Dynamic spatial heterogeneity reveals interdependence of marine faunal density and fishery removals

The spatial structure and dynamics of populations, their environment, interacting species, and anthropogenic stressors influences community stability and ecological resilience. Despite the importance of spatial processes in ecological outcomes and increasing desire to implement ecosystem-based management, fine-scale spatial dynamics have been rarely incorporated in marine fisheries management. However, advances in population modeling and data availability provide the necessary ingredients to address this disconnect between the fields of ecology and fisheries. We used random forests and spatial indices to quantify spatial heterogeneity and dynamics of US west coast demersal marine faunal density (biomass of a community or assemblage per unit area) and the total removals (catches plus discards) from the system by the groundfish bottom trawl fishery from 2002 to 2017. We expected spatial heterogeneity of removals and density to increase following implementation of depth and habitat closures - due to proximally increasing density gradients and fishing-the-line - and following catch shares because of fleet consolidation and behavioral consequences of eliminating the race to fish. However, we found mixed responses, where at the broadest community levels spatial variation in removals and density declined with habitat closures, while spatial autocorrelation of removals increased with habitat closures and declined with catch shares. Our results reveal a complex interdependence between spatial distributions of faunal density and fishery removals that has been absent in previous studies focusing on catch only, and shows how these patterns are shaped by marine policy. Values of spatial variation of density and removals were positively correlated within year (i.e., each responded with the same sign and timescale), while there was also evidence that interannual changes in the spatial variation of removals among years led those of density by one year (i.e., increases in patchiness of removals were followed by increased patchiness of density). These results hint at the presence of a stronger than expected top-down effect of fishing, given that this system is considered to be dominated by strong bottom-up effects of environmental variation on primary and secondary productivity.