A dynamic programming model of fishing strategy choice in a multispecies trawl fishery with trip limits

Dynamic programming was used to model targeting decisions made by bottom trawling vessels in the US. west coast groundfish fishery, under management-imposed limits on landings of each target species (trip limits). A model of choice of assemblage (bottom rockfish (Sebastes sp.) versus deepwater Dover sole (Microstomus pacificus) complex) within a fishing trip was parameterized with data from an observer study conducted in 1988 through 1990. The model predicted that the vessel would fish the bottom rockfish strategy exclusively without limits but would switch between strategies several times under restrictive trip limits. That higher limits increased switching was consistent with actual landings from trips made by the same vessels under the same trip limit regimes, although the actual landings were more variable. Different trip limits or different market prices for the limited species changed the predicted decisions. Changing the cost of fishing each strategy, probability of a premature trip ending, tow duration, and time between tows also changed the predicted decisions, but the input parameters had to be well outside the range of values observed in the fishery.