A consistent occupancy-climate relationship across birds and mammals of the Americas

18 February 2014

Boucher-Lalonde, Véronique; Morin, Antoine; Currie, David

At broad spatial scales, species richness is strongly related to climate. Yet, few ecological studies attempt to identify regularities in the individual species distributions that make up this pattern. Models used to describe species distributions typically model very complex responses to climate. Here, we test whether the variability in the distributions of birds and mammals of the Americas relates to mean annual temperature and precipitation in a simple, consistent way. Specifically, we test if simple mathematical models can predict, as a first approximation, the geographical variation in individual species' probability of occupancy for 3277 non-migratory bird and 1659 mammal species. We find a Gaussian model, where the probability of occupancy of a 104 km2 quadrat decreases symmetrically and gradually around a species 'optimal' temperature and precipitation, was generally the best model, explaining an average of 35% of the deviance in probability of occupancy. The inclusion of additional terms had very small and idiosyncratic effects across species. The Gaussian occupancy-climate relationship appears general among species and taxa and explains nearly as much deviance as complex models including many more parameters. Therefore, we propose that hypotheses aiming to explain the broad-scale distribution of species or species richness must also predict generally Gaussian occupancy-climate relationships.