likely an autapomorphy, highlighting the
ambiguous phylogenetic position of this taxon, as noted by others
(11, 14–16). These findings also may complicate biomechanical
interpretations, because they raise suspicions that different
functional causes correspond to the different patterns of internal
bone distribution beneath the anterior pillar.
Materials and Methods
We examined the fossil specimens using direct observation and/or exami-
nation of CT data to examine the internal morphology of the nasal margin,
maxillary sinus, and maxilla (Table 1). CT data were collected from a variety
of sources and include all available CT scans of relevant specimens from the
species analyzed. For specimens SK 11, SK 12, SK 46, SK 79, SK 83, and
SKW11, scans (0.33-mm slices; voxel size 0.3222) were made by Chris Williams
at Moot Algemen Hospital, Pretoria, Republic of South Africa, on a Phillips
Brilliance 40 CT scanner. Specimens SK 13/14, SK 48, SK 52, Sts 17, and Sts 52
were scanned by Stephany Potze of the Transvaal Museum at Little Com-
pany of Mary Hospital in Pretoria on a Siemens Sensation 16 CT scanner
(0.75-mm slices; voxel sizes ranging from 0.2929–0.3476). Other scans (Sts 5,
Sts 71, OH 5, KNM-ER 406, KNM-ER 732, KNM-WT 17000, and A.L. 444–2) are
from the digital archives at the University of Vienna and the National
Museums of Kenya. We used ImageJ 1.42 (National Institutes of Health) and
Amira 4.1.2 (Mercury Computer Systems) to examine the CT data.
The small sample sizes of well-preserved taxa that have been CT scanned
limit the potential resolution of statistical analyses. To capture the internal
East African and South African clades; however, this model requires con-
vergences for the trabecular column in A. boiei and A. robustus. B shows
a phylogenetic model that is consistent with recent phylogenetic analyses
(11, 16), and does not require any homoplasy.
Two models for the evolution of the anterior pillar. A shows distinct
| www.pnas.org/cgi/doi/10.1073/pnas.1105844108 Villmoare and Kimbel
structure, we examined CT scans on coronal, transverse (horizontal), and Download full-text
sagittal planes at multiple intervals. We also describe anatomy that is directly
visible because of breakage in both original specimens (A.L. 200–1, A.L. 417–
1, A.L. 444–2, and A.L. 922–1) and casts (Stw 73 and Stw 183).
ACKNOWLEDGMENTS. We thank Kevin Kuykendal and Todd Rae for use of
CT data for specimens SK 11, SK 12, SK 46, SK 79, SK 83, and SKW11 and for
specimens SK 13/14, SK 48, SK 52, Sts 17, and Sts 52, respectively; Fred Spoor,
Jean-Jacques Hublin, Emma Mbua, and the Department of Earth Sciences,
National Museums of Kenya, for access to CT scans and facilities for analysis;
and Yoel Rak for discussion and comments on the manuscript. Funding for
scanning specimens SK 11, SK 12, SK 46, SK 79, SK 83, and SKW11 was provided
by the University of Sheffield, and funding for scanning specimens SK 13/14, SK
48, SK 52, Sts 17, and Sts 52 was provided by Roehampton University.
1. Kimbel WH (2007) The species and diversity of australopiths. Handbook of Paleoan-
thropology, eds Rothe H, Tattersall I, Henke W (Springer, Berlin), Vol. 3, pp 1539–1573.
2. Robinson JT (1954) Prehominid dentition and hominid evolution. Evolution 8:324–334.
3. Rak Y (1983) The Australopithecine Face (Academic, New York).
4. Teaford MF, Ungar PS (2000) Diet and the evolution of the earliest human ancestors.
Proc Natl Acad Sci USA 97:13506–13511.
5. Strait DS, et al. (2009) The feeding biomechanics and dietary ecology of Austral-
opithecus africanus. Proc Natl Acad Sci USA 106:2124–2129.
6. Grine FE, et al. (2010) Craniofacial biomechanics and functional and dietary inferences
in hominin paleontology. J Hum Evol 58:293–308.
7. Broom R (1938) Pleistocene anthropoid apes of South Africa. Nature 142:377–379.
8. Gregory WK, Hellman M (1939) The dentition of the extinct South African man-ape
Australopithecus (Plesianthropus) transvaalensis. Ann. Transv. Mus 19:339–373.
9. Broom R, Robinson JT (1950) Man contemporaneous with the Swartkrans ape-man.
Am J Phys Anthropol 8:151–155.
10. Broom R, Robinson JT (1952) Swartkrans ape-man Paranthropus crassidens. Transvaal
Mus. Mem. 6. pp 1–123.
11. Kimbel WH, Rak Y, Johanson DC (2004) The Skull of Australopithecus afarensis (Ox-
ford Univ Press, Oxford).
12. Walker AC, Leakey RE, Harris JM, Brown FH (1986) 2.5-Myr Australopithecus boisei
from west of Lake Turkana, Kenya. Nature 322:517–522.
13. Asfaw B, et al. (1999) Australopithecus garhi: A new species of early hominid from
Ethiopia. Science 284:629–635.
14. Skelton RR, McHenry HM (1992) Evolutionary relationships among early hominids.
J Hum Evol 23:309–349.
15. Strait DS, Grine FE, Moniz MA (1997) A reappraisal of early hominid phylogeny. J Hum
16. Strait DS, Grine FE (2004) Inferring hominoid and early hominid phylogeny using
craniodental characters: The role of fossil taxa. J Hum Evol 47:399–452.
17. McCollum MA (1999) The robust australopithecine face: A morphogenetic perspec-
tive. Science 284:301–305.
18. Ward CV, Leakey MG, Walker A (2001) Morphology of Australopithecus amanuensis
from Kanapoi and Allia Bay, Kenya. J Hum Evol 41:255–368.
19. Kimbel WH, White TD, Johanson DC (1988) Implications of KNM-WT 17000 for the
evolution of the ‘robust’ Australopithecines. Evolutionary History of the Robust
Australopithecines, ed Grine F (Gruyer, New York), pp 259–268.
20. Johanson DC, White TD (1979) A systematic assessment of early African hominids.
21. White TD, Johanson DC, Kimbel WH (1981) Australopithecus africanus: Its phyletic
position reconsidered. S Afr J Sci 77:445–470.
22. Kimbel WH, White TD, Johanson DC (1984) Cranial morphology of Australopithecus
afarensis: A comparative study based on a composite reconstruction of the adult skull.
Am J Phys Anthropol 64:337–388.
23. Rak Y (1985) Australopithecine taxonomy and phylogeny in light of facial morphol-
ogy. Am J Phys Anthropol 66:281–287.
24. Lockwood CA, Tobias PV (2002) Morphology and affinities of new hominin cranial
remains from Member 4 of the Sterkfontein Formation, Gauteng Province, South
Africa. J Hum Evol 42:389–450.
25. Lockwood CA, Tobias PV (1999) A large male hominin cranium from Sterkfontein,
South Africa, and the status of Australopithecus africanus. J Hum Evol 36:637–685.
26. Tobias PV (1991) Olduvai Gorge (Cambridge Univ Press, Cambridge, UK) Vol. 4: The
Skulls, Endocasts and Teeth of Homo habilis.
27. McKee JK (1989) Australopithecine anterior pillars: Reassessment of the functional
morphology and its phylogenetic relevance. Am J Phys Anthropol 80:1–9.
Villmoare and Kimbel PNAS
| September 27, 2011
| vol. 108
| no. 39