Effects of catchability variation on performance of depletion estimators: application to an adaptive management experiment

Depletion methods use sequential catches to estimate abundance, and commonly assume constant capture probability. Violation of this assumption results in abundance estimates that are biased low, and confidence bounds that indicate unwarranted precision. I used simulation to compare performance of three alternative estimators that explicitly consider catchability change over passes. If catchability declined, non-constant catchability models failed to reduce bias, but better characterised uncertainty. I used non-parametric methods to examine time and treatment effects on the precision and bias of abundance estimates from depletion data for a large management experiment on the Bridge River, BC. Catchability increased over time, suggesting a concurrent change in bias. The magnitude of the change in bias was a function of the decline in catchability within depletion experiments. Because estimates of the decline in catchability were uncertain, it was difficult to assess the magnitude of change in bias.