Abstract

Winter (01 Jan – 15 Apr) habitat selection by Shiras moose ( Alces alces shirasi) within individual home ranges was investigated using global positioning (GPS) collars on 23 adults (7 M, 16 F) during 2 winters (2005, 2006) of differing snow pack in the Snowy Range of the Medicine Bow mountains, southeastern Wyoming. Although wide variation was observed among animals, moose commonly selected for riparian shrub, deciduous forest, and mixed forest cover types. Mixed mountain shrub was occupied extensively by several moose and was used more often towards the end of winter, as moose tended to minimize movements and increasingly use other cover types in place of riparian shrub. Differences in selection ratios between sexes were not detected for any cover type and differences between years were minimal among moose collared both winters.

Diets of moose during winter were also investigated through fecal analysis. Willow (Salix spp.) and subalpine fir (Abies lasiocarpa ) composed a mean 60% and 30% of moose diets, respectively, with the remaining 10% comprised mostly antelope bitterbrush (Purshia tridentata ), mountain mahogany (Cercocarpus montanus), or Saskatoon serviceberry (Amelanchier alnifolia). Diets were more diverse in the year of less snow pack. Patterns of habitat selection by collared moose did not differ between winters, as would be expected if forage availability differed considerably between years. This was likely because moose GPS locations did not overlap fecal collection sites and forage item selection within those sites occurs at a different scale than the selection of cover types within home ranges.

A literature-based winter habitat suitability index (HSI) model was developed from common geographic information system (GIS) layers and scrutinized with GPS locations of sampled moose. However, the HSI model was poorly predictive of winter habitat occupancy. More accurate resource selection function (RSF) models were constructed by integrating moose GPS locations with more refined GIS data layers. Numerous vegetative, topographic and distance variables were calculated across the study area and were used in a forward stepwise general linear regression model to identify important components of moose habitat during winter and non-winter seasons. Distance to forest edge and distance to deciduous forest were significant predictors in both seasons. Slope also influenced habitat use year-round, although slope2 was a factor in the winter model. While distance to riparian shrub was predictive of moose habitat occupancy during winter, the total area of riparian shrub within a circular 1 km radius was a better determinant of summer habitat use. The combination of variables in the winter model accounts for the distribution of willow, subalpine fir, mountain mahogany and antelope bitterbrush, in proximity to forest cover. The non-winter model demonstrated the nearly exclusive importance of riparian shrub habitat in proximity to thermal cover across a wider range of elevations than during winter.

A technique was employed to make spatial calculations of the potential range capacity for moose using the winter RSF map predictions observed within individual moose winter home ranges. A wide range of capacity estimates were computed by adjusting the minimum habitat quality and maximum size parameters observed in moose winter ranges. Because not all moose are energetically capable of occupying the maximum observed home range size or competing for the highest quality habitat, more sensible estimates were produced using the mean winter home range quality and size inputs.