Forecasting fish recruitment using machine learning methods: A case study of arabesque greenling

Abstract

Fish recruitment prediction is a central topic in fisheries science. We use a machine learning method for predicting the recruitment of Arabesque greenling in Hokkaido, Japan. Biological, fisheries-related, and environmental factors were included in the predictive models as feature variables. A gradient boosting model (GBM) showed better predictive performance compared with a linear regression model (LRM) and a random forest model (RFM) in terms of relative bias and relative root mean square error for recruitment prediction in the last 5 years. The most influential feature for GBM was spawning stock biomass in the last year, followed by the fishing rate for older fish. The sea surface temperature (SST) was a very weak predictor in the GBM but was the most influential feature in the LRM. The difference among models suggests the importance of nonlinearity and variable interactions. Machine learning methods will greatly contribute to fish recruitment forecast and thereby sustainable fisheries management.