Human postcranial morphology varies with climate and geography (ecogeography), a pattern that has long been associated with thermoregulatory adaptation. The thermoregulatory imperative model of postcranial evolution suggests that groups living in colder climates have evolved stouter bodies to reduce their surface area/volume ratio, while the opposite is true for groups in the tropics. Recent applications of quantitative genetics methods to human postcranial evolution have allowed researchers to move beyond testing whether limb lengths and measures of body size adhere to ecogeographic expectations, and begin disentangling the natural selection from neutral evolutionary processes.

However, these have continued to model postcranial traits as if they were evolving independently. By examining human evolution on a trait-by-trait basis, researchers fail to account for evolution due to changes in correlated morphology. In addition, while these studies are able to identify the direction and magnitude of selective effects, they have yet to identify the source of that selective pressure, often relying on latitude as a problematic proxy for “climate.” In this dissertation, I use quantitative genetic methods to elucidate the role of trait covariation in the evolution of the human postcranium, as well as explore the variables that drive directional selection. Using a large, globally-dispersed sample of human skeletal material, in addition to microsatellite and temperature data, I estimate the selection gradients driving among-group differentiation, compare indices of evolvability across regions, and examine the environmental variables influencing postcranial evolution.

Results indicate that 1. trait covariation plays an important role in shaping evolutionary response, 2. human groups may demonstrate emergent differentiation in evolvability across regions , and 3. the selective pressures acting on the postcranium are likely synergistic, complicating simplistic interpretations of the thermoregulatory imperative model. These findings have implications for our understanding of modern human variation, suggesting the need to develop multivariate models in which the reciprocal effects of multiple environmental variables can be examined on the covariance structure of multiple traits. Results also add to the growing evidence that population-specific trait covariance prevents evolutionary interpretations founded on modern humans to be meaningfully translated to ancient hominin lineages.