What determines the climatic niche width of species? The role of spatial and temporal climatic variation in three vertebrate clades.

Aim Climatic niche breadth (the range of climatic conditions that a species experiences over space and time) is a fundamental topic in ecology, biogeography and evolution. But what determines the climatic niche width of species? In 1967, Janzen suggested that climatic niche widths for temperature were determined by levels of seasonal fluctuation in temperature at each locality, such that niche breadths are narrow in tropical species and broad in temperate species. However, it is unclear whether climatic niche breadths of species are determined more by seasonal variability within sites as opposed to climatic variation between sites across the species' range. We address this question here. Location Global. Methods We analysed three vertebrate clades (plethodontid salamanders, hylid frogs and phrynosomatid lizards) for which we had phylogenetic information and climatic data from localities throughout each species' geographic range, collectively including 409 species. We tested how climatic niche breadths of localities (i.e. temporal variation) are related to overall species climatic niche breadths (i.e. temporal and spatial variation) using phylogenetic comparative methods, focusing both on temperature extremes and precipitation. Results Across the three clades, we find that niche breadths for single localities generally span most of the species' climatic niche breadth, and are strongly correlated with overall species niche breadths. However, species with wider climatic niches also tend to show greater climatic divergence between localities. Main conclusions The extent to which the climatic niche breadths of species are determined by variation within localities versus spatial variation between localities has been largely unexplored. Our results suggest that within-locality seasonal variation explains most variation in climatic niche breadths among species. However, between-locality variation and local adaptation may also play some role. These results require more general testing, but have several important implications. [ABSTRACT FROM AUTHOR]