The Future of Length-Based Stock Assessments

Beverton and Holt used mean length to estimate total mortality rate. Their estimator proved useful because of its limited data requirements. However, it requires strong equilibrium assumptions. Gedamke and Hoenig used mean length data for a series of years to relax the assumption of constant mortality rate. Additional types of data can be incorporated into the models to further relax assumptions and estimate additional parameters. Thus, including an index of recruitment relaxes the assumption of constant recruitment; including fishing effort (or catch divided by catch per unit effort) allows for the estimation of catchability coefficient, natural mortality rate, and year-specific total mortality rates; including aggregate catch rate data constrains the amount of change in estimated mortality rates. Analyzing data simultaneously from several species can reduce the number of parameters when changes in mortality rate occur synchronously across species.  The use of these generalized estimators can provide bridges between data-limited methods and methods with heavier data demands. For example, in the short term, one can use mean length, catch and catch rate data to estimate catchability and natural mortality; in the long term, one can develop a surplus production model from these data. Estimation is accomplished using Bayesian or classical methods.