Enzootic maintenance of sylvatic plague in Canada's threatened black-tailed prairie dog ecosystem

Data paucity can seem to hinder science-based approaches to the conservation of imperiled species. Yet, even individually limited datasets can improve understanding and management of complex ecological systems when carefully integrated. We demonstrate this approach to gain first insights on the transmission ecology of Yersinia pestis in Grasslands National Park (GNP), Canada, where both the bacterium and its rodent host, the nationally threatened black-tailed prairie dog (BTPD, Cynomys ludovicianus), reach the northern limit of their distribution in North America. Primarily flea-borne, Y. pestis causes sylvatic plague, a disease of exceptional relevance to both human health and wildlife conservation. We integrated data collected independently by multiple organizations in 2010-2017 across 17 BTPD colonies, where the species co-occur with Richardson's ground squirrels (RGS, Urocitellus richardsonii). Available data included estimates of BTPD density and occupancy from visual counts and colony mapping; information on flea distribution, abundance, and prevalence of infection with Y. pestis from burrow swabbing, animal combing, and PCR assays; and the response of these variables to deltamethrin application on BTPD colony sections. Our analyses suggest that sylvatic plague in GNP is maintained at an enzootic level (i.e., chronic presence affecting a low proportion of individuals) with no evidence of widespread mortality, at least partially due to reduced flea activity after spring (percentage of prevalence in burrows: April-May = 11.69-33.89%; June-September: 1.75-3.19%), low prevalence of Y. pestis in flea samples (95% CI = 0.42-2.27%), and relatively low BTPD densities. Nonetheless, reducing flea prevalence through insecticide application had a positive effect on BTPD abundance, suggesting that enzootic plague is causing chronic mortality. Because flea prevalence on hosts was higher following drier years and higher on RGS than on BTPD (26.69% vs. 3.27%), insecticide application may be particularly important during dry periods and may need to take RGS and their movements into consideration. Differences between flea communities sampled by burrow swabbing and host combing suggest that plague surveillance should integrate both methods. Effects of projected climate change on vector life cycles, flea community composition, and host-parasite interactions warrant continued monitoring and an adaptive approach to species recovery actions and plague mitigation measures.