The evolutionary response of virulence to host heterogeneity: A general model with application to myxomatosis in rabbits co-infected with intestinal helminths

Background: Changes in the mean and variance of traits in a host population modify selection pressures on pathogen virulence; increasing heterogeneity (variance) leads to the over- or under-exploitation of a subset of hosts and thus decreases the pathogen’s ability to spread in the population. Objective: To improve the links between theory and data, we develop a model of pathogen evolution in heterogeneous host populations that can be parameterized with a range of empirical measures of host heterogeneity to infection, and which is sufficiently flexible to capture the life history of many natural host–pathogen systems. We use this model to determine whether rabbits co-infected with gastrointestinal helminths could have contributed to the attenuation of the myxoma virus, and might explain differences in the virulence of strains more recently circulating in Australia and Scotland. Methods: We constructed a deterministic model of pathogen transmission and solved it numerically to determine evolutionarily stable strategies with respect to transmission and virulence. Using this model and empirical data from the rabbit–myxoma virus system, we examine how host heterogeneity in co-infection with gastrointestinal helminths affects the severity of an evolved pathogen in a given host type, and to what degree host heterogeneity affects a pathogen’s ability to spread in the host population. Results: Host heterogeneity to infection always decreases pathogen spread and often leads to the under-exploitation of a typical host. However, the specifics are sensitive to: the shape of the distribution describing host heterogeneity, the relationship between virulence and transmission, and the relationship between the heterogeneous host trait and pathogen virulence. In the rabbit–myxoma virus system, gastrointestinal helminths plausibly contributed to the attenuation of the myxoma virus but are unlikely to have contributed to the higher virulence of circulating strains in Scotland relative to those in Australia in the years following the release of the virus.