PARENT SESSION
Poster Session 39: Late Breaking and Newsworthy Posters
Friday, August 12, 8:30 AM - 10:30 AM, Exhibit Hall 220 A-E, Level 2, Palais des congrès de Montréal
Accounting for uncertainty in monitoring: Making the most of your money!.
Ensbey, Michelle1, Tyre, Andrew2, Wintle, Brendan 1, Taylor, Scott3, 1 Environmental Science, Department of Botany, University of Melbourne, Melbourne, Victoria, Australia2 School of Natural Resources, University of Nebraska, -Lincoln, United States of America3 Upland Game Program, Nebraska Game and Parks Commission, Lincoln, Nebraska, United States of America
ABSTRACT- Monitoring of wildlife populations is commonly undertaken by researchers, managers and community groups to provide early warning of unacceptable population declines and to evaluate the impacts of management or exploitation. However, many monitoring regimes are often poorly designed or analyzed, commonly failing to account for uncertainty arising from observation error and natural variation, and providing insufficient power to detect important population declines or increases. Few monitoring programs exploit monitoring data to their full potential because they fail to consider influences on populations at multiple spatial and temporal scales. In this study, we re-analyze a decades long time series of Northern Bobwhite Quail monitoring data to investigate the causes of population fluctuations at a national, regional and local levels and at fine (yearly) and coarse (ten-yearly) temporal scales. We also exploit newly developed methods for analyzing zero-inflated data to incorporate observation error arising from imperfect detectability in the Bobwhite data. The analysis is implemented using a hierarchical Markov Chain model fitted in the Bayesian freeware WinBUGS. The use of Bayesian analytical techniques increase inferential power by allowing multiple sources of data, including expert knowledge to be incorporated in a single analysis.
Key words: Monitoring, Northern Bobwhite (Colinus virginianus ), Zero inflated, Bayesian Model