Danah Duke, Department of Biological Sciences, University of Alberta, PO Box 1852, Banff, AB T0L 0C0, Canada; Mark Hebblewhite, Wildlife Biology Program, University of Montana, Missoula, MT 59812, USA; Paul C. Paquet, Conservation Biology Institute, PO Box 150, Meacham, SK S0K 2V0, Canada; Carolyn Callaghan, Melanie Percy, Central Rockies Wolf Project, 910 15 Street, Canmore, AB T1X 2W1, Canada

Gray wolves (Canis lupus) naturally recolonized the Bow River Valley of Banff National Park in the mid 1980's. However, research showed high levels of human activity prevented wolves from moving between critical habitats within the Bow Valley. Using empirical data, we previously developed a spatially explicit predictive model that quantified wolf habitat use and movements. The model predicted that reducing human use in dysfunctional habitat would increase carnivore use of the area. During 1997, management actions significantly reduced human use within one affected corridor of the Bow Valley to a level shown by the model to allow wolf movement. We evaluated the response of radio-collared wolves to corridor restoration by comparing two winters of post-restoration movements (1997 to 1999) with four years of pre-restoration movements (1993 to 1997). Wolf use of the corridor, as measured by number of wolf radio-telemetry days and transect crossing indices, was significantly higher after corridor restoration. Careful consideration of other potential confounding factors (snow depth, prey distribution) failed to offer alternative explanations for the increased use. These results suggest that empirically based habitat/movement models can be used successfully to identify travel linkages and determine levels of human activity that impede wolf movements. We believe this is a promising approach for the identification and restoration of wildlife movement corridors.