Dental Morphology and the Phylogenetic “Place” of Australopithecus sediba

See allHide authors and affiliations

Science  12 Apr 2013:
Vol. 340, Issue 6129, 1233062
DOI: 10.1126/science.1233062


To characterize further the Australopithecus sediba hypodigm, we describe 22 dental traits in specimens MH1 and MH2. Like other skeletal elements, the teeth present a mosaic of primitive and derived features. The new nonmetric data are then qualitatively and phenetically compared with those in eight other African hominin samples, before cladistic analyses using a gorilla outgroup. There is some distinction, largely driven by contrasting molar traits, from East African australopiths. However, Au. sediba links with Au. africanus to form a South African australopith clade. These species present five apomorphies, including shared expressions of Carabelli’s upper first molar (UM1) and protostylid lower first molar (LM1). Five synapomorphies are also evident between them and monophyletic Homo habilis/rudolfensis + H. erectus. Finally, a South African australopith + Homo clade is supported by four shared derived states, including identical LM1 cusp 7 expression.

Given their propensity to outlast other hard tissues, teeth provide sizable, although restricted, data sets to define and characterize fossils. Thus, when alternative remains are available, as with Australopithecus sediba (1), the focus shifts readily to elements that give insight into hominin abilities such as cognition (2), manipulation (3, 4), parturition (5), and locomotion (59). Still, teeth have three key attributes that permit a singular perspective on hominins and their origins. First, crowns come into actual contact with the environment, so that direct inferences can be made about diet and certain behaviors (10). Second, because teeth preserve their record of incremental growth, a fine-grained understanding about the pace of hominin life histories is possible (11). Third, their high genetic component, among other features in expression (12), facilitates phylogenetic interpretations. Focusing on the latter attribute, we describe 22 nonmetric crown and root traits. These descriptions are then contrasted with those of six fossil and two recent hominin samples. Finally, after coding characters across these samples and an outgroup for cladistic analysis, we present additional phylogenetic information concerning this 1.977 million-year-old (13) South African australopith.

Some dental data in Au. sediba have been presented (1, 14). However, many of those characters are unique to Plio-Pleistocene hominins (1517). The traits used here are from the Arizona State University Dental Anthropology System (ASUDAS) (18), most of which are expressed in modern and fossil hominins, plus various other primates. Traits in the ASUDAS are highly heritable and minimally affected by sexual dimorphism, among other attributes (12). Further, as “minor” variants they should be little affected by selection, and pairwise correlations are negligible (18, 19). A lack of selection would reduce potential homoplasy (12). Low or no correlations address character redundancy that would bias cladistic analyses; this issue is crucial for molar traits, given the identified linkage, albeit minor, between accessory cusp formation and intercusp spacing and crown size (20). Only low intertrait correlations were detected for the present traits (12).

We inspected >340 fossil and 4571 recent hominin specimens retaining teeth, along with 44 gorilla dentitions; recording of the requisite traits (12) yielded the samples presented herein. The Au. sediba dental remains derive from MH1, a juvenile male, and MH2, an adult female (1). Data from the fossil comparative samples were collected in original specimens and 11 high-resolution casts (table S1) (21, 22). The east African australopith sample consists of Au. anamensis and Au. afarensis. Shared ASUDAS expression in these two time-successive species permitted pooling [see (21)] so as to most comprehensively characterize the Australopithecus genus in this region. Homo habilis/rudolfensis includes specimens assigned by some to different, albeit contemporaneous, species that also share ASUDAS expression. Similarly, H. erectus (also known as ergaster) includes specimens from what some maintain are different species (23). The remaining samples comprise Au. africanus, Paranthropus robustus, and P. boisei. The rationale to assign specimens to species is presented elsewhere (21, 22). In brief, sample composition is based on a majority consensus of researchers, while maximizing characterization for comparative purposes.

The recent hominin samples consist of post-Pleistocene H. sapiens from sub-Saharan (n = 2309) and North Africa (n = 2262) (2426). Crown data in Gorilla gorilla were recorded in a collection from Cameroon; root descriptions are extrapolated from published sources (27, 28). The latter species is used as an outgroup for cladistic analysis, which is appropriate given that it and Pan are the closest extant relatives of hominins (15).

