Seasonal resource selection of woodland caribou (Rangifer tarandus caribou) across a gradient of anthropogenic disturbance

Megan L Hornseth,Robert S Rempel

doi:10.1139/cjz-2015-0101

Megan L Hornseth, Robert S Rempel

Open Access

https://doi.org/10.1139/cjz-2015-0101

Copy DOI

Abstract

Resource selection functions are useful tools for land-use planning, especially for wide-ranging species with sensitivity to anthropogenic disturbance. We evaluated five a priori hypotheses describing seasonal habitat selection of woodland caribou (Rangifer tarandus caribou (Gmelin, 1788)) across three regions in northern Ontario. Two regions were Boreal Shield dominated, one area with relatively high anthropogenic disturbance (due to commercial forestry) and the other with relatively low anthropogenic disturbance. The final region was located on the wetland-dominated Hudson Bay Lowlands. Each region encompassed two caribou management ranges: one was used for model development and the other for model evaluation. We developed seasonal resource selection probability functions using seasonal utilization distributions and isopleths derived from GPS collar data (from 212 caribou) to identify high- and low-use areas. We explored selection across five spatial scales; selection patterns were strongest at the 10 000 ha scale. We found temporal and spatial variations in all environmental predictors across ranges and seasons, especially in the Hudson Bay Lowlands. Our results consistently supported the integrated global model (with common variables but range-specific coefficients) where caribou habitat use is related to minimizing apparent competition with moose (Alces alces (L., 1758)) while avoiding disturbed areas, and utilizing areas with adequate forage.

Highlights

Ontario’s Far North constitutes one of the world’s last remaining intact landscapes
We examined environmental variables across a range of spatial scales and developed seasonal resource selection probability functions (RSPFs) for caribou under each of the proposed hypotheses to (i) identify the important features associated with habitat selection, (ii) determine if there are distinct differences in caribou habitat selection among different analysis regions across the north of Ontario and among seasons, and (iii) assess the likelihood of alternative hypotheses about the mechanisms that drive habitat selection
During a preliminary assessment of analysis across five spatial scales (100, 500, 1 000, 5 000, and 10 000 ha spatial analysis units), we used Akaike’s information criterion (AIC) selection procedures to select the spatial analysis scale at which the model was best supported and we found that patterns of selection were best supported at the broadest spatial extent (10 000 ha) for all ranges (Fig. 3)

Summary

Introduction

Ontario’s Far North constitutes one of the world’s last remaining intact landscapes (together with the Russian taiga and the Amazon; Keddy 2010). This area is of global significance because it contains the largest wetland complex in North America and provides important ecosystem services such as carbon storage and cycling on a global scale and regulates the effects of climate change (Abraham and Keddy 2005; Abraham and McKinnon 2011; Riley 2011). The future of the Far North is uncertain due to several drivers: climate change, commercial forestry, development associated with transportation and energy for both industry and local communities, and mineral exploration and development (Far North Science Advisory Panel 2010)

Methods

Results

Conclusion