APPLICATION OF A SURPLUS PRODUCTION-MODEL TO A SWORDFISH-LIKE SIMULATED STOCK WITH TIME-CHANGING GEAR SELECTIVITY

Should a surplus production model be applied to a stock that exhibits pronounced age structure? Should it be applied to a stock that has exp erienced changing fishing mortality rates on fish of different sizes o ver time (i.e., changing selectivity)? These questions are of general interest to those engaged in stock assessment and of particular intere st in the assessment of North Atlantic swordfish Xiphias gladius. In a n attempt to answer them, we simulated an age-structured population, w ith fishery, similar to that of swordfish in the North Atlantic. The 3 0-year simulation included biological characteristics from the literat ure on swordfish and simulated fishing with increasing mortality of yo ung fish over time; simulated catches approximated the actual catches of swordfish from 1962 through 1991. We fit a lumped-biomass, dynamic surplus production model to summary non-age-structured data from the s imulated fishery. The resulting parameter estimates were compared to m anagement benchmarks, including maximum sustainable yield (MSY), compu ted by age-structured methods from the growth and recruitment characte ristics of the underlying simulation. The changing selectivity resulte d in a small (<10%) decrease in MSY from the beginning of the series t o its end; estimates from the production model were close to these und erlying MSY values. Nine additional population trajectories were simul ated with the same biological characteristics but other fishing histor ies; in most cases, the production model provided qualitatively correc t estimates of stock status. Two new reliability statistics appear to be of value in judging the quality of production model fits. We conclu de that for stocks similar to swordfish, the presence of strong age st ructure and moderate changes in selectivity should not proscribe the a pplication of simple production models.