Informing improved management of mixed fisheries through comparative modelling of fleet dynamics

Abstract

Mixed fisheries, where several species are caught in the same fishing operation, are ubiquitous and a major challenge for fisheries management. Overexploitation occurs in mixed fisheries where fishers catch species for which they have no quota and then discard. Understanding how these ‘technical interactions’ lead to decisions by fishers about where to fish in response to management is key to addressing the sustainability of mixed fisheries. The objectives of this thesis were to i) improve understanding of how fishers exploit different populations in space and time, and ii) develop a comparative framework for modelling location choice to better predict how fishing effort is allocated in response to population and fishery dynamics subject to management interventions. Addressing exploitation in space and time, Chapter 2 developed a spatiotemporal dimension-reduction framework to understand how community and fishery dynamics interact to determine species composition. We identified where species can be effectively decoupled through changes in spatial fishing patterns. Chapter 3 developed a highly resolved discrete-event simulation model of mixed fisheries to understand how data source and resolution impact inference on mixed fisheries interactions. To improve prediction of effort allocation, Chapter 4 compared process-based and statistical location choice models from theoretical and applied perspectives. We found theoretical equivalences among simplified models but important differences in application. By implementing alternative location choice models as operating models in mixed fishery management strategy evaluation (MSE, Chapter 5), we demonstrated significant impact on inferred sustainability of given management plans. This thesis advances the scientific basis for mixed fisheries advice by a) providing a basis for understanding co-occurrence and separability of species, b) critiquing the utility of different sources of data to support management, c) providing a comparative understanding of location choice models in theory and application, and d) demonstrating how these can be used in an MSE framework capturing structural model uncertainty.