Identifying hotspots of parasite diversity from species-area relationships: host phylogeny versus host ecology

Interspecific variation in parasite species richness among host species has generated much empirical research. As in comparisons among geographical areas, controlling for variation in host body size is crucial because host size determines resource availability. Recent developments in the use of species-area relationships (SARs) to detect hotspots of biodiversity provide a powerful way to control for host body size, and to identify 'hot' and 'cold hosts' of parasite diversity, i.e. hosts with more or fewer parasites than expected from their size. Applying SAR modelling to six large datasets on parasite species richness in vertebrates, we search for hot and cold hosts and assess the effect of other ecological variables on the probability that a host species is hot/cold taking body size (and sampling effort) into account. Five non-sigmoid SAR models were fitted to the data by optimisation; their relative likelihood was evaluated using the Bayesian information criterion, before deriving an averaged SAR function. Overall, the fit between the five SAR models and the actual data was poor; there was substantial uncertainty surrounding the fitted models, and the best model differed among the six datasets. These results show that host body size is not a strong or consistent determinant of parasite species richness across taxa. Hotspots were defined as host species lying above the upper limit of the 80% confidence interval of the averaged SAR, and coldspots as species lying below its lower limit. Our analyses revealed (1) no apparent effect of specific ecological factors (i.e. water temperature, mean depth range, latitude or population density) on the likelihood of a host species being a hot or coldspot; (2) evidence of phylogenetic clustering, i.e. hosts from certain families are more likely to be hotspots (or coldspots) than other species, independently of body size. These findings suggest that host phylogeny may sometimes outweigh specific host ecological traits as a predictor of whether or not a host species harbours more (or fewer) parasite species than expected for its size.