Résumé
Outbreaks of generalist pathogens are influenced by host community structure, including population density and species diversity. Within host communities predation can influence pathogen transmission rates, prevalence and impacts. However, the influence of predation on community resilience to outbreaks of generalist pathogens is not fully understood.The role of predation on host community resilience to disease was assessed using an epidemiological multi‐host susceptible‐exposed‐infectious‐recovered model. Sphaerothecum destruens, an emerging fungal‐like generalist pathogen, was used as a model pathogen. Six cyprinid and salmonid fishes, including an asymptomatic carrier, were selected as model hosts that are known to be impacted by S. destruens, and they were used within a model host community.Pathogen release into the host community was via introduction of the asymptomatic carrier. Mortality from infection, pathogen incubation rate, and host recovery rate were set to a range of evidence‐based values in each species and were varied in secondary consumers to predict top‐down effects of infection on the resilience of a host community. Predation pressure within the fish community was varied to test its effects on infection prevalence and host survival in the community.Model predictions suggested that predation of the asymptomatic hosts by fishes in the host community was insufficient to eliminate S. destruens. Sphaerothecum destruens persisted in the community due to its rapid transmission from the asymptomatic host to susceptible host fishes. Following transmission, pathogen prevalence in the community was driven by transmission within and between susceptible host fishes, indicating low host community resilience. However, introducing low densities of a highly specific piscivorous fish into the community to pre‐date asymptomatic hosts could limit pathogen prevalence in the host community, thus increasing resilience.The model predictions indicate that whilst resilience to this generalist pathogen in the host community was low, this could be increased using management interventions. The results suggest that this model has high utility for predicting community resilience to disease and thus can be applied to other generalist parasites to determine risks of disease emergence.