GO NORTH YOUNG BIRDER! THE BBS FRONTIER HIGHLIGHTS THE NEED FOR A DISCUSSION ON WHAT WE WANT FROM PUBLISHED ESTIMATES OF STATUS AND TREND (AND THE NEED FOR MORE DATA!).
The sparsely surveyed northern edge of the BBS presents particular challenges for status and trend assessment in Canada. For example, estimating stratum-level trends as fixed effects allows trends to vary regionally, but can result in spurious and extreme trend estimates in the sparsely surveyed, northern strata. Modifications to the hierarchical structure of the BBS model could retain the regional variability in trend estimates and reduce extreme estimates, but require additional assumptions. I will briefly compare trend estimates from models with added hierarchical structure on trends to estimates from the standard model. I will use this comparison as a starting-point for a broader discussion on the most appropriate combinations of “reasonable” assumptions and model structures. BBS estimates are used in status and trend assessments and as data in further modeling studies, but the estimates vary depending on the model used to produce them, and particularly for sparsely surveyed strata. The structure and assumptions of a given model are more or less appropriate for different species and regions. However, practical and computational considerations limit the fine-tuning of the analysis to particular situations, so the published estimates are not always an optimal assessment of a species’ status and trend, even for relatively common types of inference that users of BBS estimates require. By demonstrating some aspects of this variability and some alternative models, I hope to contribute to a broader discussion on the types of model structures and assumptions that are most appropriate for the estimates published annually by the CWS and the USGS.
Smith, A. C., Canadian Wildlife Service, Canada, email@example.com
Location: Emerald Mountain - Bible Point
Presentation is given by student: No