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

L. Scott Mills

Commitee Members

Daniel H. Pletscher, Howard B. Quigley, Michael S. Mitchell, Toni K. Ruth


habitat use, harvest, population contribution, RSF, source-sink, spatially structured populations


University of Montana


The structuring of populations within a metapopulation, connected through dispersal, is important to basic and applied ecology. However, a considerable gap exists in our knowledge of the influence landscape heterogeneity has on dispersal and its consequences. We examined landscape effects on dispersal and its consequences for puma (Puma concolor) populations using data from three separate populations in the Central Rocky Mountains including the Northern Greater Yellowstone Ecosystem (NGYE), the Southern Greater Yellowstone Ecosystem (SGYE) and the Garnet Mountains of Montana. We found human-induced mortality reduced inter-population vital rates with population consequences. The NGYE population was dependent largely on immigration for its own growth and emigration for its overall contribution to the metapopulation. The Garnet population, subject to high human induced mortality, was a population sink until a 915 km2 area was closed to hunting, after which that area became a source, largely from a 16x emigration increase. Human-induced mortality affected emigration, dispersal distance, and establishment success. Effective dispersal of subadult males (n=59) was reduced primarily through direct mortality, whereas females (n=67) were more influenced by indirect effects reflected in reduced emigration and dispersal distances. We further examined dispersal-landscape relationships through disperser habitat preferences. Generalized linear mixed-effects models were constructed from a priori models of disperser habitat to test the importance of forest cover, topographic cover, suitable hunting habitat, and anthropogenic disturbance. Models were fit to location data from GPS-marked (n=11) and VHF-marked (n=123) dispersers from all three study areas. Model selection, using Akaike’s Information Criterion, found landscape characteristics associated with successful hunting of ungulate prey combined with anthropogenic disturbance parsimoniously explained locations of GPS-marked individuals. For VHF-marked dispersers the hunting habitat model ranked highest with the combined hunting habitat and anthropogenic disturbance model second. Model fitting from both datasets indicate habitat characteristics important to dispersers is similar to resident adults. A resource selection function estimated from the top GPS model was highly predictive of disperser locations from the independent VHF dataset. This model can identify areas important to dispersing individuals and suggests adult habitat is a useful surrogate for landscape connectivity.



© Copyright 2011 Jesse R. Newby