Sampling Bias of Hook-and-Line Gear Used to Capture Rainbow Trout in Gertrude Creek, Alaska

Abstract

Hook-and-line gear is commonly used for monitoring the size and abundance of resident fish populations in southwest Alaska. The selectivity of the gear, however, may contribute to bias in the catch, making it difficult to detect changes in population structure over time. This study assessed the potential bias in sampling a population of rainbow trout Oncorhynchus mykiss with angling gear. A bidirectional weir was operated on Gertrude Creek from 1997 to 1999 to intercept the annual upstream spring migration of rainbow trout into the stream. Captured fish were measured and tagged to establish a population of marked individuals with a known length distribution above the weir. The marked population was then sampled with angling gear on three occasions during each of three summers. A variety of terminal tackle was used, and types were selected to maximize the catch of rainbow trout. Size selectivity was evaluated by use of an exact Kolmogorov–Smirnov (KS) test of the hypothesis that the length distribution of all marked fish was equivalent to that of the subset of marked fish recaptured by anglers during a single recapture event. Test results were not significant for any of the recapture events, providing strong evidence that length selectivity did not exist for fish within the size range of the marked population. An exact KS test was also used to test length distribution equality between the marked population and all fish captured during an angling event. All of these KS tests were significant, suggesting that the marked group was a size-biased subset of the rainbow trout population in the drainage. The results of this study support the use of angling gear as a potentially viable sampling method for monitoring the size composition of rainbow trout in small southwest Alaskan streams.