Distributions of threatened skates and commercial fisheries inform conservation hotspots

Abstract

Bycatch in commercial fisheries is a pressing conservation concern and has spurred global interest in adopting ecosystem-based management practices. To address such concerns, a thorough understanding of spatiotemporal relationships among bycatch species, their environment and fisheries is required. Here we used a generalized linear mixed model framework incorporating spatiotemporal random effects to model abundance patterns for 3 skate species caught as bycatch in commercial fisheries (thorny skate Amblyraja radiata, winter skate Leucoraja ocellata and smooth skate Malacoraja senta), as well as 10 target species on the Scotian Shelf, NW Atlantic. Spatiotemporal estimates of relative abundance for at-risk skates within the years 2005-2015 were modelled from research trawl survey data and overlaid with those for target species to identify hotspots of bycatch risk. In addition, abundance estimates for at-risk skates within the years 1975-1985, a period of higher stock abundance, were used to identify areas of previously important habitat. Historically, skate species densely occupied areas near Sable Island and Banquereau Banks, Georges Bank and the Bay of Fundy. Bycatch hotspots between at-risk skates and commercial targets were identified in regions across the Scotian Shelf. These hotspots were independently validated by predicting species presence from at-sea observer data that monitor skate bycatch directly. We discuss spatial relationships between target and bycatch species, highlighting limitations of at-sea observer programmes that this method helps to address. This framework can be applied more broadly to inform ecosystem management and priority areas for conservation or fisheries regulation.