Canonical trend surface analysis of morphometric variation in Africanized honey bees from Brazil

Canonical trend surface analysis (CTSA) is a powerful multivariate statistical method that finds linear combinations of morphological data which have maximum correlation with geographic space. In this paper, CTSA was applied to morphometric variation among 42 Africanized honey bee populations from Brazil. Results were compared with principal component analysis (PCA) and matrix comparison between D-2 Mahalanobis and geographic distances, using the Mantel test. CTSA can be more effective than other multivariate techniques in describing multivariate spatial patterns and, consequently, in suggesting evolutionary explanations for their origin.