COS 193-6 - Coevolutionary rescue in disease models: Tasmanian devils (Sarcophilus harrisii) as a case study

Abstract: Emerging infectious diseases (EIDs) are a leading threat to natural populations, and strong selection pressures often mean populations must evolve rapidly or go extinct. Recently, a growing body of theory and empirical studies predict when and how evolutionary rescue may occur in host-pathogen systems. Despite the well-developed literature on host-pathogen coevolution, these studies have focused primarily on rapid evolution in the host species. Here, we develop a theoretical model of coevolution rescue in host-disease systems. As a case study, we focus on the joint eco-coevolutionary response of Tasmanian devils to the Tasmanian devil facial tumor disease (DFTD), a lethal EID that has caused dramatic population declines. Genomic studies have found substantial genetic variation among both devils and tumors for disease traits such as force of infection and disease-related mortality, raising the possibility of rapid coevolution in this system. We use an age-structured individual-based model of a devil population to examine the effect of rapid coevolution on the likelihood of devil persistence and tumor extirpation. We simulate devil-tumor eco-coevolutionary dynamics over 100-years following the introduction of DFTD, using parameters that produce devil age-structures and disease dynamics consistent with those observed in natural populations. We examine rapid coevolution in three disease-related traits: tumor transmissibility / devil susceptibility to infection, tumor growth rate / devil suppression thereof, and tumor virulence / devil tolerance. We show that coevolution in transmissibility / susceptibility leads to an evolutionary arms race that results in tumor extirpation if the devil “wins” the arms race and extinction if the devil “loses”. When there is coevolution of tumor growth the results are more complex; at moderate levels of devil genetic variation coevolution increases tumor prevalence and results in devil-tumor coexistence, while at higher levels of genetic variation it results in tumor extirpation and population recovery. Finally, while coevolution in virulence / tolerance results devil-tumor coexistence and high levels of devil genetic variance, it results in increased extinction risk at intermediate levels of devil genetic variance by increasing tumor prevalence without sufficiently increasing the lifetime reproductive success of infected devils. Thus, our results demonstrate that coevolution significantly complicates the process of evolutionary rescue in host-disease systems.