Extensive studies of fossil skeletons of Australopithecus sediba provide fascinating details of the anatomy of this hominin species, but do not convincingly indicate its position on the evolutionary route to modern humans.
The evolutionary events that led to the origin of the Homo lineage are an enduring puzzle in palaeoanthropology, chiefly because the fossil record from between 3 million and 2 million years ago is frustratingly sparse, especially in eastern Africa. Much attention has been paid to two fossilized skeletons, found in approximately 2-million-year-old sediments at the Malapa cave site in South Africa, that are recognized as representing the species Australopithecus sediba. These have been the focus of scrutiny because of both their excellent preservation and claims1,2 that this hominin — a species more closely related to humans than to chimpanzees — lies at the base of the Homo lineage. A series of reports published in Science3,4,5,6,7,8sheds light on the morphology of A. sediba but, in my view, does little to elucidate its role in later human evolution.
Dental morphology is a frequent source of information about hominin phylogeny but, in the first of these new papers, Irish et al.3 take the unconventional step of using only the Arizona State University Dental Anthropology System — a graded series of minor crown variants originally devised to distinguish recent human populations from one another — to decipher relationships between hominin species that are millions of years old. I have serious doubts about the phylogenetic meaning of morphological similarity in this case. These concerns are compounded by the authors' reliance on the gorilla as the sole outgroup in their cladistic analysis. Their results link A. sediba exclusively to Australopithecus africanus, an older (approximately 2.7 million to 2.3 million years old), potentially ancestral, southern African species with which it also shares some key cranial features1. If this finding is borne out by further work, then the relevance of A. sediba to the origin of Homo would be inextricably tied to that of A. africanus, whose own position in hominin phylogeny is by no means settled9.
De Ruiter and colleagues' analysis of the A. sediba mandible4 includes a measurement-based comparison in which the sub-adult individual MH1(with only its second molar erupted) is treated as though its growth had been completed. However, for most dimensions, hominoid mandibles achieve only around 75–90% of their adult values by the time of the second molar eruption10. So, although the A. sediba mandibles seem to be small and lightly built (and thus Homo-like) by australopith standards, it is unclear how much of this impression is due to the authors' use of a sample comprising a sub-adult and a presumed adult female (MH2).
Much of the value of the Malapa material lies in the extremely rare association of upper and lower limb parts with elements of the axial skeleton in two individuals of the same species (Fig. 1). These skeletons paint a portrait of a pectoral girdle that retains more ape-like anatomy than the pelvic girdle5,11. Churchill et al.5 report that a fairly complete scapula (from MH2) features an upwardly tilted articulation for the humerus and a relatively broad attachment area for a muscle that helps to lift the arm over the head, a familiar australopith upper-limb pattern that also includes long, strong forearms and curved fingers. Although these features are embedded in a terrestrial bipedal frame, they are often interpreted as signs of retained ancestral arboreal climbing behaviour5,12. Still unsettled is what led to the refashioning of the hominin shoulder by the time, around 1.6 million years ago, of Homo erectus, a species that shows modern upper-limb and shoulder morphology (this anatomy is unknown in the approximately contemporaneous Homo habilis). Simply leaving the trees seems to be an insufficient explanation.
Schmid et al.6 used the low curvature of the upper ribs of A. sediba to argue for a conical ribcage and elevated shoulders similar to those of the great apes, even though second- and fourth-rib curvatures do not actually distinguish apes from humans. However, it is clear that the unusually strongly curved first rib articulates only with the first thoracic vertebra, as in humans and Australopithecus afarensis. This configuration is at odds with a completely ape-like upper thorax and has been associated with descent of the shoulder after the upper limbs were freed from locomotion13, although this interpretation has been contested14. A further puzzle is the A. afarensis partial skeleton KSD-VP 1/1, which, although 1.6 million years older than the A. sediba skeletons, has an upper thorax more similar to that of modern humans15. A lower (ninth) rib from MH2, which is more like those of modern humans in its curvature and torsion, is consistent with the less flaring pelvic rim of A. sediba when compared with A. afarensis and A. africanus6,11.
The 'long-backed' lumbosacral vertebral formula of six lumbar and four sacral vertebrae that is seen in species from australopiths through to Homo erectus is the probable primitive condition for humans and the great apes16. The evolution of the most common modern-human condition of five lumbar and five sacral vertebrae occurred by 'sacralization' of the lowermost lumbar vertebra. Williams et al.7 argue that MH2 is unusual for an early hominin in having this derived pattern, but the MH2 formula depends on how one defines a lumbar vertebra — by a lack of rib articulations or by functional criteria that relate to intervertebral movement. In early hominins, the first of the six functional lumbar vertebrae carries rib articulations that are similar to those of a thoracic vertebra. MH2, in fact, resembles other Australopithecus specimens in having six functional lumbars, but this is unexpected in an early hominin with five sacral vertebrae.
