Using LASSO regularization to project recruitment under CMIP6 climate scenarios in a coastal fishery with spatial oceanographic gradients

As climate change disrupts fisheries, scientists are interested in fisheries projections under climate change scenarios. However, projections that account for spatial oceanographic gradients use increased variable selection power and output high spatial resolution climate data are needed to improve strategic fisheries management. This study uses the least absolute squares and selection operator, a regularization technique, and improved, climate change projections from phase 6 of the Couple Model Intercomparison Project to relate Atlantic surfclam, Spisula solidissima solidissima, recruitment to climate variables. Results show a longitudinal gradient in New York State waters where western recruitment displays a negative relationship with sea surface temperature and eastern recruitment displays a negative relationship with eastward spring wind intensity. Models project that recruitment in 2050 will decrease 100% in western waters and remain sporadic, but high, in eastern waters. This study provides insight regarding surfclam responses to climate change and considerations (methodological and statistical) for improved climate-based fisheries projections.