Coping with uncertainty in ecological advice: lessons from fisheries

All ecologists are familiar with uncertainty, at least at the level of whether they should reject a null hypothesis. Uncertainty is, however, pervasive and its characterization is essential if we are to understand our effects on ecosystems. Traditional fisheries management has a poor track record for confronting uncertainty, but most management authorities are now committed to a precautionary approach. As a result, some of the most interesting methods for taking account of uncertainty in ecological systems have been developed by fisheries scientists. These methods evaluate the relative performance of different management procedures with the use of mathematical and statistical models that synthesize knowledge and speculation about the system of interest. Recent advances in computer-intensive statistics have made it possible to combine this approach with model fitting, so that the uncertainties and risks associated with different outcomes of management can be quantified. We show how this methodology can be applied to a range of ecological problems where the advice that scientists provide to decision makers is likely to be clouded by uncertainty.