The last of the papers presents DeSilva and colleagues' reconstruction of the A. sediba gait8 (based on the MH2 skeleton), which will be controversial. The proposed 'hyperpronation' of the foot and extreme inward rotation of the leg and thigh suggest an ungainly bipedal stride that might have made it into Monty Python's 'Ministry of Silly Walks' sketch. The presumed inversion of the foot at the heel-strike of the surprisingly ape-like calcaneus, combined with a vertical shank (tibia), outwardly angled thigh (femur) and a long, lordotic lower back — all hallmarks of terrestrial bipedality in Australopithecus and Homo species — constrains the reconstruction. Prominent osteophytic growths on the pelvis and fibula at the attachment sites of the thigh musculature raise the possibility of a gait that was pathologically impaired, but DeSilva and colleagues argue that this locomotor pattern was adaptive. However, if A. sediba was a descendant of A. africanus, which, similarly to the even older A. afarensis (dating to between 3.7 million and 3.0 million years ago), shows no trace of this pattern, then it is hard to imagine the selective advantage that would accrue from such a kinematically peculiar gait.
Given the mix of features seen in A. sediba, it is difficult to understand why these researchers insist that it lies at the base of the Homo lineage1,2. Similar intellectual gymnastics are required to comprehend the authors' argument that no African Homo fossils exist from before the time of A. sediba2,17. Although the recent papers constitute a fascinating further analysis of the A. sediba fossils, I do not think that they provide compelling evidence that this species is anything other than an unusual australopith from a Pliocene–Pleistocene time period that is already populated by a fair number of them18.
Berger, L. R. et al. Science 328, 195–204 (2010).
Pickering, R. et al. Science 333, 1421–1423 (2011).
Irish, J. D., Guatelli-Steinberg, D., Legge, S. S., de Ruiter, D. J. & Berger, L. R. Science http://dx.doi.org/10.1126/science.1233062 (2013).
de Ruiter, D. J. et al. Science http://dx.doi.org/10.1126/science.1232997 (2013).
Churchill, S. E. et al. Science http://dx.doi.org/10.1126/science.1233477 (2013).
Schmid, P. et al. Science http://dx.doi.org/10.1126/science.1234598 (2013).
Williams, S. A. et al. Science http://dx.doi.org/10.1126/science.1232996 (2013).
DeSilva, J. M. et al. Science http://dx.doi.org/10.1126/science.1232999 (2013).
Grine, F. E. in The Paleobiology of Australopithecus (eds Reed, K. E., Fleagle, J. G. & Leakey, R. E.) 73–104 (Springer, 2013).
Humphrey, L. T. in Human Growth in the Past: Studies from Bones and Teeth (eds Hoppa, R. & FitzGerald, C.) 65–87 (Cambridge Univ. Press, 1999).
Kibii, J. M. et al. Science, 333, 1407–1411 (2011).
Larson, S. G. in The Paleobiology of Australopithecus (eds Reed, K. E., Fleagle, J. G. & Leakey, R. E.) 247–2621 (Springer, 2013).
Ohman, J. C. Am. J. Phys. Anthropol. 70, 209–229 (1986).
Stern, J. T. Jr & Jungers, W. L. Am. J. Phys. Anthropol. 82, 431–439 (1990).
Haile-Selassie, Y. et al. Proc. Natl Acad. Sci. USA 107, 12121–12126 (2010).
McCollum, M. A., Rosenman, B. A., Suwa, G., Meindl, R. S. & Lovejoy, C. O. J. Exp. Zool. 314B, 123–134 (2010).
Berger, L. R. J. Anthropol. Sci. 90, 1–16 (2012).
Kimbel, W. H. in Handbook of Paleoanthropology Vol. 3 (eds Henke, W. & Tattersall, I.) 1539–1574 (Springer, 2007).
About this article
Statistical estimates of hominin origination and extinction dates: A case study examining the Australopithecus anamensis–afarensis lineage
Journal of Human Evolution (2020)
Expanded character sampling underscores phylogenetic stability of Ardipithecus ramidus as a basal hominin
Journal of Human Evolution (2019)
American Journal of Physical Anthropology (2018)
Journal of Human Evolution (2017)
Annual Review of Anthropology (2017)