Adaptive Management of Stock–Recruitment Systems

Publisher Website
Google Scholar
Abstract
Management of escapements can be formulated as a problem in stochastic optimal control. Using the Ricker model as a reasonable representation of the average stock–recruit relationship, approximate optimal policies are found for cases where there is parameter uncertainty as well as environmental variation, by employing a "wide-sense adaptive dual control" algorithm. This algorithm accounts for the future information value of any escapement choice as well as the effect on immediate yield. The algorithm indicates that escapements should be deliberately varied to obtain better information for future management, provided that historical escapement levels have not been extremely variable. Such actively adaptive, or probing policies will become progressively more important in the future as managers learn how to control escapements more precisely and thus eliminate natural sources of informative variation.Key words: adaptive management, stock–recruitment, escapement

This publication has 6 references indexed in Scilit: