A general framework for modeling pathogen transmission in co-roosting host communities

Cross-species transmission of pathogens can be facilitated by frequent contact among wildlife. Cross-species transmission is often driven by phylogenetic similarity between host species, but the role this plays when multiple host species co-roost is unknown. We developed a generalizable framework for understanding how cross-species transmission is driven by contact among co-roosting species spanning evolutionary similarities and the net impact on roost-level infection prevalence. We developed ordinary differential equation models describing population and infection dynamics between two and three co-roosting species. We derived conditions for pathogen invasion and parameterized models using co-roosting Neotropical bat systems, with interspecific transmission exponentially declining with phylogenetic distance. To assess the relative contribution of contact rates and phylogenetic similarity, we co-varied intraspecific transmission rates and phylogenetic distances while considering sensitivity to epidemiological structure and pathogen traits. For both susceptible-infected-recovered-susceptible and susceptible-infected-latent-infected models, we show that relatedness between co-roosting hosts facilitates pathogen invasion, particularly for poorly transmissible pathogens with short durations of infection and immunity or latency. These models converged on similar equilibria, and roost-level prevalence was greatest when hosts were most closely related. However, we also identified regions of parameter space where roost-level prevalence increased when hosts were distantly related. Our generalizable models are adaptable to other co-roosting systems with low-virulence pathogens that are directly transmitted and inform our understanding of pathogen spillover.