Modelling migration and aggregation: Applications in fisheries, environmental change, and evolution

Commercially important fish stocks worldwide are in danger of collapse due to fisheries exploitation compounded by environmental change. Many species exhibit seasonal migratory homing behaviour to and from spawning grounds. In this thesis, models are developed to explore the consequences of migratory behaviour and spatial structure on fisheries management and responses to environmental change. Initially, a simple population model of a migratory fish population is presented. Migration occurs between an exploited feeding area and a protected spawning area in which no fishing occurs (a marine protected area, or MPA). The model supports the notion that MPAs can be useful management tools, especially in combination with other management techniques. However, the efficacy of MPAs is likely to decrease as the propensity for migration increases. Results which include stochastic recruitment differ from deterministic conclusions, highlighting the importance of considering environmental variability in models of fisheries. A novel individual based model (IBM) is then developed, characterised by behavioural information retrieved from fish tagged with electronic data storage tags. The IBM enables more realistic simulations of movement to be undertaken, with population-scale breeding aggregations forming via simple rules governing individual behaviour. The model is used to analyse the response of aggregated breeders to range expansion following environmental change. Despite relatively rapid individual movement, breeding aggregation locations change slowly and aggregations are therefore at risk of collapse during rapid environmental change. Importantly, this slow response occurs within an homogeneous environment, highlighting the potential for responses to be limited by behavioural rather than environmental constraints. The model is then extended to explore fisheries exploitation. Fishing may cause aggregation collapse and fragmentation, but MPAs can prevent collapse and enhance yields. It is determined that large-scale measures of population health may not reveal the collapse of individual aggregations, highlighting the need for management at appropriate spatial scales.