The effects of artificial and natural barriers on the movement of small mammals in Banff National Park, Canada

Barriers that dissect continuous habitat, such as roads, initiate the process of habitat fragmentation and can filter or eliminate animal movement through otherwise pristine areas. A severe road barrier appears to exist in Banff National Park where the Trans‐Canada Highway (TCH) is known to impede the movement of some animals, but the effects of this barrier relative to natural barriers and continuous habitat have not been assessed. To compare animal movement in response to these three contexts, we translocated three species of murid rodents (meadow voles, Microtus pennsylvanicus, deer mice, Peromyscus maniculatus, and red‐backed voles, Clethrionomys gapperi) across the TCH (an artificial barrier), a forested strip in the median of the TCH (a natural barrier), in the adjacent highway verge (continuous grassy) and in nearby forest (continuous forested). We coated murids with fluorescent dye, released them at standardized distances, and followed the resulting trails to obtain detailed information on movement paths. Overall, individuals were 20% less successful crossing the TCH than natural (forested) barriers and 10% less successful crossing natural barriers than continuous habitat, but exhibited marked variation among species. Path tortuosity was negatively related to return success; more convoluted paths were associated with lower success across all treatments. Deer mice returned to their home‐ranges more often than the other two species, perhaps because of their generalist habitat preferences, nocturnal activities, or larger home ranges. In all treatments, the fractal dimensions of movement paths were very qualitatively similar to those observed for return success and tortuosity, indicating that the responses of these species were not influenced by spatial scale. Together, our results suggest that (a) murid species differ in the movement limitation imposed by both natural and artificial barriers, perhaps as a consequence of differences in habitat preferences and activity patterns, and (b) murid responses to a road barrier are only moderately different from responses to the natural barrier of a forest edge in the context of translocation experiments.