Michelle Drapeau and colleagues (2005) report on the AL 438-1 specimen from Hadar. The specimen consists of "part of the mandible, a frontal bone fragment, a complete left ulna, two second metacarpals, one third metacarpal, plus parts of the clavicle, humerus, radius, and right ulna" (1). At 3 million years, the specimen is one of the youngest of the A. afarensis sample.
The AL 438-1 individual was evidently relatively large compared to the rest of the Hadar sample. The ulna length is 278 mm, which is larger than the mean for any of the human samples examined by Aiello et al. (1999) in their comparative study of the OH 36 ulna. It is about average for a chimpanzee, although chimpanzees have relatively longer forelimbs than would have been true of A. afarensis, so again this is evidence of a relatively large body size.
Drapeau et al. (2005) make a point of the proportion of the ulna and the mandible being similar to that found in AL 288-1 (Lucy), which they take as evidence that large teeth in this late specimen may be attributed to larger body size rather than greater megadonty:
Both Australopithecus afarensis mandibles have a larger corpus (breadth and height at M1 relative to the ulnar size surrogate than those of African apes. Similarly, mandibular corpus shape (breadth/height x 100 at M1) is similar in the two fossils (A.L. 288-1, 57%; A.L. 438-1, 60%). This difference in mandibular size corresponds to what would be expected from two extant ape conspecifics with ulnae of such different sizes. Since there are no differences between the two Hadar skeletons in mandibular to ulnar proportions, there is no evidence for an increase in mandibular size relative to the rest of the skeleton between the points in time represented by these two individuals. We cautiously offer this as support for Lockwood et al.'s (2000) suspicion that the observed temporal trend toward larger mandible size reflects a body size increase late in the Hadar time span of A. afarensis (Drapeau et al. 2005:41-42).
The paper has a substantial discussion of the morphology and comparative anatomy of the ulna. The bottom line of this analysis is that the ulna is similar to that of AL 288-1 in most respects, except for its larger size and somewhat greater curvature. It is, however, smaller and somewhat less curved than the later Omo L40-19, and substantially less curved than the OH 36 ulna. The authors write this about its similarities to other homionids:
While phenetically A.L. 438-1 presents a mix of ape-like and human-like morphology, when considered in a phylogenetic context, the Australopithecus afarensis forelimb shares synapomorphies exclusively with humans among extant hominoids taxa. It resembles non-hominins only in plesiomorphic character states. In this context, it is apparent that A. afarensis forelimb anatomy reveals the results of selection for a more human-like humeroulnar joint, larger thumbs, and altered carpometacarpal joints that reflect an emphasis on manipulative aptitude at the expense of forelimb-dominated climbing ability (Drapeau et al. 2005:43).
That is a relatively powerful statement of the adaptive qualities of the A. afarensis forelimb, which appers more or less necessary to explain the differences between early hominids and apes in this respect. If the early hominids were really climbing a substantial proportion of the time, then we might hypothesize that their forelimbs ought to look more like ape arms. But they don't; there are clear differences that make the early hominid arms look more similar to human arms. Thus, the authors turn to the hypothesis that the A. afarensis forelimb is additionally adapted to "an emphasis on manipulative aptitude."
At the moment, this hypothesis remains to be strongly tested. Most of the human-like features of the A. afarensis arm are arguably the result of not being used in quadrupedal weight support. Thus, the fact that "the Australopithecus afarensis elbow joint appears to reflect habitual loading the elbow at or near 90 degrees, rather than optimization for loading in a more extended posture as in extant apes" (43-44), as well as the anatomy of the joint and the form of the carpometacarpal joints may all be explained by the fact that early hominids were not knuckle (or fist) walkers. The large thumbs are the strongest piece of evidence for any kind of manipulative behavior in A. afarensis.
On the subject of retained similarities with apes, Drapeau et al. (2005:46) have this to say:
The retention of African ape symplesiomorphies in A. afarensis may be attributed to either stabilizing selection fore a partially arboreal locomotor repertoire, or to lack of selection against these traits (see discussion in Stern, 2000; Ward, 2002). It is inherently difficult to test these alternative hypotheses. Thus, the significance of these retained traits for reconstructing the behavior of A. afarensis is difficult to determine with certainty in the context of demonstrable selection for a human-like elbow and hand joints. Australopithecus afarensis shares some apomorphies with humans that suggest emphasis on use of the forelimb in flexed postures, and improved grip capability relative to apes. The presence of these synapomorphies suggests similarities in forelimb function among hominins, likely reflecting selection for expanded manipulative capabilities and flexed forearm postures relative to that found in apes and a diminished capacity for ape-like arboreal behaviors. Only later did humans display evidence of further selection for manipulation coupled with reduced forelimb robusticity. We conclude that in Australopithecus afarensis, selection for natural manipulation outweighed selection for arboreal activities, but that selection for refined manipulative ability had not yet come into play in human evolution.
A fine balance, if it is true, and fitting within the generally understood picture that, with regard to its arm and hand functions, A. afarensis was either Homo habilis nor a chimpanzee.
Aiello LC, Wood B, Key C, and Lewis M. 1999. Morphological and taxonomic affinities of the Olduvai Ulna (OH 36). Am J Phys Anthropol 109:89-110.
Drapeau MSM, Ward CV, Kimbel WH, Johanson DC, and Rak Y. 2005. Associated cranial and forelimb remains attributed to Australopithecus afarensis from Hadar, Ethiopia. J Hum Evol Advance before print.
Lockwood CA, Kimbel WH and Johanson DC. 2000. Temporal trends and metric variation in the mandibles and dentition of Australopithecus afarensis. J Hum Evol 39:23-55.