Understanding how diversity is maintained in natural populations is a major goal of evolutionary biology. In coevolving hosts and parasites, negative frequency‐dependent selection is one mechanism predicted to maintain genetic variation. While much is known about host diversity, parasite diversity remains understudied in coevolutionary research. Here, we survey natural diversity in a bacterial parasite by characterizing infection phenotypes for over 50 isolates in relation to 12 genotypes of their host, Daphnia magna. We find striking phenotypic variation among parasite isolates, and we discover the parasite can infect its host through at least five different attachment sites. Variation in attachment success at each site is explained to varying degrees by host and parasite genotypes. A spatial correlation analysis showed that infectivity of different isolates does not correlate with geographic distance, meaning isolates from widespread populations are equally able to infect the host. Overall, our results reveal that infection phenotypes of this parasite are highly diverse. Our results are consistent with the prediction that under Red Queen coevolutionary dynamics both the host and the parasite should show high genetic diversity for traits of functional importance in their interactions.

Knowing the determinants of the geographic ranges of parasites is important for understanding their evolutionary ecology, epidemiology and their potential to expand their range. Here we explore the determinants of geographic range in the peculiar case of a parasite species - the microsporidian Hamiltosporidium tvaerminnensis - that has a limited geographic distribution in a wide-spread host - Daphnia magna. We conducted a quantitative trait loci (QTLs) analysis with monoclonal F2 D. magna populations originating from a cross between a susceptible northern European genotype and a resistant central European genotype. Contrary to our expectations, long-term persistence turned out to be a quantitative trait across the F2 offspring. Evidence for two QTLs, one epistatic interaction and for further minor QTL was found. This finding contrasts markedly with the previously described bimodal pattern for long-term parasite persistence in natural host genotypes across Europe and leaves open the question of how a quantitative genetic trait could determine the disjunct geographic distribution of the parasite across Europe.