Mapping elk distribution on the Canadian prairies: Applying local knowledge to support conservation
Once abundant across the Great Plains of North America, prairie-parkland elk (Cervus canadensis manitobensis) underwent a catastrophic population collapse and dramatic contraction of their overall range through the late 1800’s and the early 1900’s due to habitat loss (primarily from agricultural expansion) and unregulated hunting. Populations were able to recover in some areas following new hunting regulations and the establishment of protected areas. Prior to this study, the current distribution of prairie parkland elk was poorly understood, though it was established that they were largely relegated to large protected areas and made use of adjacent agriculture dominated landscapes. In order to effectively manage prairie-parkland elk so populations remain resilient to ongoing habitat loss, population reduction and disease risks, detailed mapping of their range and an understanding of the environmental factors most important to elk is essential. The purpose of my thesis is to characterize elk distribution and resource selection patterns in the prairie-parkland of Saskatchewan and Manitoba, Canada, at a landscape scale and assess the implications of distribution for species persistence, by using both local ecological knowledge (LEK) and biological research techniques in tandem. Integrating LEK with more conventional biological research can provide complementary data at contrasting time and spatial scales and facilitates comparison of multiple independent datasets. Furthermore, LEK research creates important opportunities to engage stakeholders in contributing knowledge and may facilitate relationships and contribute toward more effective resource management. I used three sets of biotelemetry-collar data from across Saskatchewan and Manitoba (n = 328 collared elk; 1998–2012), in conjunction with LEK from hunters, biologists and enforcement officers (n = 71 participants) to create a series of resource selection functions (RSFs) characterizing elk distribution across Saskatchewan and Manitoba. I gathered LEK in workshops where participatory mapping was conducted with local experts across the study area. The RSF analysis determined that prairie-parkland elk selected locations close to protected areas and far from high road density. Elk also selected areas with moderate amounts of mixed-wood and deciduous forests and herbaceous vegetation. Models developed with LEK only, biotelemetry collar data only and a combined dataset were all validated against an independent dataset of elk crop damage locations. All models predicted elk presence well. The RSF scores of the LEK only and radio-collar only models were not significantly different. Successful conservation management requires identifying which areas are most important to a species, and assessing if these areas are vulnerable to threats, as well as balancing human resource needs. Using the RSF-based maps of prairie-parkland elk distribution, I identified locations of high quality habitat (top 10% of RSF values) and determined which of those areas were vulnerable to agricultural expansion, forestry, disease, and hunting. I identified 81 high quality habitat areas with a combined total area of 30 753 km2. One or more vulnerability factor impacted 87% of the identified core area. High quality areas were clearly clustered around the boreal-prairie transition zone and large protected areas. The majority (88%) of high quality core areas were located within protected areas. A connectivity analysis using least cost path analysis determined that core habitat areas endemic with chronic wasting disease (CWD) in wild elk are highly connected to other high quality habitat areas. CWD thus has the potential to reduce cervid populations within the study area. My thesis results highlight that prairie-parkland elk populations in Saskatchewan and Manitoba remain threatened by habitat loss and disease, and emphasize the need for habitat conservation to prevent further population reduction. While elk populations have regained a small fraction of the range lost at the turn of the last century, they have not been able to re-establish with much success in the prairie portion of their range. I also determined that LEK can be as effective as conventional biological research approaches to develop RSFs. I also confirmed that stakeholders within the study are knowledgeable about elk behaviour. The distribution maps and identified areas of priority concern created in this thesis can provide important insights to support the management and maintenance of abundant elk populations.
LEK, elk, prairie-parkland, conservation, RSF, habitat selection
Master of Science (M.Sc.)
Animal and Poultry Science