Measurement Errors and Uncertainty in Parameter Estimates for Stock and Recruitment

Abstract

A procedure is developed for estimating stock and recruitment parameters in the presence of measurement errors. It requires an independent assessment of the ratio of environmental and measurement error variances, and provides maximum likelihood estimates of the time series of errors as well as the average stock–recruit parameters. Measures of parameter uncertainty are also provided and are incorporated into an analysis of optimum spawning stocks. This analysis indicates that much higher, or at least more variable, spawning runs should be allowed in many Pacific salmon stocks. An immediate need in salmon management is to obtain estimates of the measurement error variance, so that recent historical data can be made more useful.Key words: statistics, stock recruitment, optimum spawning, uncertainty