Explanations and predictions in fisheries science - problems andchallenges in a historical and epistemological perspective

摘要

The state of explanations and predictions in fisheries science is discussed with particular emphasis on the extent to which fish stock assessments utilise basic scientific methods and knowledge. It is concluded that in general fish stock assessments do not fully utilise available oceanographic and biological knowledge. The problem of testability, and the potential of using long time series of biological and physical data in testing basic hypotheses in population dynamics, have not got the attention they deserve. For increasing the predictive power, our theories should make testable assertions about structural and relational properties which penetrate deeper and are of a higher universality than the observed pattern of events ("appearances") we try to explain. Evolutionary thinking may help in specifying possible hypotheses explaining main features of the population dynamics of fish stocks and in gaining a deeper understanding of how exploitation may perturb the stocks' adaptation to their environment. We are far from any limits set by nature on predictability, but our predictions should take account of the probabilistic character of many physical and biological processes by considering the probabilities as properties of the whole physical/biological system under investigation.