Feral Horse Ecology in the Rocky Mountain Foothills of Alberta, Canada.
Feral horses have been present in western Canada since the early 1720s, though little is known about fundamental components of their ecology. In the foothills of the Rocky Mountains in Alberta, Canada, increasing socio-political conflict regarding feral horse management, their role, and their impacts in the Foothills ecosystem highlights a growing need for nuanced management approaches predicated on robust ecological information. My objective was to assess several key aspects of feral horse ecology within the largest known population of horses in western Canada, located west of the township of Sundre, Alberta. In Chapter 1, I provide a general introduction to the history and ecology of feral horses in western Canada and North America, to the study area, and to the key questions motivating the research. In Chapter 2, I begin by reviewing the causes and consequences of a lack of robust ecological information for many feral horse populations, and feral wildlife generally, and found that both political and biological asymmetries in feral populations can favor the persistence and expansion of populations of feral species relative to native species. Ideological opposition to the necessary study of feral populations as wildlife also has important implications for increasingly contested population management of feral species. In Chapter 3 I assessed key spatial and social characteristics of the population within the Sundre equine management zone (EMZ) using a combination of GPS telemetry and camera-trap data. Social characteristics such as band size and sex-ratio were consistent with earlier work and with feral horse populations generally. Home-ranges overlapped considerably between individual bands with sizes ranging from 47.6 to 93.0 km2. These were larger than those that were identified by Salter in 1978 within an area now encompassed in the Sundre EMZ, though were similar to those identified by Girard in nearby Bragg Creek, Alberta with the difference likely representing differences in methodology. Movement and detection rates of horses were greater in summer, and horses showed reduced diel activity in the middle of the day in summer, compared to winter. The number of foals detected, and the mean number of foals within each band was highest from June to August and composed less than 15% of the total individuals detected each year. Identification of individuals based on natural markings was possible, though the influence of non-independent movement and social grouping was evident, and problematic for capture-recapture analyses. In Chapter 4, I assessed density and total abundance of the population of horses in the Sundre equine management zone using space-to-event (STE) density models. Feral horse density was 0.602/km2, 0.606/km2 and 0.522/km2 respectively, from 2017 – 2019. Total abundance estimates were similar to minimum aerial count data with confidence intervals from estimates overlapping aerial counts in all years except 2017, suggesting the population declined by approximately 14% from 2018 to 2019. These results are similar to recent trends detected in minimum aerial counts that show a decline in counts of approximately 22% (2019 – 2021) and contrast with expectations based on increasing minimum counts in the Sundre population in recent decades, and populations of many other free-ranging feral horse populations. Reasons for the decline are likely to be multi-factorial and are difficult to determine based on the current data alone. I discuss potential causes and the implications these have for future management. In Chapter 5, I assessed habitat selection of feral horses with respect to several key landscape and vegetation characteristics of interest using GPS telemetry data, and compared summer occupancy of feral horses, cattle, and elk using camera-trap data. Variation was high among tracked individuals in selection for vegetation type, and counter to expectations, horses avoided native rangeland in summer, compared to greater selection of forestry cutblocks in all but one individual. This was also supported by higher summer horse occupancy probability with increasing areal coverage of cutblocks. This contrasted to cattle occupancy which declined strongly with increasing cutblock coverage. Cattle occupancy was also negatively influenced by terrain, though positively influenced by the presence of linear features and lower distance to roads. These results have important implications for the spatio-temporal partitioning of cattle and horses over summer and suggest that horses may avoid areas with high overlap with cattle. In Chapter 6, I summarize the findings from each chapter, and discuss their wider management implications.
feral horses, ecology
Doctor of Philosophy (Ph.D.)