A cautionary note on some phylogenetic dissimilarity measures

Aims Measures of plot-to-plot phylogenetic dissimilarity and beta diversity are providing a powerful tool for understanding the complex ecological and evolutionary mechanisms that drive community assembly. Methods Here, we review the properties of some previously published dissimilarity measures that are based on minimum or average phylogenetic dissimilarity between species in different plots. Important Findings We first show that some of these measures violate the basic condition that for two identical plots the measures take the value zero. They also violate the condition that the dissimilarity between two identical plots should always be lower than that between two different plots. Such erratic behavior renders these measures unsuitable for measuring plot-to-plot phylogenetic dissimilarity. We next propose a new measure that satisfies these conditions, thus providing a more reasonable way for measuring phylogenetic dissimilarity.