Year of Award


Document Type


Degree Type

Master of Science (MS)

Degree Name

Wildlife Biology

Department or School/College

College of Forestry and Conservation

Committee Chair

Lisa Eby

Commitee Members

Fred Allendorf, P. Stephen Corn


beaver, Castor canadensis, Columbia spotted frog, ecosystem engineer, landscape genetics, landscape pattersn, lentic habitat, population structure, Rana luteiventris


University of Montana


Amish, Stephen J. M.S., Autumn 2006 Wildlife Biology ECOSYSTEM ENGINEERING: BEAVER AND THE POPULATION STRUCTURE OF COLUMBIA SPOTTED FROGS IN WESTERN MONTANA Chairperson: Lisa Eby Beavers (Castor canadensis) are considered ecosystem engineers, altering hydrologic regimes, ecosystem processes, and modifying community structure. Effects of beaver on the spatial pattern of lentic habitat and populations using those habitats have not been examined. I used a landscape database and eight microsatellite markers to compare the scale and pattern of lentic sites, their occupancy, and population structure by Columbia spotted frogs (Rana luteiventris) between watersheds with and without beaver activity. Across all watersheds frog breeding sites were more clustered than the underlying pattern of lentic habitat. Beaver watersheds had four times as many lentic and breeding sites than non-beaver watersheds. Non-beaver watersheds often had only one frog breeding site. Frog breeding sites were more dispersed within beaver drainages. In addition, frog breeding sites were evenly distributed across the elevational gradient in beaver watersheds while they were centered above the watershed midpoint in non-beaver watersheds. Columbia spotted frog breeding sites were more dispersed within drainages with evidence of beaver presence than would be expected given the configuration of the underlying lentic habitat and have persisted despite being separated by distances larger than its dispersal ability. The genetic divergence seen within watersheds revealed that landscape configuration affected the fine scale population structure of Columbia spotted frogs. Landscape patterns of breeding sites were reflected in the presence and strength of isolation by distance equilibriums and the overall level of population subdivision within watersheds. Watersheds with beaver presence and an average distance of less than five kilometers between breeding sites showed higher levels of connectivity than did non-beaver watersheds with an average distance of more than five kilometers between breeding sites. More importantly, short beaver watersheds had lower levels of genetic divergence between breeding sites than those in long non-beaver watersheds separated by the same distance, even when distances were within the commonly observed dispersal ability of the frogs. Typical beaver watersheds in southwestern Montana with similar habitat configurations are likely composed of a single population, while non-beaver watersheds likely contain a single or a few isolated population/s. Careful consideration of potential population effects for species dependent upon habitat beaver create is required.



© Copyright 2006 Stephen Joseph Amish