Designing Adaptive Management Policies

Abstract

It does not take much practical experience with stock-assessment methods and models to realize that the available data usually cannot provide unique and reliable estimates of the best management policy. Assessment calculations based on historical data often reveal gross uncertainties about sustainable yields, optimum effort levels, effects of various stock enhancement measures, and so forth. These uncertainties are not just a matter of annoying imprecision in estimates of a few parameters such as natural mortality rates; usually they reflect a fundamental lack of experience about how the stock behaves under alternative policy options. In other words, the prediction of the best policy choice is usually either a gross extrapolation about how the stock will respond to conditions that have never occurred in the available historical data, or is only one of a collection of alternative predictions that each appear equally credible based on the data