Morphological size and shape variation among local human populations

Abstract

A common method in morphometric studies is the partitioning of biological distance into components representing size and shape (usually defined as relative size). This is a critical procedure when samples from different species are compared since size differences obscure shape relationships, and may thus confuse taxonomic interpretation. It has been suggested that for analyses within a single species, shape variation contributes more than size variation to among-group distances. This hypothesis is tested using data on 14 anthropometric variables for 347 adult males and 261 adult females from 12 towns in western Ireland. For each sex, Mahalanobis' generalized distance (D2) was computed between each pair of towns. Spielman's (1973) method was used to partition each inter-town anthropometric distance into a size component and a shape component. For both sexes, over 93% of the average anthropometric distance between towns is due to shape variation. Rank order correlations show that shape and overall distance (D2) are highly correlated in both sexes (rs=0·99). These results support the hypothesis that local interpopulational variation within a species is influenced more by shape variation than size variation.