This study evaluates 7 mark-recapture projects using grizzly bear (Ursus arctos) DNA from bait stations with hair snags in British Columbia, 1996-98. The emphasis of these projects was to estimate population size. We evaluate how well this objective was met in terms of mark-recapture assumptions and ultimate precision and bias of estimates. We used the Pradel model in program MARK to compare smaller grids and cells with more intensive hair-snag coverage to larger grids and cells. Assuming demographic closure, the Pradel model estimates fidelity, recapture rate, and immigration rate from each of the grids. Results from this analysis suggest (1) closure violation was more likely in smaller grids than larger grids, and (2) recapture rate decreased as cell area increased (and sample intensity decreased). Sex-specific differences were suggested but not verified due to large standard errors of Pradel parameter estimates. We used newer mixture models in program MARK to evaluate heterogeneity with each sampling design and produced meta-analysis population estimates. These estimates were compared with corresponding estimates from the program CAPTURE. Results suggest that heterogeneity is evident with all sampling designs and the newer mixture models provide higher precision than CAPTURE heterogeneity models in most cases. This study highlights the tradeoff in study design faced when designing DNA mark-recapture studies. Namely, intensive sampling using smaller cells is needed to detect and model individual heterogeneity, but smaller grids are most vulnerable to closure violations. Strategies to confront this issue, such as use of radiocollared bears to index movements, are suggested. We discuss applications of recently developed mark-recapture models to DNA, radiotelemetry, and other sources of data that estimate apparent survival, population rate of change, and movements of bears in sampling areas. We argue that use of these newer methods is the most advantageous step in the application of the DNA mark-recapture technique to bear populations.
- Volume 13
- Issue
- Pages 137-152
- Publication Date 1 January 2002
- DOI
- File Size 811.28 KB
- Download