Statistical Evaluation of a Large-Scale Fishing Experiment Designed to Test for a Genetic Effect of Size-Selective Fishing on British Columbia Pink Salmon (Oncorhynchus gorbuscha)

Publisher Website
Google Scholar
Abstract
Since 1950, stocks of British Columbia pink salmon (Oncorhynchus gorbuscha) have shown up to a 34% decrease in mean adult body weight, causing significant reduction in economic value of commercial harvests. Previous research suggests that this trend is due to size-selective harvesting of large fish, but changes in oceanographic conditions are a plausible alternative. Corrective action by management agencies requires that the true causal mechanism be identified. We therefore examined several possible designs for a large-scale fishing experiment devised to test the size-selective fishing hypothesis. These designs would generate accurate and precise field estimates of the heritability (h2) of growth rate, which is important because it, in combination with the selection differential (D) caused by fishing, determines how rapidly body size changes. Monte Carlo simulations showed that block designs with three to six spatial replicates and relatively short durations generated high statistical power. For example, for h2 = 0.22, D = 0.25 kg, and four spatial replicates, an 8-yr experiment resulted in power = 0.87, which gave a SE < 0.10 for h2 = 0.22. We conclude that some experimental designs have good potential to test the possible effects of size-selective fishing on mean adult size of British Columbia pink salmon.