| Summary: | Dissertation (Ph.D.) University of Alaska Fairbanks, 2014 Snowmobiling is a popular winter activity in northern regions of North America. Although snowmobiles are important utility vehicles and serve as a means of outdoor recreation, their activity is known to affect plants and animals. These effects have been a growing concern over the past 20 years as a result of increased snowmobile activity into once inaccessible natural areas. Minimizing the impacts of snowmobiles on biota, preserving the quality and character of wilderness areas, and providing adequate access to snowmobilers for traditional activities has been a challenge for public land managers in Alaska. To address the effects that snowmobiles have on ecological systems at site-specific and landscape-level scales, I conducted a study in the Kenai National Wildlife Refuge, a boreal ecosystem located in southcentral Alaska, to determine 1) the response of woody wetland plants to snowmobile traverses at varying snow depths, 2) thetemporal and spatial variation of a winter soundscape with emphasis on anthrophony, in general, and snowmobile noise, specifically, 3) the effects of snowmobile noise on wilderness character and naturalness, and 4) the spatial and physiological response of moose (Alces alces) to snowmobile activity and noise. I used a combination of traditional experimental designs and statistics, machine learning, and spatially-explicit predictive modeling to assess the effects snowmobile activity has on these four issues. I found that snowmobile activity reduced the number of living stems and inhibited the growth of woody wetland plants by direct contact with protruding vegetation above the snow and indirectly from snow compaction. Snowmobile noise was not a large contributor of noise to the soundscape but was pervasive in remote areas. Snowmobile noise affected a significant area of Congressionally-designated wilderness altering the naturalness and character of the wilderness soundscape. Moose exhibited a distinct spatial partitioning and ...
|
|---|