Technical Comments

Response to Comment on “Early Homo at 2.8 Ma from Ledi-Geraru, Afar, Ethiopia”

See allHide authors and affiliations

Science  19 Jun 2015:
Vol. 348, Issue 6241, pp. 1326
DOI: 10.1126/science.aab1122


Hawks et al. argue that our analysis of Australopithecus sediba mandibles is flawed and that specimen LD 350-1 cannot be distinguished from this, or any other, Australopithecus species. Our reexamination of the evidence confirms that LD 350-1 falls outside of the pattern that A. sediba shares with Australopithecus and thus is reasonably assigned to the genus Homo.

Hawks et al. (1) claim that we misinterpreted the mandibular anatomy of the Malapa hominins and, as a result, failed to satisfactorily distinguish the LD 350-1 jaw from Australopithecus.

We stand by our assessments of the mandibular and dental anatomy of A. sediba. On the right side of the MH2 mandibular corpus, the inferior border is damaged posterior to mid-M2, but any reasonable estimate of “minimum corpus depth” [as defined in (2); Hawks et al. do not specify their method] below M2 is 2.5 to 3.5 mm less than the depth below P3. The left side of the MH2 corpus [specimen UW88-55 (3)] is also preserved, however, and is undamaged from the M1 level to posterior to M3. The inferior margin of the damaged right corpus (4) can be inferred by photographically superimposing the left side onto the right (Fig. 1), which reveals the primitive (australopith-like) height relationships of the corpus, deeper anteriorly and considerably less so posteriorly. On a cast of MH2, we measure minimum lingual corpus height at left M2 as 26.6 mm and at right P3 as 30.3 mm.

Fig. 1

(Left) Reversed image of the lingual aspect of the MH2 left mandibular corpus (cast of UW88-55) superimposed on the lingual aspect of the right side, with the damaged right inferior corpus margin indicated by a black dotted line. The left side was overlapped to match the size and position of the M3, the inferior margin of the corpus, and the preserved contours of the ramus of the right. Note the visible shallowing of the corpus toward the rear of the toothrow, indicated by the dashed red lines at P3 and M2 (Right) MH2 right mandible fragment UW 88-54 (cast) in anatomical position, lateral view. The origin of the anterior ramus margin (arrow) is clearly located at the M2/M3 level.

With regard to the juvenile MH1 mandible (M3 unerupted), our data on chimpanzees and A. afarensis show that, although corpus depth does indeed change with emergence of the final molar, the relative anterior to posterior depth (P3:M2) is essentially stable from the time of M2 emergence. Therefore, in MH1, the clear discrepancy in corpus depth [see figure 1 in (3)] is unlikely to have changed substantially with the attainment of adulthood, confirming our assessment of a posteriorly shallowing corpus in the Malapa hominins, as commonly seen in Australopithecus.

Hawks et al. argue that the orientation of the mental foramen in MH2 is lateral, rather than anterior, as we described. Without prejudicing our assessment of MH2, we note that a lateral orientation of the foramen fits the description of A. africanus [as noted in (2)], whereas a clear posterior opening of the foramen into a short groove on the corpus, as in LD 350-1, is typically seen only in Homo among nonrobust early hominins. With regard to the root of the ascending ramus, our definition [as noted in the legend of figure S7 in (2)] is the point at which the anterior ramus margin becomes independent of the corpus, which on MH2 clearly occurs at M2/M3, as we described (Fig. 1). It is the concave vertical segment of the anterior margin that leaves the M3 visible in lateral view, not an extreme posterior origin of the margin as in LD 350-1.

The flaring buccal face of the molars in MH1, the standard condition in Australopithecus, may or not have been present in MH2. Heavy occlusal wear, to the point of substantially reducing crown height and eliminating virtually all cusp detail, precludes assessing the trait for MH2. Although Hawks et al. acknowledge that the LD 350-1 M3 is shorter than the M2, a trait that is not seen in A. sediba, they propose that M3 length cannot be used to discriminate LD 350-1 from this taxon. Our argument was not about overall length of M3, however; rather, it is the relative length of the M3 to the M2 that discriminates LD 350-1 (and Homo more broadly) from the great majority of Australopithecus mandibular tooth rows. We made no taxonomic case with regard to mesial/distal tapering of M2, so it is unclear why it is part of Hawks et al.’s argument. We do note that, in this feature, MH1 falls well within the range of A. afarensis and A. africanus and does not overlap the Homo distribution [see table S7 in (2)].

Hawks et al. state that “the evidence cannot reasonably reject” the taxonomic allocation of LD 350-1 to A. afarensis, A. africanus, or A. sediba. We find this confusing, because irrespective of whether or not it would be reasonable to assign LD 350-1 to A. sediba, we provided a detailed account of the characters distinguishing LD 350-1 from both A. afarensis and A. africanus, which these authors do not contest. We acknowledge that there is variation in some of these traits across early hominins, as noted in our analysis, but it is their appearance in combination that supports assigning the LD 350-1 mandible to an early part of the Homo lineage.

There is further confusion in Hawks et al.’s argument that LD 350-1 cannot reasonably be attributed to Homo. Although they assigned the Malapa hominins to Australopithecus based on a methodological choice that emphasizes primitive characters in taxonomy, Berger and colleagues identified derived characters in these fossils that they argued supported a unique phylogenetic link of A. sediba to the Homo lineage (3, 5). Now, however, they contend that our identification in the LD 350-1 mandible and teeth of derived features seen in later Homo cannot support the assignment of the specimen to this genus because of their (supposed) occurrence in Australopithecus sediba. Our emphasis on derived traits marries taxonomic practice to phylogenetic inference; their emphasis on primitive traits divorces the two. This contradiction is a common source of confusion in paleoanthropology and here renders Hawks et al.’s objections to our taxonomic conclusions regarding LD 350-1 logically incoherent.


View Abstract

Navigate This Article