Human impact modulates chytrid fungus occurrence in amphibians in the Brazilian Atlantic Forest

Abstract

Here, we investigate the influence of scale on different drivers influencing the occurrence of the chytrid fungus, Batrachochytrium dendrobatidis (Bd), in the Atlantic Forest, Brazil. We used gridded values of proxies of the abiotic, biotic and anthropogenic components of landscapes where Bd infects amphibians. Building upon disease prevalence data obtained from a previous work, we fitted GLS multiple regression models using extracted values of the three predictors for each prevalence centroid in space, explicitly controlling for spatial autocorrelation among predictors. To test for the effect of scale on driving the macroecology of Bd infection, we performed tests at different spatial scales. We then used model selection procedures to evaluate the relative contribution of the different predictors on the occurrence of the fungus. The Human Footprint Index better explained a pathogenic species occurrence than largely studied biotic and abiotic factors (i.e., host species distribution and minimum monthly potential evapotranspiration). That effect was, however, not observed at landscape scale, where we found no difference among the relative influence of predictors. Our results indicate that human-mediated impacts on environments can be strong drivers of spread of infectious diseases on native faunas worldwide, thus, suggesting that anthropogenic landscapes may create favourable conditions for the occurrence of this and other infectious diseases.