Differential space use inferred from live trapping versus telemetry: northern flying squirrels and fine spatial grain

Small mammal space use is inferred from live-capture data or various methods of tracking, with differences between these methods potentially affecting the input and subsequent inferential abilities of resulting wildlife-habitat models. Unlike tracking via radio telemetry, live trapping employs use of bait, which is known to change proximate animal density as evident in many food addition studies (the 'pantry effect'), and conceivably bias individuals' space use, particularly if measured over small spatial extents in heterogeneous areas. The present study analysed both trapping and telemetry data from northern flying squirrels (Glaucomys sabrinus) to assess whether different habitat associations could be generated based on methods alone. Conditional on sampling method, two different space-use patterns were identified from the same group of squirrels and two significantly different sets of habitat model input were associated with each. Trap areas were not used post capture; once enumerated, animals on average (n = 34) spent over 80% of their time from 100 to 200+ m, upwards of 800 m, away from trap areas. Using telemetry and fine-grained habitat structure data, this study found 33% of sampled squirrels used areas not identified via habitat-stratified trap effort (specifically black spruce habitat). It is concluded that wildlife-habitat investigations dealing with fine spatial grain are likely to acquire different results using trapping versus telemetry, especially if animals are relatively mobile and habitat structure is relatively heterogeneous.