Modelling the behaviour of tunas in relation to their environment

The diversity of fields covered by this thesis is most apparent in Chapter 1, which gives an overview of the underlying philosophy and present level of understanding of the research. The role of modelling in science and the capabilities of different types of model are discussed. The academic sub-disciplines within which this thesis may be categorised are defined. The ecology of tunas is described as are satellite sensors that have been or could be used for fisheries applications. Case studies of such applications are reviewed. For Chapter 2, I analysed a large data set from the New Zealand surface longline fisheries for tuna, in order to identify whether and at what spatial scales tunas are aggregated. I applied methods that have not previously been used in the analysis of longline data and determine that these adult tunas are often caught in loose schools within a larger, sub-mesoscale area in which they are aggregated. In Chapter 3, I review present knowledge in tuna physiology and sensory biology, and develop and apply analytical models to determine reaction distances. This was an essential pre-cursor to the development of theoretical models in Chapters 4 and 5. The optimal foraging model had been previously anticipated in the literature but had not previously been realised. It provides a framework for addressing a long-standing and still unresolved debate as to what is 'controlling' tuna behaviour at ocean fronts. In Chapter 5, I describe and apply an important methodological synthesis for fisheries oceanography. While still requiring further development, this work nonetheless proves that it is possible to combine behavioural models for fish with whole-ocean circulation and production models that incorporate data from satellite sensors. Chapter 6 discusses further work that might follow from this thesis. Successful proposals have been made to the European Space Agency and the New Zealand Foundation for Research Science and Technology, which may now build on the work carried out here. I strongly advocate that further work to develop models linking fish distributions with environmental properties should include sea-going studies of pelagic trophic dynamics in specific areas of interest.