Seasonal and altitudinal influences on the home range and movements of common wombats in the Australian Snowy Mountains

Abstract

Many animals throughout the world are excluded from areas because of seasonal snow cover. The aim of this study was to determine how snow influences the home range use and movements of the common wombat, a large burrowing mammal that remains active in the subalpine zone of the Australian Snowy Mountains throughout winter but is not resident in the alpine zone (above treeline). Global positioning system collars were deployed on wombats to monitor nightly movements continuously over both the winter and non‐winter periods. Home ranges of wombats (six male, five female) were far larger than previously reported (mean = 172 ha; 95% kernel method), and increased significantly with altitude. Wombats typically remained in their non‐winter home range during winter, but they contracted their range (by 7–43%) and shifted their centre of activity. Some wombats also moved more slowly and did not travel as far each night during winter. This study has shown that wombats at their upper range limit in marginal habitat exhibit a high degree of behavioural flexibility and have a surprising capacity for long‐distance movements over large home ranges, despite their need to burrow. This suggests that the alpine zone is easily within their dispersal range, but they are currently constrained by snow depth. If the snow cover continues to decline, then wombats will be limited only by the suitability of the habitat in the alpine zone, such as for burrowing.

Using a resource selection approach to predict the suitability of alpine habitats for a common herbivore

“…Differences in the selection of resources by individuals may be largely dependent on the availability or quality of the habitat (Gustine et al . ), and this was also reflected in the larger size of home ranges of wombats at higher altitudes (Matthews & Green ). The general geographic location of the wombat's home range resulted in individual specialization in resource selection because of a patchy availability of particular resources across the landscape.…”

“…This study investigated the within‐home range resource selection of wombats in the subalpine zone of the Australian Snowy Mountains to provide a more detailed understanding of the habitats important to them, and to build a model of habitat suitability over the broader landscape, incorporating both the subalpine and alpine zones. Matthews and Green () found that wombat home range size increased with elevation, and building on this work, we hypothesized that differences in wombat spatial behaviour over an altitudinal gradient would lead to high variation in the selection of resources by individuals, but there would be commonalities for the most important resources at a population level. The main objectives of this study were to (i) develop estimates of habitat selection to determine which resources individual wombats were selecting relative to what was available, and (ii) identify resources selected across individuals, and map those resources over the landscape to predict areas of suitable wombat habitat in the subalpine and alpine zones.…”

“…). Details of the ecology of wombats in the study area, including feeding and spatial behaviour are provided by Green (), Matthews (), and Matthews and Green ().…”

“…All GPS locations gathered over the year were used in analyses, without stratification by season, because wombats were shown to be faithful to their home range throughout the year (i.e. the summer range overlapped the winter range (Matthews & Green )). In this situation, bridging temporal scales was necessary because winter habitats are important determinants of summer home range selection (Boyce ).…”

“…Briefly, conditional logistic regression was used to estimate the model coefficients of the RSF using a subset (80%) of the location data. The RSF values were then predicted in ArcGIS to create a predictive map, with the extent of the map based on the MCP home range of the wombat (Matthews & Green ) with a buffer equal to its 95% movement distance (http://onlinelibrary.wiley.com/doi/10.1111/aec.12053/suppinfo). The predictive map was reclassified into 10 ordinal bins to compare the expected use in each bin with the observed use from the remaining location data subset (20%).…”