Monday, August 3, 2009 - 3:40 PM

OOS 6-7: Connecting Alaskan landscapes into the future

John M. Morton, US Fish and Wildlife Service, Karen A. Murphy, U.S. Fish and Wildlife Service, Falk Huettmann, University of Alaska, and Nancy Fresco, University of Alaska.

Background/Question/Methods Sustaining and enhancing landscape-level connectivity will be critical for facilitating fish and wildlife adaptation to accelerated climate change. In early 2008, we initiated a multi-partner project to examine landscape connectivity in Alaska under different climate change scenarios. Our goals were to identify lands and fresh waters that likely serve as landscape linkages now and in the future given climate change, and to identify conservation strategies that will help maintain connectivity by focusing conservation efforts, minimizing redundant research and monitoring, and sharing data and information.  We used decadal precipitation and temperature predictions over the next 100 years from a composite of the five best IPCC models for Alaska, down-scaled to 2 x 2 km using PRISM data. We used RandomForests™ to model current and future distributions of biomes, vegetation communities (based on LANDFIRE), productivity/biodiversity (NDVI), caribou, Alaska marmot, trumpeter swan, and reed canary grass. These four species were chosen for pilot modeling because of their disparate (but representative) life histories and data availability. We essentially modeled the future climatic envelope for each of these resources, but with additional constraints imposed by our collective knowledge and data availability; e.g., reed canary grass, an invasive, is being dispersed along highways and therefore on streams intersected by those roads.  Our approach is known to be robust, used in climate modeling, and deals with non-linear interactions well. 

Due to the developmental nature of this project, results are evolving.  We have the methods and data sets to model some resources reasonably well. However, it’s apparent that spatial data for most species are sparse, and existing data sets will have to be supplemented to make modeling feasible.  We are producing an initial report in 2009 that will provide the conceptual framework developed through this effort, identify shortcomings and research needs, recommend steps for bringing these concepts to finer scales, and identify conservation strategies to be explored.  In Alaska, which still has mostly intact ecosystems, adaptation strategies are likely to focus on retaining the extant matrix through the use of less-than-fee approaches to land conservation and partnering with international conservation efforts.  We found that biomes, vegetation, and wildlife in interior Alaska will experience substantial longitudinal and configurational shifts (in addition to elevation and latitude) that makes international conservation efforts like the Yellowstone-to-Yukon initiative relevant to sustaining the viability of Alaska in the next century