The epidemiology of bovine tuberculosis (Mycobacterium bovis) in the Greater Riding Mountain Ecosystem

The overall objective of this thesis is to provide an enhanced understanding of the epidemiology of Mycobacterium bovis in the Greater Riding Mountain Ecosystem (GRME) and provide a scientific basis for disease management from a systems perspective now and into the future. M. bovis prevalence has been consistently higher in elk compared to white-tailed deer, and higher within a defined Core area compared to areas outside. Prevalence in both species declined significantly between 2003 and 2013. Only one infected elk was detected in 2013; the last infected white-tailed deer was detected in 2009 and the last infected cattle herd was detected in 2008. Parallel interpretation of three blood-based assays resulted in effective selective culling of elk within Riding Mountain National Park (RMNP) with predictive value negative of 100%. A lymphocyte stimulation test (LST) was the most sensitive single blood-based assay, but was difficult to perform under field conditions. Combinations of humoral antibody tests and cell-mediated tests performed better than any single test, likely detecting the broad spectrum of host pathology present. Seven of 14 risk factors were identified for wild cervids testing culture positive with the three being most strongly associated with culture positivity being geographical location (within core area), elk density and year category (sampling phase). Age, sex, and surveillance method were also significant factors, but species was not. A rapid decline in elk density in combination with fencing of hay storage yard and non-selective culling were likely key factors resulting in the M. bovis prevalence decline observed in elk, and an overall decline in prevalence from 1997 for both species. Elk were the primary reservoir species in this episystem, but are now considered a spillover host, while white-tailed deer have always been a spillover host due to lower densities and shorter life expectancy. Very limited strain diversity exists within the GRME with one spoligotype restricted to cattle and associated with a limited outbreak in five herds in the early 1990’s, and three other shared strains between cattle and wildlife. A single monomorphic type was present in white-tailed deer. Significant spatial overlap of wildlife and cattle isolates delineated a core area where management activities are now focused. The relative simplicity of this episystem has allowed significant progress on control and management to be achieved, despite being located within a national park. Wildlife surveillance will need to continue until at least 2022 in order to achieve a 95% probability of freedom using three different surveillance streams. Latent cases are likely to be extremely rare in future and unlikely to result in ongoing transmission as the factors that created this wildlife reservoir no longer exist. Wild cervids should not be considered ideal maintenance hosts for M. bovis in North America but rather facultative hosts; acting either as a reservoir or spillover host dependent on regional/local density and presence/absence of baiting and feeding