Climatic niche properties shape treefrog diversity.

A suite of ecological hypotheses has been proposed to explain why more species are found in the tropics. Most of these hypotheses consider the geography of climate as a major determinant of richness, while ignoring the relationships between organisms and climate, in other words, their climatic niche. In this study, we evaluate three hypotheses that link species richness to niche properties. The niche breadth hypothesis predicts higher richness where species have narrower niches; the niche marginality hypothesis predicts higher richness where species have less marginal niches; and the niche position hypothesis predicts higher richness where species have niches similar to those of their ancestors. We estimated niche properties for 441 (70%) treefrog species from the Americas using both univariate and multivariate approaches based on minimum volume ellipsoids and projected them onto geography to relate them with the geographical pattern of species richness. We used an assemblage-based approximation to map niche properties under both approaches and performed simultaneous autoregressive models to test our hypotheses. We found support for the niche-breadth hypothesis under both methodological approaches, for the niche position hypothesis under the multivariate approach, and only for precipitation position under the univariate approach. Contrary to expectations under the niche position hypothesis for temperature, we found that treefrog species richness increased with distance to the ancestral niche temperature. We found no support for the niche marginality hypothesis under either approach. In general, our results indicate that places with high richness contain species with narrower niches, closer to their ancestral niche, but far from the mean conditions available in the Americas. These results support the long-standing hypotheses of niche packing and niche conservatism, while suggesting that some niche dimensions are more constrained than others.