Comparing estimates of landscape resistance to animal movement

Wildlife corridors are an essential tool to enable wildlife to continue moving through the landscape in face of habitat loss and fragmentation. Focal species based models used to identify likely corridors rely on resistance surfaces that estimate how much landscape elements promote or impede movement. I examined whether we can predict long-distance movements based on habitat suitability estimates. I also assessed which type of data is best used to estimate landscape resistance to animal movement.;For elk and desert bighorn sheep I tested whether resistance estimated by experts, from habitat suitability models, or from within-home range movement analysis best predicts long distance movements. For bighorn sheep, a habitat specialist, I found that a negative exponential function best described the relationship between habitat suitability and resistance values, indicating that long-distance movers readily travel through moderately-suitable areas and avoid only the least suitable habitat. All three suitability estimates performed better than chance, but habitat suitability and movement analysis outperformed expert opinion models. For elk, a habitat generalist, all three suitability estimates performed about the same as chance.;For kinkajous I estimated landscape resistance from habitat suitability and movement analysis as well as from dispersal and mating movements. The results suggest that kinkajous perceive the landscape as more connected during dispersal and mating movements than while in the home range. Kinkajous are tightly linked to forest in the home range, but no features in the study landscape posed resistance to mating and dispersal movements. Thus, kinkajous readily traverse low-quality habitat during mating and dispersal movements.;Tolerance to low-quality habitat during long-distance movements has now been observed in several wildlife species. This is good news for corridor designers, because it indicates that for mobile species the matrix between protected areas is more permeable for dispersal and other long-distance movements than expected from habitat suitability models. If mobile animals can readily traverse habitat of lower quality, planning for wildlife corridors would be more flexible. Further research is necessary to determine whether distances between patches of good-quality habitat that animals can traverse are species or landscape specific. For desert bighorn sheep, any of the 3 habitat suitability estimates can inform corridors. Habitat generalists like elk are of little utility in a multi-species corridor design. Studies of additional habitat specialists are needed to quantify the tradeoff between cost and reliability of different suitability estimates.