Twenty-two traits that could be recorded in the two Au. sediba specimens, their ASUDAS grades, and expressions in the species are listed in Table 1. Comprehensive trait and grade definitions are available elsewhere (18, 19). Similar descriptions and character states for the cladistic analysis are listed in Table 2 for all samples. In this table, state values replicate ASUDAS grades, with one exception. The ASUDAS grade range of 0 to 7 for Carabelli’s upper first molar (UM1) was increased to 0 to 8, with definitions (19) shifted up for each value to account for the gorilla cingulum variant (now given a value of 0); grade 0 is treated here as the primitive state, although not all researchers may concur. This modification accounts for the change in Au. sediba Carabelli’s values (from 4 to 5) between tables.

Table 1

ASUDAS traits, grades, and descriptions in Australopithecus sediba specimens.

View this table:
Table 2

ASUDAS traits and grades and the distribution of their states across the gorilla outgroup and hominin samples.

View this table:

Fourteen of 22 character states are identical between Au. sediba and east African australopiths. The count between Au. sediba and Au. africanus is 15, and both are similar for a 16th state (protostylid lower first molar, LM1). All Australopithecus hypodigms are somewhat similar, but the South African taxa are notably alike for lower molar cusp number, size, and pattern expression. Australopithecus sediba has 15 states in common with H. habilis/rudolfensis, and 13 of 21 with H. erectus; it is less akin to recent H. sapiens and is distinct from Paranthropus—owing to the latter’s mass-additive molar morphology. Shared states among other species are also instructive; for example, Au. africanus shares additional ASUDAS traits with early Homo. The most expeditious way to present these links is with a dendrogram (fig. S1), where an Au. sediba/Au. africanus group is adjacent to an early/later Homo cluster, and east African australopiths and Paranthropus are more distant. Of course, character states of these traits were not determined for this phenetic comparison; some are symplesiomorphic, so results should be interpreted with caution.

Both primitive and derived dental features are present in Au. sediba. The former, like many hominins including Au. africanus, includes two-rooted upper third premolar (UP3), hypocone UM2, UM3 presence, Tome’s root lower third premolar (LP3), two-rooted LM1, two-rooted LM2, five-cusped LM1, five-cusped LM2, and LM2 Y-groove. The last three traits in east African australopiths often include six cusps for LM1 and LM2, and some occurrence of an X-groove pattern (below). Derived traits in Au. sediba include a decrease in labial curvature upper first incisor (UI1), slight UI1 shoveling, and increases in Carabelli’s UM1, LM1 protostylid, and LM1 cusp 7 expression (fig. S2). Other derived dental features, including overall size reduction and a specific decrease in size and complexity of the canine and posterior teeth, are detailed elsewhere (1, 14).

Numeric states in Gorilla, Au. sediba, and the comparative samples were next submitted for cladistic analysis with PAUP 4.0b10 (29). Coding of binary and multistate characters emulates a standard approach in the above (1, 1517) and other systematic studies. Polarity was determined via rooting of the outgroup, and all characters were treated as ordered and of equal weight. Wagner parsimony and the branch and bound method were used to identify optimal cladograms. Of the 22 characters, root number LM1 is constant, and three were deemed parsimony-uninformative (double shovel UI1, UM3 agenesis, and root number LM2); the expression of all four represents the primitive condition. The remaining 18 traits yielded one cladogram (Fig. 1) with a tree length of 76, consistency index of 0.62, retention index of 0.48, and rescaled consistency index of 0.3 [interpretation in (12)].

Fig. 1 Maximum parsimony cladogram of gorilla outgroup and nine hominin samples based on ASUDAS characters.

Numbers are from a separate, although analogous 50% majority consensus tree of 10,000 bootstrapped replicate data sets; they represent the proportion of included trees that support the given node. SSA, sub-Saharan Africans; NAF, North Africans. See text for all methods employed and sample compositions.

