Contrasting distribution and biomass trends of two sympatric redfish (Sebastes) species off Southeast Greenland

Sustainable harvest of fish stocks requires regular monitoring of abundance and biomass in space and time. In Southeast Greenland, two Sebastes species, beaked redfish ( Sebastes mentella ) and golden redfish ( S. norvegicus ), coexist and have been harvested together. However, the species differ greatly in stock status as beaked redfish are depleted, while golden redfish have a surplus of exploitable biomass. Detailed habitat affinities of both species along the continental shelf and slope are largely unknown despite documented differences in depth distribution. Yet, such information is crucial for effective management and conservation given policies greatly constraining beaked redfish catches. Furthermore, investigating the distribution of these predator species is important to understand the changing sub-Arctic ecosystem of Southeast Greenland. Here we combine three fisheries survey time series to compare and predict distribution patterns of population density and identify their drivers. Spatiotemporal generalized linear mixed models were implemented in the sdmTMB framework. The most appropriate models were identified based on 5-fold cross-validation and ecological realism of partial effects. Our models confirmed that 1) spatial aggregations are common for both species, however, beaked redfish showed a broader spatial autocorrelation compared to golden redfish, shown in greater Matérn spatial range (43 km vs 20 km), 2) both species are widely distributed over the shelf and slope but with local/regional hotspots that remain relatively constant over time despite interannual changes in biomass. Golden redfish displayed larger spatiotemporal variability in distribution compared with beaked redfish, and 3) beaked redfish occupy deeper areas compared to golden redfish indicating vertical segregation. Depth and density of Atlantic cod ( Gadus morhua ), which is another important predator in the area, were the most important drivers for beaked redfish distribution, while depth and temperature were most important for golden redfish. These findings highlight the importance of distribution models to understand co-occurrence patterns of closely related species and provide opportunities to improve fisheries management.