Multispecies spatial assessment models for the British Columbia groundfish trawl fishery

A multispecies spatial model is proposed for evaluation of trawl management strategies and tested with data from British Columbia groundfish. Minimum stock sizes are estimated for 16 species using spatial catch per unit effort data; these data indicate that only small proportions of most stocks are exposed to fishing at any time and that exchange rates of fish between the grounds and other habitats are rapid. A gravity model for spatial effort allocation predicts substantial side effects of marine protected areas or discarding of particular species on nontarget species/grounds due to redistribution of fishing effort. Multispecies, multiground delay-difference models are used in conjunction with the gravity model for spatial effort distribution to predict long-term population responses to various regulatory strategies. This simulation suggests that species composition and spatial structure of the fishery are changing slowly due to reduction in abundance of long-lived species, but rapid collapses of target stocks are unlikely due to spatial redistribution of effort in response to local declines. We question whether quota management (individual vessel quotas) can be safely implemented due to extreme uncertainty about current stock sizes and suggest that a better strategy may be to simply denominate fishing rights as individual effort quotas while providing protection for sensitive stocks through carefully placed marine protected areas.