Life on the edge: Industrial footprint and edge effects variably influence the distribution of a boreal small mammal

Biodiversity is lost through habitat loss and fragmentation, but the independent role of landscape configuration in moderating ecological processes is empirically contentious. Landscape and fragmentation theories suggest that the spatial arrangement of habitat patches influences species distribution, but testing these effects in terrestrial landscapes undergoing anthropogenic change is challenging due to replication demands from multiple patches in multiple landscapes.

We investigate the role of landscape structure altered by natural resource development (hydrocarbon extraction and timber harvest), measured at multiple spatial scales, on the spatial distribution of American red squirrel ( Tamiasciurus hudsonicus ), a small mammal forager. With occurrence data from a uniquely large dataset of 430 camera traps across 10 regions of Canada's industrialized western boreal forest, we tested whether industrial development affected red squirrel distribution through changes to landscape composition, configuration or context‐dependent combinations of both.

Red squirrel occurrence was best explained by the interacting effects of landscape‐level industrial footprint and edge creation. Squirrel occurrence increased with cumulative footprint, consistent with resource complementation theory. Red squirrel occurrence decreased with a higher amount of edge created by industrial features. Edge density interacted with cumulative industrial footprint, such that these edge effects were more strongly negative in intensively disturbed landscapes, and less so in more intact landscapes.

We suggest that the negative population‐level change due to edge density arises from increased encounters with mammalian and avian predators at forest edges. Anthropogenic disturbances that create a high amount of edge habitat, such as seismic lines, have a disproportionately negative impact on red squirrels.

Synthesis and applications. While management in the oil sands typically focuses on the amount of habitat lost and looks to ‘spatial offsets’ of undisturbed habitat elsewhere as a solution, we demonstrate that both composition and spatial configuration must be considered on appropriate spatial scales when developing land‐use strategies in patchy ecosystems undergoing anthropogenic change.