Résumé
Aim Geographical gradients of vertebrate species richness are determined jointly by evolutionary and environmental factors. Support for these factors comes mostly from tetrapods and, recently, marine fishes, but their validity and relative importance in freshwater fishes is not well understood. Here, we describe the species richness gradient for the major component of the viviparous freshwater fishes of the Americas, the poeciliids, and test the effects of evolutionary and environmental factors on this pattern. Location The Americas. Time period Approximatley 56–0 Ma. Major taxa studied Poeciliidae (Teleostei: Cyprinodontiformes). Methods We constructed geographical ranges for 93% (256) of poeciliids to describe and evaluate their species richness gradient. Evolutionary factors (evolutionary time and speciation rate) were derived from a recent phylogeny. Environmental factors were represented as basin area, topographical heterogeneity, energy, climate seasonality and past climatic stability. We tested the influence of these factors with a piecewise structural equation model (pSEM). Results The distribution of Poeciliidae is biased to the Atlantic coast, with species richness showing a bimodal latitudinal gradient, peaking in middle latitudes near the Tropics of Cancer and Capricorn, and exhibiting the highest richness in Middle America. The pSEM showed that this species richness pattern was influenced positively by evolutionary time and past climatic stability and negatively by climate seasonality. Main conclusion The species richness gradient of Poeciliidae has been shaped by the interplay of evolutionary time in addition to current and historical climate. Indeed, regions with high poeciliid richness were those containing by ancient lineages, supporting the time‐for‐speciation effect, and that have experienced low historical stability in temperature and currently show low temperature seasonality. Conversely, species‐poor regions contained younger lineages and experienced greater temperature seasonality. Our study highlights the need to assess jointly the evolutionary, historical and climatic drivers of species richness in order to unravel the causes of diversity gradients.