Threshold Management Policies for Exploited Populations

Abstract

Under a threshold management policy, harvesting occurs at a constant rate but ceases when a population drops below a threshold. A simulation model of an age-structured population with stochastic recruitment was constructed with such a harvest policy with several threshold levels. Other factors were fishing mortality, recruitment, and initial biomass. The objective function was a weighted function of average yield and standard deviation over a planning horizon. First, we determined the optimal threshold given fishing mortality. Secondly, we determined optimal threshold and fishing mortality, simultaneously. In application to eastern Bering Sea pollock, a threshold management policy always increased average yield over a non-threshold policy. For the first problem, optimal threshold levels ranged from 20 to 30% of pristine biomass. For the second problem, each scenario had a unique threshold and fishing mortality, with fishing mortality slightly above the maximum sustainable yield (MSY) level and a threshold range of 25–50%. These results were robust in regard to other factors. Benefits of the threshold policy were greater with a Ricker spawner-recruit model and with higher fishing mortality. The success of the threshold management policy is due to the relatively rapid rebuilding of a population to levels producing MSY.