Estimation of gear selectivity, growth and mortality parameters, and alternative harvesting strategies: A multiple gillnet fishery for corvina reina (Cynoscion albus) population

he corvina populations inhabiting the Gulf of Nicoya, Costa Rica, are an important part of the small scale fisheries in the region. The four primary corvina species, aguada, coliamarilla, picuda, and reina, account for more than 30% of the total catch. These high valued species are caught in drift gillnet fisheries utilizing mesh sizes 3.5, 5.0, 6.0 and, more recently, 8.0 inches.;In addition to the well known difficulties in ageing tropical fish, these artisanal fisheries are characterized by the absence of long time series of catch data, which hinders the use of methods that follow cohorts over time. Frequently, the only available data are single length frequency distributions of catch, collected sparsely in time. Furthermore, highly selective gillnet gear, almost eliminates the prospects of assuming knife-edge selection which is very commonly done as part of the estimation of population parameters. Thus, the estimation of gillnet selectivity represents an integral part of some fish stock assessment and should be a major aspect for any search for alternative harvesting strategies. Recognition of this problem and implementation of appropriate research is long overdue.;The method of estimating gear selectivity used and the selectivity model adopted will affect greatly the shape and behavior of the estimated selection function. This in turn influences population parameter estimates and the analysis of alternative exploitation strategies. A nonlinear least squares procedure is used to estimate the corvina reina selectivity curves. This method uses the ratio approach to eliminate the unknown population size distribution. The estimation of population parameters was accomplished using a new maximum likelihood approach that take into account selectivity of the fishing gear in describing the population decay. Harvesting strategies for this fishery were investigated using a size-based population model with stochastic growth, wherein fishing effort is allocated to the four gillnets. A hybrid simplex algorithm seeks the target harvest rate and its optimal allocation among gillnets.;The results indicate that estimates of growth and mortality parameters can differ by up to 50% depending on the approach used to estimate selectivity. Harvesting strategies such as...