Population genetics, fragmentation and plague in black-tailed prairie dogs (Cynomys ludovicianus)

Abstract

Black-tailed prairie dogs (Cynomys ludovicianus) have declined dramatically over the past century from habitat loss, poisoning, and plague, a disease caused by the bacterium Yersinia pestis. Because of these factors, populations have become fragmented, and, in areas with plague, exist in metapopulations, undergoing regular extinction and recolonization. The focus of this study was two primary sources of decline in black-tailed prairie dogs: habitat fragmentation and plague. First, the genetic consequences of fragmentation from habitat loss and metapopulation dynamics resulting from plague were examined using six microsatellite loci and a 454 bp portion of the mitochondrial DNA control region. Prairie dog towns were sampled in 2000-2001 in two fragmented areas with plague in Colorado, in a naturally fragmented area without plague in South Dakota, and in a relatively unfragmented area without plague in South Dakota. To document temporal changes in genetic variability, prairie dogs were sampled on the Central Plains Experimental Range, Colorado in 1997, 2000 and 2001. This analysis found high levels of gene flow between populations, but that gene flow was inversely related to levels of landscape fragmentation. There was significant isolation by distance in the three fragmented sites, leading to the conclusion that conservation of genetic variability in fragmented populations of prairie dogs with and without plague should emphasize the maintenance of gene flow between populations. Second, using long-term (1981-2005) monitoring of prairie dog towns on the Pawnee National Grasslands, Colorado, logistic regression models were fitted to examine factors influencing plague outbreaks, focusing on intertown connectivity, soils, and climate. The data revealed little spatial or temporal autocorrelation, but plague extinctions were spatially clustered and spatial dependence could be modeled by town connectivity. High summer temperatures were associated with reduced risk of plague. In spatial logit association models, increased summer rainfall raised the risk of plague events. However, plague outbreaks were not significantly linked with past precipitation in any models tested. As such, our models support the concept that plague occurrence depends upon direct climatic effects and not on a trophic cascade from lagged precipitation.