Correcting for Body Size Variation in Morphometric Analysis

Abstract

Using an allometric growth model to correct for body size variation has been known for many decades to be superior to several other widely used methods such as ratios, analysis of covariance, principal components analysis, and residual analysis. However, this technique remains relatively obscure and rarely applied. We optimize the implementation of this method through a newly developed and easy-to-use R package GroupStruct and use empirical datasets to test its relative efficacy compared to several commonly used methods. Our results demonstrate the superiority of the allometric method and highlights the negative impacts of applying improper body size correction methods.