Spatio‐temporal co‐occurrence of cougars (Felis concolor), wolves (Canis lupus) and their prey during winter: a comparison of two analytical methods

摘要: Aims  To examine the spatio-temporal co-occurrence of cougars (Felis concolor), wolves (Canis lupus), and their prey during winter using monthly (November–March) species–environment relationship models. In addition, to contrast predictions across two methods: logistic regression Geographic Information System (GIS) image correlation. Location  The eastern front ranges Canadian Rocky Mountains (south-central Alberta), approximately 100 km west Calgary, including portions Banff National Park Kananaskis Country. Methods  Snow-tracking data were collected simultaneously for cougars, wolves, elk (Cervus elaphus), deer (Odocoileus virginianus O. hemionus) between November March, 1997–2000. Track synthesized in a GIS. Logistic Akaike's information criterion (AIC) used select optimal environmental models each species. We first examined by iteratively species as dependent variable (presence/absence) analysis all other track-density estimates independent variables. built predictive surfaces GIS exponent form models, assessed model accuracy with receiver operating characteristic curve. then re-examined pairwise correlations probability month. correlation results compared illuminate mechanisms investigate consistency methods. Results  Cougars showed trend distribution from higher elevation less rugged terrain December, lower more March. This differed that which stable affinity low valley bottoms months. indicated positive negative associations month, changes over time. Notably, there was shift both predators found high surfaces, except month January. Our comparison spatial amongst increased winter, negatively correlated February. Combining approach we converged spatially at landscape scale (i.e. valley), while showing discrete use space time habitat attributes (e.g. forest cover, topographic complexity, track density). Main conclusions  Mountains, distributions into floor progressed. distinct intensity this shift. determined alone fails explain co-occurrence. must be coupled investigation respective account temporal associations. suggest approaches represent different ecological scales: may best landscape- (valley) level analysis, is site-level analysis. Ultimately, critical our Finally, variability observed suggested annual seasonal obscure important patterns processes, especially cougars.