Testing for changes in rate of evolution and position of the climatic niche of clades

There is solid recognition that phylogenetic effects must be acknowledged to appreciate climatic niche variability among species clades properly. Yet, most currently available methods either work at the intra-specific level (hence they ignore phylogeny) or rely on the Brownian motion model of evolution to estimate phylogenetic effects on climatic niche variation. The Brownian motion model may be inappropriate to describe niche evolution in several cases, and even a significant phylogenetic signal in climatic variables does not indicate that the effect of shared ancestry was relevant to niche evolution. We introduce a new phylogenetic comparative method which describes significant changes in the width and position of the climatic niche at the inter-specific (clade) level, while making no a priori assumption about how niche evolution took place. We devised the R function phylo.niche.shift to estimate whether the climatic niches of individual clades in the tree are either wider or narrower than expected, and whether the niche occupies unexpected climates. We tested phylo.niche.shift on realistic virtual species' distribution patterns applied to a phylogeny of 365 extant primate species. We demonstrate via simulations that the new method is fast and accurate under widely different climatic niche evolution scenarios. phylo.niche.shift showed that the capuchin monkeys and langurs occupy much wider, and prosimian much narrower, climatic niche space than expected by their phylogenetic positions. phylo.niche.shift may help to improve research on niche evolution by allowing researchers to test specific hypotheses on the factors affecting clades' realised niche width and position, and the potential effects of climate change on species' distributions.