Triangulation-Measurement Schemes in the Multivariate Analysis of Size and Shape

Publisher Website
Google Scholar
Abstract
Skulls of seven rodent species were measured by a triangulation-measurement scheme and subjected to multivariate analysis. Results were compared to studies from the literature based on traditional measurements for the proportion of variance explained by overall body size, the number of shape factors, and the ease of interpretation of these factors. Principal-components analysis of triangulation measurements showed that size variation accounted for less than half of the variance within samples of Akodon, Gerbillurus, Chaetodipus, and Dipodomys. In contrast, size accounts for more than half of the variance in most studies in which traditional skull measurements were analyzed. Triangulation measurements appear to capture more information about variation in shape than do traditional measurements. In addition, shape factors extracted from these species can be interpreted in terms of variation in spsecific regions of the skull. Regional patterns of covariation maybe more amenable to explanations involving underlying biological processes than are results of multivariate analysis of more traditional measurements.