Critique of the Leslie Method for Estimating Sizes of Crab and Lobster Populations

The Leslie method can be used to estimate the size of a decapod population by regressing catch per trap haul on cumulative catch of the fleet at several time intervals within a fishing season. The data requirements are modest and the calculations simple; however, demands for a time series of catch–effort data and for the capture of a large portion of the population entail assumptions that are easily violated. These assumptions are that the target species has constant catchability over the sampling period, the fishing effort is distributed uniformly over the fishing ground, the fishing methods do not change, the target population is closed, and landings and effort arc reported correctly. The assumptions were investigated by reviewing the literature for examples of deviations, and by simulating the deviations in a simple model. Petersen tag–recapture studies have been used to verify the accuracy of Leslie estimates. However, important assumptions are common to both methods, and violations bias stock size estimates in the same direction. Also, the cause of bias cannot be diagnosed from shape of the Leslie regression. Without knowing the cause, a correction cannot be made, However, some biases can be corrected with supplementary data if their cause is identified. Mean catch per unit effort used as an index of stock abundance is an alternative to the Leslie method. Simulations showed this method to be equally or more accurate than the Leslie method for several scenarios of decapod catchability and recruitment.