Compared to their small body mass, the forelimbs of early hominids are both longer and more muscular than those of recent humans. The arms are shorter than in chimpanzees, but the areas of muscle attachment have greater strength. Strength is especially evident in a large humerus from the Ethiopian site of Maka, dating to 3.4 million years ago (White et al. 1993). The prominent muscle attachments on this large specimen indicate that the individual was very strong, but also that most muscle exertion was in a single preferred pattern. The bone is thicker than chimpanzee humeri of equal length, again reflecting its mechanical strength.
Several other forelimb fossils show a similar pattern. These include a relatively large distal humerus fragment from Kanapoi, a large radius from the contemporary site of Sibilot Hill, and a distal section of humerus exhibiting large muscular crests from Lukeino (Senut et al. 2001). Additionally, the hamate bone preserved at the Kenyan site of Turkwel preserves part of a very large carpal tunnel region, indicating strong tendon attachments into the hand (Ward 1999). Finally, the finger bones of early hominids are curved. This feature occurs wherever Early Pliocene hominid finger bones are preserved (Stern and Susman 1983).
The most probable interpretation for the strength of the forelimbs among these sites is that early hominids were effective climbers. Hominids do not have grasping feet. Apes use all four limbs in a variety of climbing positions, but hominids use their arms mainly to pull the body upward with the legs providing upward propulsion but no gripping support. Such use would lead to large muscle development in the arms, both because they bore more of the force of climbing and because they functioned in a more specialized way.
But despite the strength of the large fossil arms, smaller individuals show a somewhat different pattern. For example, the AL 288-1 skeleton preserves much of both humeri and ulnae, and these small bones bear minimal muscle markings. The difference in forelimb anatomy between large and small individuals may mean either that males climbed more frequently than females or that the biomechanics of climbing among maleswith a mass much larger than femalesplaced much greater muscle requirements on the male forearm.
Males may have performed other tasks with their arms, including wielding weapons or other competitive behaviors such as threat displays. The hands of early hominids, as represented at Hadar, have fingers that are similar to living humans in their relative sizes. These proportions are very different than in chimpanzees, which have a much shorter thumb that departs the hand much closer to the wrist, as well as much longer fingers. The human-like proportions of the hands underline the fact that climbing in these obligate bipeds was done in a human-like manner, and their hands did not function in a chimpanzee-like suspensory role. Also, early hominids may have gripped clubs or other items that do not require forceful fingertip control. Chimpanzees wield large branches in the context of threat displays, and it is possible that early hominids also had such behaviors or even more menacing ones, enabled by the power of arms like those from Maka.
However, early hominid hands were clearly not used for making stone tools. Two sources of evidence argue against the ability of these early hominids to modify stone. First, no stone tools have yet been found earlier than about 2.6 million years ago, long after the early Hadar sample. Second, the Hadar hands and other early hominid hand bones lack important features that reflect a powerful grip useful for tool production (Marzke 1983). Most noticeably, the distal finger bones, or phalanges, lack the large fingertip surfaces, called apical tufts, which are found in living and fossil humans (Bush et al. 1982). These large fingertips increase the surface area used for gripping, and allow the forceful grip necessary for tool production. The apical tufts at Hadar are relatively much smaller than those in human fingers (Stern and Susman 1983).