Bootstrap Calculation of Catch-Per-Unit-Effort Variance from Trawl Logbooks: Do Fisheries Generate Enough Observations for Stock Assessments/

Abstract

Abstract The variance in trawl catch per unit effort for four fisheries along Canada's Pacific coast was examined with bootstrap analysis. The resulting confidence limits, based on the bias-corrected percentile method, indicated that the variance in catch rates varied widely among fisheries. Depending on the fishery, from 20 to 100 randomly selected observations per year were required to provide minimally sufficient precision for stock assessment.