Incorporating environmental heterogeneity and observation effort to predict host distribution and viral spillover from a bat reservoir.

Abstract

Predicting the spatial occurrence of wildlife is a major challenge for ecology and management. In Latin America, knowledge of the true number and locations of vampire bat colonies precludes informed allocation of measures intended to limit lethal rabies spillover to humans and livestock. We inferred the complete spatial distribution of vampire bat roosts while accounting for observation effort and environmental covariates by fitting a Log Gaussian Cox Process model to the locations of 563 roosts in three regions of Peru. Our model explained 47% of the variance in the observed roost distribution and identified landscape correlates of roost establishment. Our model estimated that 1,795 roosts (76%) remain undiscovered and identified hotspots of undetected roosts in currently rabies-free areas, implying high risk for viral incursion. Incorporating the locations of undetected roosts improved spatial predictions of rabies spillover to livestock, revealed areas with disproportionate underreporting to surveillance systems, and indicated a higher rabies burden than previously estimated. We provide a robust approach to infer the distribution of a mostly unobserved bat reservoir that can inform strategies to prevent the re-emergence of an important zoonosis.