Uncertainties due to within-species variation in comparative studies: Measurement errors and statistical weights

Abstract

Comparative studies investigating evolutionary questions are generally concerned with interspecific variation of trait values, while variations observed within species are inherently assumed to be unimportant. However, beside measurement errors, several biological mechanisms (such as behaviors that flexibly change within individuals, differences between sexes or other groups of individuals, spatial, or temporal variations across populations of the same species) can generate considerable variation in the focal characters at the within-species level. Such within-species variations can raise uncertainties and biases in parameter estimates, especially when the data are hierarchically structured along a phylogeny, thus they require appropriate statistical treatment. This chapter reviews different analytical solutions that have been recently developed to account for the unwanted effect of within-species variation. However, I will also emphasize that within-species variation should not necessarily be regarded as a confounder, but in some cases, it can be subject to evolutionary forces and delineate interesting biological questions. The argumentation will be accompanied with a detailed practical material that will help users adopt the methodology to the data at hand.