Interlimb Proportions in Humans and Fossil Hominins: Variability and Scaling

Our understanding of the evolutionary trajectory of hominin limb proportions and overall body shape is severely hampered by the paucity of fossil individuals with associated elements from early in the human career (Walker, 1973; McHenry and Coffi ng, 2000; Richmond et al., 2002; Reno et al., 2005). It is now apparent that hind-limb elongation and modern human interlimb proportions emerged at least by the time of early Homo erectus (=ergaster) as represented by the fossils from Dmanisi at almost 1.8 million years ago (Lordkipanidze et al., 2007) and the “Nariokotome boy”, KNM-WT 15000, at approximately 1.6 million years ago (Ruff, 1993; Ruff and Walker, 1993). Interlimb proportions based on complete bone lengths in australopithecines are poorly known, and “Lucy” (A.L. 288-1, Australopithecus afarensis) still represents our best example of the “primitive” hominin condition (Johanson and Edey, 1981). With a humer-ofemoral index (100 × humerus length/femur length) near 85, A. afarensis is “intermediate” between African apes and modern humans (Jungers, 1982, 1991; Jungers and Stern, 1983; Richmond et al., 2002). This difference from humans is driven by a relatively short femur (but not as short as in apes), not by long arms (Jungers, 1994). It is also important to note that interlimb proportionality based on lengths need not correspond to estimates of interlimb size and shape based on dia-physeal and articular dimensions (McHenry, 1978; McHenry and Berger, 1998; Green et al., 2007). Long bone lengths, especially of the hind-limb elements, are most relevant to questions about the evolution of locomotor effi ciency and the kinematics of bipedalism (e.g., Jungers, 1982; Bramble and Lieberman, 2004; Pontzer, 2005; Steudel-Numbers, 2006).

BOU-VP-12/1 is another skeleton of a later australopith (possibly A. garhi), but its humerus and femur are both quite damaged and incomplete (Asfaw et al., 1999). Estimated and reconstructed lengths of these long bones are controversial and exhibit large confi dence intervals, whether based on regression or by eye (Richmond et al., 2002; Haeusler and McHenry, 2004; Reno et al., 2005). Although possible, it seems premature to conclude with any confi dence that its humerofemoral proportions were already “human-like” (Reno et al., 2005). Similarly, Homo habilis as represented by OH 62 has essentially indeterminate interlimb proportions, probably lying somewhere between those of gorillas and modern humans (Korey, 1990; Reno et al., 2005; but see Haeusler and McHenry, 2004). It appears risky at this time to develop complex evolutionary scenarios about limb length proportions predicated on BOU-VP-12/1 and/or OH 62.