Statistical Tests of Salmon Fishery Contribution Rates Based on Coded-Wire-Tag Recoveries

摘要

Statistical procedures for testing hypotheses concerning fishery contribution rates based on recoveries of coded-wire-tagged (CWT) salmon are described. Observed recoveries of CWT fish from one or more release groups to one or more fisheries are modeled with compound probability distributions, one component reflecting natural variation and the other representing the catch sampling variation. We describe a flexible bootstrap testing procedure suitable for a wide range of hypotheses, verify its accuracy by simulation of compound probability distributions, and apply it to a set of real CWT recoveries.