This 2018 paper found that a non-linear method known as empirical dynamic modeling (EDM) outperformed conventional fishery models in predicting recruitment—the number of young fish that enter a population in a given year. The study compared the accuracy of the two types of models using data for 185 fish populations from a large international database. It considered all populations for which at least 20 years of data were available. For these stocks, EDM produced more accurate predictions of recruitment than conventional “stock recruitment models” about 90 percent of the time. Accuracy was lower for longer-lived species, likely because EDM performs best with several generations of data.
The authors suggest that EDM is more accurate because it implicitly accounts for multiple factors that affect recruitment—not just stock size but also things like food availability and age structure. It has not yet been used in management, but research is underway on options for integrating it with conventional methods.
Munch, S. B., Giron‐Nava, A., & Sugihara, G. Nonlinear dynamics and noise in fisheries recruitment: A global meta‐analysis. Fish and Fisheries.