Multiple spatial-scale resource selection function models in relation to human disturbance for moose in northeastern China

Guangshun Jiang,Jianzhang Ma,Minghai Zhang,Philip Stott
DOI: https://doi.org/10.1007/s11284-008-0519-8
2008-01-01
Ecological Research
Abstract:The moose ( Alces alces cameloides ) population in northeastern China is on the southernmost edge of its distribution in Asia. A survey was conducted to determine moose resource selection and the effects of human disturbance on moose in a study area of 20,661 ha located on the northwestern slope of the Lesser Khingan Mountains, located in northeastern China. Predictive models of resource selection were developed using logistic and autologistic regression. All models considered resource variable selection at two spatial scales, patch and landscape. At the patch scale, moose preferred larger birch ( Betula platyphylla ) patches, but avoided larger tamarack ( Larix gmelinii ) patches. At the landscape scale, moose preferred higher densities of tamarack patches, i.e., heterogeneity of tamarack stands, selected areas with more abundant annual shoots, terrain conducive to better concealment, higher altitudes and areas saturated with soil moisture. Roads and forest harvest intervals were identified as important human disturbance factors. This is the first time that moose have been reported to avoid roads, and the avoidance distance was nearly 3 km. We believe that in this region moose under the influence of roads are behaviorally plastic, compared with the indifference of moose to the presence of roads in other regions. Moose avoided forest areas logged more than 3 years previously and preferred areas logged 1–2 years previously. In addition, it may be necessary to monitor the effect of the dynamic of density of roe deer on the spatial distribution of the moose population.
What problem does this paper attempt to address?