Aim Climatic factors are known to influence species distributions. However, elucidating
the underlying mechanisms is challenging because direct and indirect effects of climatic and non-climatic factors are correlated. In the absence of this covariation and at fine-grain resolutions the direct effect of climate via physiological constraints should be stronger on the distributions of ectothermic organisms.
So far, no comprehensive study has explicitly tested the influence of climate on species distributions by quantitatively comparing ectothermic and endothermic vertebrates.
Location Peninsular Spain.
Methods Presence-absence data of native terrestrial vertebrates in Peninsular Spain were modelled using generalized additivemodels to disentangle the influence of climate and other contemporary correlated factors (topography and plant cover). We performed partial regressions to partition the deviance explained by climatic and non-climatic effects into independent and shared components. We compared the independent contributions of climatic and non-climatic effects between ectothermic and endothermic vertebrates, and among mammals, birds, reptiles and amphibians.
Results After reducing the covariation with non-climatic factors, climate explained a greater proportion of deviance in ectotherms than in endotherms. Also, the contribution of temperature was highest for reptiles, and the contribution of precipitation was highest for amphibians, after extracting their overlaps with precipitation and temperature, respectively. The contribution of topography and plant cover remained high for birds after extracting the overlap with climate.
Main conclusions Our results are consistent with the prediction that, at fine resolutions, the direct influence of climate (via physiological constraints) on range distributions is stronger in ectothermic vertebrates. Also, at least for birds, indirect effects of climate (via plant productivity) and other habitat characteristics remained relatively important once their covariation with climate was reduced. This study shows that controlling the direct effects of climate by their indirect effects and/or other correlated factors, combined with comparisons among functional groups, can be a useful approach to elucidate causal links with the spatial
patterns of organisms.