High annual survival in infected wildlife populations may veil a persistent extinction risk from disease

Host response to emerging pathogens is variable, causing uncertainty about population-level impacts and challenging effective disease management. White-nose syndrome (WNS) has caused catastrophic declines in some bat species, while others appear less impacted. Developing predictive models based on observed survival patterns can generate testable hypotheses about mechanisms driving population dynamics and contribute to the development of targeted approaches to disease management. We conducted a mark–recapture study of federally endangered Indiana bats (Myotis sodalis) during 2011–2016. Annual survival decreased from 0.78 (95% CI: 0.59, 0.89) and 0.79 (95% CI: 0.70, 0.86) for females and males, respectively, in 2011 to 0.74 (95% CI: 0.33, 0.94) and 0.75 (95% CI: 0.53, 0.89) for females and males, respectively, in 2015. We then modeled two explanatory mechanisms potentially driving the observed patterns: (1) phased exposure to disease through the spatial spread of the pathogen within the hibernaculum; and (2) cumulative mortality risk from iterative yearly WNS infection. Under a phased exposure scenario, models suggest that infected individuals have an average survival probability of 0.68, and disease prevalence is predicted to reach 100% within 9 yr of disease emergence. Under the cumulative mortality risk hypothesis, survival probability of individuals decreases with each infection cycle. In either case, infected populations are predicted to stabilize at a negative growth rate. Results suggest that Indiana bats tolerate a pathogen load prior to onset of infection, leading to a less pronounced population decline than for other susceptible species. However, the long-term risk of WNS to Indiana bats may be more severe than current population trends suggest. To inform current conservation management, we performed a vital rate sensitivity analysis, which suggested that modest increases in survival (4–5%) through targeted intervention may return declining populations to stability (λ = 1.0). Demographic modeling approaches coupled with continued population monitoring can highlight important differences in disease response, and ultimately extinction risk, in host species allowing conservation practitioners to tailor intervention actions so that they will be most effective.