The geographic distributions of highly visible species, such as large mammals, are usually well known, but less is known about the distributions of cryptic or nocturnal species that vector pathogens. Many species of bat are... [ view full abstract ]
The geographic distributions of highly visible species, such as large mammals, are usually well known, but less is known about the distributions of cryptic or nocturnal species that vector pathogens. Many species of bat are small, reclusive, and rarely observable during daytime. Species distribution models (SDMs) have been used to suggest sites with high potential for occurrence of rare or cryptic species, and for development of disease management plans. We analyzed the distributions of 2 North American bat species that vector rabies virus, each with different geographic distributions and natural histories: hoary bats (Lasiurus cinereus; N=2,753 occurrence records), which roost individually in trees, undergo continental-scale migrations, and are occasionally encountered by humans; and vampire bats (Desmodus rotundus; N=7,094 occurrence records), which can roost in large congregations, do not migrate, and can pass rabies virus to cattle, causing substantial impacts to agricultural economies. We analyzed and mapped the distributions of these species using 5 approaches to species distribution modeling: logistic regression, multivariate adaptive regression splines, boosted regression trees, random forest, and maximum entropy. SDM performance using standard metrics was higher for vampire bats (AUC = 0.94, sensitivity = 0.91, specificity = 0.85) than for hoary bats (AUC = 0.88, sensitivity = 0.75, specificity = 0.83). For hoary bats, seasonally-dynamic SDMs suggested that hoary bats winter in regions with relatively long growing seasons where temperatures are moderated by proximity to oceans, then move to the continental interior for summer. For vampire bats, SDMs suggested that the northern portion of the distribution is constrained by winter temperatures, and that the northern edge of the distribution may gradually move northward under the influence of future climate conditions. These results demonstrate how species distribution models and maps can help generate insights into the current and future distributions of wildlife species that vector pathogens.
Topics: Infectious Disease , Topics: Terrestrial Mammals