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

Christopher Servheen

Commitee Members

David Patterson, Fred Allendorf, L. Scott Mills


bear rubs, grizzly bear, mark–recapture, noninvasive genetic sampling, Northern Continental Divide Ecosystem, Pradel model, trend monitoring, Ursus arctos


University of Montana


Wildlife managers need reliable estimates of population size, trend, and distribution to recover at–risk populations, yet obtaining these estimates is costly and often imprecise. The threatened grizzly bear (Ursus arctos) population in northwestern Montana has been managed for recovery since 1975, yet no rigorous data were available to evaluate the program’s success. We assessed population status using data from a large noninvasive genetic sampling project and 33–years of physical captures. Our abundance estimate, Nˆ= 765 (CV = 3.8%), was more than double the working estimate. Based on our results, the total known, human–caused mortality rate was 4.6%, slightly above the 4% level considered sustainable. Genetic diversity approached levels seen in relatively undisturbed populations, with the only signal of population fragmentation that aligned with landscape features being across U.S. Highway 2. I used these encounter data to parameterized a series of simulations to assess the ability of noninvasive genetic sampling, specifically surveys of naturally occurring bear rubs, to estimate population growth rates. I used data on 379 grizzly bears identified from bear rub surveys in a range of Pradel model simulations in program MARK. I evaluated model performance in terms of: (1) power to detect declines in population abundance, (2) precision and relative bias of estimates, and (3) sampling effort required to achieve 80% power to detect a decline within 10 years. Simulations suggest that annual bear rub surveys would exceed 80% power to detect a 3% annual decline within 6 years. Robust design models with 2 surveys per year provide precise and unbiased estimates of trend and abundance. Designs with 1 survey per year are less expensive but only yield trend and apparent survival estimates. I provide recommendations for designing a program to monitor population trends by sampling at bear rubs. Systematic bear rub surveys may provide a viable alternative to telemetry–based methods for monitoring trends in grizzly bear populations. This study illustrates the power of molecular techniques to rapidly assess population status and trends at landscape scales and provide detailed demographic and genetic data to guide and evaluate recovery efforts.



© Copyright 2008 Jeffrey Brian Stetz