PARENT SESSION
Contributed Oral Session 48: Landscape Ecology: Animal Dynamics
Tuesday, August 9, 8:00 AM - 11:30 AM, Meeting Room 520 B, Level 5, Palais des congrès de Montréal

Patch occupancy by the North American red squirrel, Tamiasciurus hudsonicus, in a fragmented agricultural landscape: the relative role of local factors and landscape composition and configuration.

Darlow, Neil^*,1, ¹ University of Alberta, Edmonton, Alberta, Canada

ABSTRACT- One central theme of ecology is to understand how pattern and scale influence the distribution and abundance of organisms. Landscape ecologists typically address this issue by focusing on patterns of species response attempting to infer the process causing those patterns. This research aims to examine the occurrence of the North American red squirrel in Canada′s sub-boreal aspen parkland which is both highly fragmented and of low habitat quality, to address key questions about species response to habitat loss and fragmentation. Specifically this research has three objectives, to: 1) examine the magnitude and direction of local (patch and habitat) and landscape scale effects on red squirrel patch occupancy, 2) determine the relative contribution of landscape composition and landscape configuration effects to patch occupancy, and 3) examine the effect of a temporary increase in habitat quality, in the form of a spruce mast on each of objectives 1 and 2. Over 3 years we used small mammal trackplates to determine red squirrel occupancy in 118 landscapes within Alberta′s aspen parkland ecoregion. Using generalized linear models, we related a series of independent variables, chosen a priori, at the habitat, patch and landscape levels to our estimates of occurrence. Preliminary analyses indicate that local factors explained 32% of the deviance in occurrence and in particular that the presence of beaked hazel, Corylus cornuta was a key driver. At the landscape scale, configuration variables had greater predictive ability than composition variables, explaining 34% and 12% of the deviance respectively. Importantly, although accounting for merely 1% of landscape composition, variables related to the amount and distribution of white spruce in the landscape were significant predictors for all models. We examined further the ecological importance of white spruce to red squirrel occurrence by assessing occupancy the year following a spruce mast at sites deemed absent the previous year. We found an increase in occupancy but one that did not swamp landscape effects. In particular we found the significant predictors to be the amount and distance to white spruce, and the total length of wooded fencerows in landscapes, the latter suggesting a potential role for wooded fencerows in landscape connectivity.

Key words: habitat quality, presence-absence, landscape composition, landscape configuration