Comparative performance of ADAPT and Laurec-Shepherd methods for estimating fish population parameters and in stock management

Abstract

Monte-Carlo simulations are used to compare the performance of two methods in fitting populations and fishing mortalities to error-free catch-at-age observations and independent indices of stock size with a variety of specified error-structures. The two methods are the Laurec–Shepherd method that has been extensively used in the ICES arena and the ADAPT method that has been widely used in Eastern North America. We use simulated data sets with known error-structures to explore numerically the behaviour of the two algorithms in situations of uncertainty with the additional information used to tune the series. We find the ADAPT method to yield more precise and less biased population estimates than the Laurec-Shepherd method under the simulated error distributions, and in a series of management simulations we show that the use of the ADAPT method results in reduced inter-year TAC variability and in the improved attainment of a fixed fishing mortality management target. However, under conditions of high uncertainty in the indices of stock size, application of either method in a management simulation resulted in fishing mortalities that substantially exceeded the management targets.