Drivers of the likelihood of evolutionary rescue in response to novel infectious diseases

The introduction of novel infectious diseases has caused significant declines in many species, and in some cases, local extinction. Simultaneously, hosts might adapt to these novel pathogens, allowing population recovery, an example of evolutionary rescue (ER). Novel infectious diseases differ from sudden, abiotic stressors in that disease only impacts a subset of individuals within the population and individuals might be affected for a finite period. Whether ER occurs in an infectious disease context, and how likely it is relative to abiotic stress, might depend on both host and pathogen characteristics such as the scale of transmission (e.g., local density-dependent transmission only or additional infection through an external reservoir), the degree of fatality or immunity, and the adaptive pathway in terms of whether adaptive alleles reduce mortality, reduce likelihood of infection, or enhance recovery. Here, we use a stochastic mathematical model of host-pathogen dynamics with evolution of a novel allele conferring disease robustness to quantify how population trajectories, including the probability of and time to ER, vary across adaptive pathways and disease characteristics. We find that ER is more likely to occur with dynamical disease transmission compared to a lasting stressor that affects the entire population, indicating that protecting a subset of the population from stress facilitates ER. Among cases with disease transmission, ER is generally more likely to occur with the adaptation of mortality-blocking alleles and sources of transmission that include an external reservoir.