In parallel with the phenetic findings, Au. sediba is depicted as a sister species of Au. africanus to form a South African australopith clade. These species share five apomorphies: (i) weak labial curvature UI1; (ii) faint shovel UI1 (both are parallel with the east African australopiths, Paranthropus, and Homo clades); (iii) large depression-to-cusp Carabelli’s UM1 (parallel with P. robustus and North African H. sapiens); (iv) trace-small cusp protostylid LM1; and (v) small cusp 7 LM1 (parallel with H. habilis/ rudolfensis + H. erectus clade and sub-Saharan African H. sapiens). Synapomorphies between the South African gracile australopiths and monophyletic H. habilis/rudolfensis + H. erectus include (i) weak labial curvature UI1; (ii) faint shovel UI1 (both parallel with the east African australopiths, Paranthropus, and H. sapiens); (iii) faint distal accessory ridge UC (parallel with H. sapiens); (iv) >2 root number UP3; and (v) small cusp 7 LM1 (parallel with sub-Saharan African H. sapiens). A South African australopith + Homo clade is supported by four shared derived states: (i) weak labial curvature UI1, (ii) faint shovel UI1 (both parallel with east African australopith and Paranthropus), (iii) faint distal accessory ridge UC, and (iv) small cusp 7 LM1.

Although the focus is on Au. sediba, the remaining clades are defined by characters of interest as well, and warrant brief mention, i.e., H. habilis/rudolfensis + H. erectus (e.g., Carabelli’s cusp UM1), Homo sapiens (e.g., Bushman canine), and east African australopiths + Paranthropus. The latter grouping is seemingly nonintuitive but not unique; comparable classifications were identified using a wide range of characters (30, 31). Here, Paranthropus and, as noted, the east African australopiths—especially Au. afarensis, are characterized by some unique molar states such as trace-small cuspule for cusp 5 UM1, six-cusped LM1, absent-pit protostylid LM1, absent cusp 7, and six-cusped LM2. These shared expressions were reported elsewhere (12, 22) and were suggested to be symplesiomorphies and/or homoplasies.

A 50% majority-rule consensus tree of 10,000 bootstrapped replicate data sets was used to assess the stability of all clades; it is analogous to the maximum parsimony cladogram, so percentages of included trees supporting each clade were simply added to Fig. 1. Overall, the location of Au. sediba + Au. africanus at the stem of the Homo clade corresponds well with prior phylogenetic findings using >60 craniodental characters (1), which suggested that Au. sediba was derived from Au. africanus. This proposal is sustained here. Berger et al. (1) also stated that Au. sediba is more derived toward Homo than are east African australopiths and Au. africanus; as such, A. sediba may represent (i) a sister group to a later ancestor that was an initial contemporary of early Homo, or (ii) a candidate ancestor to the latter genus (1). Both scenarios are supported, although, at least for these new data, Au. africanus does share additional ASUDAS traits with early Homo; as such, Fig. 1 and the study’s (1) 60-character cladogram differ in that the latter depicts Au. africanus as more distant than Au. sediba from the Homo clade.

The conclusions from the dental data presented here concur with those derived from the virtual brain endocast, and hand, spine, pelvis, foot, and ankle elements of Au. sediba (27); each area presents a mosaic of primitive and derived characters that, assuming they are homologies, is not unexpected in a transitional species or one that, minimally, reflects a close phyletic relationship with Homo (5). Moreover, with regard to east African australopiths, it is possible that disparity in expression of molar states with their southern counterparts is also paralleled by other elements; differences in the foot, for example, prompted Zipfel et al. (6) to remark that an ancestor-descendant relationship between Au. afarensis and Au. sediba is unlikely or, if it existed, would have required several evolutionary reversals.

Although the phylogenetic place of Au. sediba has not been settled, the dental data serve to further define its position relative to other genera. It is distinct from Au. afarensis, close to Au. africanus and, along with the latter, shares a number of apomorphies with Homo. The relationship to H. habilis/rudolfensis had been suggested to reflect an adaptive evolution some two million years ago toward a definitive Homo grade (13); that is, Au. sediba could just as likely be the candidate ancestor of H. erectus as that of the majority consensus H. habilis and/or H. rudolfensis (13). However, determination of which taxon (or taxa) led to Homo erectus is, at present, not explicitly demonstrable. The difficulty is, to paraphrase Pickering et al. (13), under an assumption of substantial homoplasy, to discern phylogenetic affinities in such closely related species based on so few remains. Only recovery of additional Au. sediba and other Plio-Pleistocene remains can fill in the missing pieces of this evolutionary puzzle.

