Abstract
Aim Climate and host richness are essential drivers of global gradients in parasite diversity, and the few existing studies on parasites have mostly investigated their effects separately. The advantages of combining these factors into a single analytical framework include unravelling the relative roles of abiotic and biotic drivers of parasite diversity. We compiled a dataset of helminths of amphibians to investigate the direct and indirect effects of temperature seasonality, annual precipitation, precipitation seasonality and host richness as drivers of parasite diversity at the global scale. Our analyses focus not only on the least studied group of vertebrates regarding macroecology of parasite diversity, but also the host group most sensitive to climatic conditions, especially temperature seasonality and water availability. Location Global. Time period 1955–2017. Major taxa studied Helminth parasites of amphibians. Methods We used piecewise structural equation modelling on a global dataset of helminths of amphibians, comprising 613 populations of 319 anuran species and 94 populations of 43 salamander species from 10 zoogeographical realms. Results We found that precipitation seasonality and host richness both affect parasite diversity positively, but the latter presented a stronger effect. Additionally, we found that both temperature seasonality and total precipitation indirectly affected parasite richness through their respective negative and positive effects on host richness. Main conclusions Future studies on global gradients in parasite diversity should include both direct and indirect effects of climatic factors as drivers of parasite diversity. Integrating multiple predictors into a single statistical framework that measures both direct and indirect effects increases our theoretical understanding of the relative importance of and interactions among different diversity drivers at the macroecological scale. The indirect effects of temperature seasonality and total precipitation on parasite diversity are an interesting new insight brought by our study, with implications for future studies dealing with host–parasite coextinctions due to climate change.