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. How-ever, projections that account for spatial oceanographic gradients use increased variable selection power and output high spa-tial 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. Re-sults 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.