Modeling the effects of dispersal on predicted contemporary and future fisher (Martes pennanti) distribution in the U.S

Abstract

Many species at high trophic levels are predicted to be impacted by shifts in habitat associated with climate change. While temperate coniferous forests are predicted to be one of the least affected ecosystems, the impact of shifting habitat on terrestrial carnivores that live within these ecosystems may depend on the dispersal rates of the species and the patchiness of the environment. Much of the Northern U.S. Rocky Mountains is comprised of high to moderate elevation mountain ranges separated by large river valleys, extensively used for agriculture. Connectivity among patches is dependent on patch configurations and an animal's ability to disperse. We aim to understand how dispersal affects projected future available habitat for a threatened, endemic carnivore, the fisher (Martes pennanti). We designed a survey to detect fisher using non-invasive genetic sampling across their historical distribution in Montana and Idaho. We subsequently modeled the distribution of fisher across western Montana and northern Idaho using a suite of vegetative, topographic, and climatic variables. To model future distribution under a changing climate, we used only topographic and climatic variables found to be important in the current distribution model, since vegetative variables could not readily be projected into the future. We considered a global climate model with two climate change scenarios (high emissions [A2] or reduced emissions [B2]) and three time steps (2030, 2060, and 2090). We incorporated the effect of dispersal ability into our species distribution model by varying the distance at which newly created patches of habitat could be colonized. We found that the probability of current fisher occurrence was highest given the presence of mesic forest types with tall trees, high annual precipitation, and mid-range winter temperatures. Future predictions of fisher habitat show an increase in area of high-probability habitat under most dispersal assumptions.

Citation

Olson, Lucretia; Schwartz, M. 2013. Modeling the effects of dispersal on predicted contemporary and future fisher (Martes pennanti) distribution in the U.S. Rocky Mountains. In: 43rd Annual Meeting of the Idaho Chapter of The Wildlife Society; 11-14 March 2013; Coeur d'Alene, Idaho. Idaho Chapter of The Wildlife Society. p. 31. Online: http://www.ictws.org/2013AnnualMeetingFinalProgram.pdf