Supplementary Materials

Materials and Methods

Supplementary Text

Figs. S1 and S2

Table S1

References (3265)

References and Notes

  1. Methods and background are available as supplementary materials on Science Online.
  2. Acknowledgments: Collection of the recent African comparative data was funded by monies to J.D.I. from the National Science Foundation (BCS-0840674, BNS-0104731, BNS-9013942), National Geographic Society, Wenner-Gren Foundation, Institute for Bioarchaeology, University of Alaska, and the Combined Prehistoric and Hierakonpolis Expeditions. The innumerable individuals at >25 institutions from which the data were collected cannot be listed in this restricted space, but their help is greatly appreciated. These and the Au. sediba ASUDAS raw data are in the possession of J.D.I. Collection of the early hominin comparative data was funded by grants to D.G.-S. from the Leakey Foundation and Ohio State University. The authors thank D. Johanson and W. Kimbel for allowing D.G.-S. to study the Hadar fossils, M. Yilma of the National Museum of Ethiopia, C. Kiarie of the National Museums of Kenya, P. Msemwa and E. Maro of the National Museum of Tanzania, the Ethiopian government, and the governments of Kenya and Tanzania. H. Fourie and F. Thackery provided access to fossils at the Transvaal Museum in Pretoria; P. Tobias, B. Kramer, and K. Kuykendall provided access to the fossils at the University of the Witwatersrand. I. Tattersall, J. Schwartz, J. Lukacs, H. McHenry, I. Pike, J. McKee, and T. Harrison advised D.G.-S. regarding the logistics of her research in east and South Africa. The ASUDAS hominin data are in the possession of D.G.-S. The assistance of A. Gill and M. Harman at the Powell-Cotton Museum, Birchington, UK, from which the gorilla crown data (in the possession of S.S.L.) were recorded, is greatly appreciated. Funding for S.S.L. was provided by the Wallace Travel Fund, Macalester College, and the Paul Anderson Interdisciplinary Summer Research Fund; funding for D.J.D. was provided by the Ray A. Rothrock Fellowship, Texas A&M University.For the entire project, we also thank the South African Heritage Resource agency for the permits to work at the Malapa site; the Nash family for granting access to the Malapa site and continued support of research on their reserve; the South African Department of Science and Technology, the African Origins Platform (AOP), the Gauteng Provincial Government, the South African National Research Foundation, the Institute for Human Evolution, University of the Witwatersrand, the University of the Witwatersrand’s Vice Chancellor’s Discretionary Fund, the National Geographic Society, the Palaeontological Scientific Trust, the Andrew W. Mellon Foundation, the Ford Foundation, the U.S. Diplomatic Mission to South Africa, the French Embassy of South Africa, the A. H. Schultz Foundation, Duke University, Ray A. Rothrock ’77 Fellowship and International Research Travel Assistance Grant of Texas A&M University, and the Oppenheimer and Ackerman families and Sir Richard Branson for funding; the University of the Witwatersrand’s Schools of Geosciences and Anatomical Sciences and the Bernard Price Institute for Palaeontology for support and facilities; the Gauteng Department of Agriculture, Conservation and Environment and the Cradle of Humankind Management Authority; and our respective universities for ongoing support. For access to other comparative specimens, E. Mbua, P. Kiura, V. Iminjili, and the National Museums of Kenya; B. Billings, B. Zipfel, and the School of Anatomical Sciences at the University of the Witwatersrand; and S. Potze, L. C. Kgasi, and the Ditsong Museum. For technical and material support, Duke University, the University of Zurich 2009 and 2010 Field Schools, A. B. Taylor, C. E. Terhune, and C. E. Wall. Numerous individuals have been involved in the ongoing preparation and excavation of these fossils, including C. Dube, C. Kemp, M. Kgasi, M. Languza, J. Malaza, G. Mokoma, P. Mukanela, T. Nemvhundi, M. Ngcamphalala, S. Jirah, S. Tshabalala, and C. Yates. Other individuals who have given significant support to this project include B. de Klerk, W. Lawrence, C. Steininger, B. Kuhn, L. Pollarolo, B. Zipfel, J. Kretzen, D. Conforti, J. McCaffery, C. Dlamini, H. Visser, R. McCrae-Samuel, B. Nkosi, B. Louw, L. Backwell, F. Thackeray, and M. Peltier. J. Smilg facilitated computed tomography scanning of the specimens. The Au. sediba specimens are archived at the Institute of Human Evolution at the University of the Witwatersrand.
View Abstract

Stay Connected to Science

Navigate This Article