Abstract
Parasites can represent formidable selection pressures for hosts, but the cost of infection is sometimes difficult to demonstrate in natural populations. While parasite exploitation strategies may, in some instances, actually inflict low costs on their hosts, the response of hosts to infection is also likely to determine whether or not these costs can be detected. Indeed, costs of infection may be obscured if infected individuals in the wild are those that are the most tolerant, rather than the most susceptible, to infection. Here we test this hypothesis in two natural populations of Anolis sagrei, one of the most common anole lizard of the Bahamas. Plasmodium parasites were detected in > 7% of individuals and belonged to two distinct clades: P. mexicanum and P. floriensis. Infected individuals displayed greater body condition than non-infected ones and we found no association between infection status, stamina, and survival to the end of the breeding season. Furthermore, we found no significant difference in the immuno-competence (measured as a response to phytohemagglutinin challenge) of infected versus non-infected individuals. Taken together, our results suggest that the infected individuals that are caught in the wild are those most able to withstand the cost of the infection and that susceptible, infected individuals have been removed from the population (i.e., through disease-induced mortality). This study highlights the need for caution when interpreting estimates of infection costs in natural populations, as costs may appear low either when parasites exploitation strategies truly inflict low costs on their hosts or when those costs are so high that susceptible hosts are removed from the population.