The role of ecology in the geographical separation of blood parasites infecting an insular bird

Aim: Niche modelling is increasingly used to predict species' geographical distributions or to infer the evolutionary or ecological processes that constrain them, but relatively few studies have examined the ecological processes governing the distributions of parasites. Among such processes, niche divergence is frequently invoked to explain species range variation. Here, we test whether the geographical distributions of two lineages of Leucocytozoon (Haemosporida) avian parasites are linked to climatic conditions and whether niche divergence can explain their geographical separation. Location: Réunion, Mascarene archipelago, south-western Indian Ocean. Methods: Leucocytozoon prevalence data were obtained by PCR screening of avian blood samples. Prevalence data and 20 environmental layers were used to build species distribution models (SDMs). SDMs were built by averaging the predictions of five different models: random forests (RF), generalized linear models (GLM), generalized additive models (GAM), multivariate adaptive regression splines (MARS) and support vector machines (SVM). Niche identity and background tests were used to test for a role of niche divergence in explaining parasite distributions. Results: The geographical ranges of the two lineages of Leucocytozoon under study showed little overlap. Species distribution modelling suggested that niche divergence may explain the spatial variation observed in Leucocytozoon distribution, implying that the geographical separation of parasites is linked to environmental conditions. The variables that best explained parasite distribution were all related to precipitation patterns. Main conclusions: As precipitation cannot directly affect endosymbiotic Leucocytozoon parasites, we suggest that the geographical separation of Leucocytozoon lineages is the result of an underlying geographical structure in the dipteran vectors that transmit these parasites. This illustrates the need to consider the ecology of vectors when predicting the distribution of vector-borne parasites. Our study also shows that different parasite lineages, contained within broadly defined parasitic taxa, may have very different ecologies, and that these differences should be taken into account when attempting to understand the ecological determinants of parasite distribution and disease emergence.