Variability in the Selection Patterns of Pronghorn: Are they Really Native Prairie Obligates?

In Canada, pronghorn (Antilocapra americana) are primarily considered a native prairie obligate because of their reliance on open grassland vegetation communities, although an assessment of local ecological knowledge suggests that pronghorn in Alberta select a variety of habitat from native prairie to cultivated lands. The primary objective of our study was to assess whether pronghorn in Alberta and Saskatchewan are native prairie obligates. Specifically, we addressed the following questions: 1) do individual pronghorn show similar selection patterns for native prairie and, therefore, support the notion that they are prairie obligates; 2) do pronghorn show consistent resource selection patterns at multiple scales (landscape and within-seasonal range); and 3) to what extent are selection patterns of pronghorn influenced by highways and roads. Within Alberta, we captured, collared, and monitored for one year individual female pronghorn in December 2003 (n = 24), March 2005 (n = 25), and March 2006 (n = 25). A detrended correspondence analysis of patterns of habitat selection revealed three distinct groups of pronghorn (r2 = 0.96, n = 55) that we labeled native, cultivated, and mixed, referring to the dominant land cover in their parturition ranges. We used logistic regression to model resource selection patterns of the three groups of pronghorn during the parturition and winter periods at the landscape and within-seasonal range scales. At the landscape scale, each group of pronghorn had top models consisting of the variables land cover, landform, distance to express highways, distance to arterial roads, and distance to collector roads for both periods. The native and mixed groups were less likely to use annual and perennial cropland than native prairie habitats, whereas the cultivated group was more likely to use annual and perennial cropland. At the within-seasonal range scale, the top models for each group in both seasons consisted of one or more road variables, but the top models exhibited poor model fit. Our results do not show a clear association for native prairie, which we would have expected if pronghorn were native prairie obligates, suggestive of plasticity in behavior. We acknowledge that patterns of habitat selection do not indicate habitat quality or fitness; therefore, to understand the individual- and population-level consequences of selecting sub-optimal habitats, such as agricultural landscapes, further research is needed.