Radius morphology variation in an adult South African cadaveric sample.

Abstract

Sex estimation is a vital first step in developing a biological profile, and is usually metrically or morphoscopically assessed. Geometric morphometric analyses allow re-evaluation of both size and shape variation of skeletal elements, and more meaningful comparisons of sexual dimorphism of different ancestry groups. Additionally, it has shown that some postcranial bones may be more useful for ancestry estimation than previously thought. Geometric morphometric analysis was performed on the radii of 1049 South African individuals (417 females, 578 males) of the country's three largest ancestry groups to assess morphological variation according to sex and ancestry. Three-dimensional landmark datasets of radii were superimposed and scaled to a common centroid size using Generalized Procrustes Analysis. Mean centroid sizes of groups were compared using parametric statistics. Morphological variation was assessed using Discriminant Function Analysis, coupled with leave-one-out cross-validation to assess accuracies of sex and ancestry estimation. Covariation with size, age and year-of-birth were assessed through regression analysis. Male radii were larger and more robust than those of females, as generally observed following secondary sexual development. Black individuals had absolutely larger radii but more gracile morphologies, White individuals had both absolutely and proportionately larger radii, and Mixed individuals had the smallest radii but with similar robust morphologies as White individuals. Similar differences were also detected when variation was assessed according to sex and ancestry combined. Covariation was detected with size, age and year-of-birth. Differences between groups were sufficient to allow 70.9% classification accuracy of sex, 80.6% for ancestry, and 77.2% for sex and ancestry combined. These results show that postcranial bones like the radius show sufficient morphological variation to allow some accuracy in the estimation of sex, though variation between population groups tends to overshadow this variation, even when sex and ancestry are simultaneously considered. The greater ancestry variation allows for relatively good accuracies in estimation, even for highly heterogeneous South African groups that are subject to significant genetic and environmental influences.