Evaluation of models and assumptions for closed population abundance estimators from from cutthroat trout mark-recapture data

Abstract

The goal of this project was to improve cutthroat trout Oncorhynchus clarki mark-recapture experiments in southeast Alaska lakes. A mark-recapture experiment was conducted at Hasselborg Lake, Admiralty Island, southeast Alaska. The sampling design, nine abundance models and their assumptions were evaluated. Evaluation of model assumptions led to conclusions that the population was closed and no tag loss occurred. Probability of capture varied due to effects of time (sampling occasion) and heterogeneity (differences among individual fish) during 1992, and time and behavior (capture effects) during 1993. Chao M(h) was selected as the best abundance estimator for 1992, and Chao M(t) for 1993. Evaluation of sampling design led to several conclusions: angling was the most effective gear type; catch per unit effort decreased with depth; dispersal distance was not related to time; length distributions were depth specific; and gear selectivity took place.