Dispersal is a major driver of the latitudinal diversity gradient of C arnivora

Abstract

Aim Understanding the relative contribution of diversification rates (speciation and extinction) and dispersal in the formation of the latitudinal diversity gradient – the decrease in species richness with increasing latitude – is a main goal of biogeography. The mammalian order C arnivora, which comprises 286 species, displays the traditional latitudinal diversity gradient seen in almost all mammalian orders. Yet the processes driving high species richness in the tropics may be fundamentally different in this group from that in other mammalian groups. Indeed, a recent study suggested that in Carnivora, unlike in all other major mammalian orders, net diversification rates are not higher in the tropics than in temperate regions. Our goal was thus to understand the reasons why there are more species of C arnivora in the tropics. Location World‐wide. Methods We reconstructed the biogeographical history of C arnivora using a time‐calibrated phylogeny of the clade comprising all terrestrial species and dispersal–extinction–cladogenesis models. We also analysed a fossil dataset of carnivoran genera to examine how the latitudinal distribution of C arnivora varied through time. Results Our biogeographical analyses suggest that C arnivora originated in the East Palaearctic (i.e. C entral Asia, C hina) in the early P alaeogene. Multiple independent lineages dispersed to low latitudes following three main paths: toward A frica, toward I ndia/ S outheast A sia and toward S outh A merica via the B ering S trait. These dispersal events were probably associated with local extinctions at high latitudes. Fossil data corroborate a high‐latitude origin of the group, followed by late dispersal events toward lower latitudes in the Neogene. Main conclusions Unlike most other mammalian orders, which originated and diversified at low latitudes and dispersed ‘out of the tropics’, C arnivora originated at high latitudes, and subsequently dispersed southward. Our study provides an example of combining phylogenetic and fossil data to understand the generation and maintenance of global‐scale geographical variations in species richness.

Bibliographic Information