Commercial fishing patterns in Australia's northern prawn fishery and implications for management

This thesis is focussed on the variation in fishing patterns realised by commercial fishers in Australia's Northern Prawn Fishery (NPF). An analytical framework is developed for use as a tool to help fishery managers understand commercial fishers' decision making and the subsequent fleet dynamics in commercial fisheries and their implications for fishery management. The thesis provides a framework based on Markovian theory that permits a description of the spatial and temporal allocation of fishing effort, and allows for the simulation of fleet movement in response to fishery management policy changes. The relocation of fishing effort between fishing grounds, over time, defines the spatial and temporal dynamics of fleet movement. Data on fishing locations of individual vessels are used to obtain a spatial and temporal series of transition probabilities, which are then used for describing and forecasting commercial fishing behaviour. These transition probabilities represent the likelihood of effort allocation to selected fishing grounds, over any two consecutive time periods. The ordinary Markov model relies solely on the historical transitions for the entire fleet across particular fishing grounds during a selected fishing period. However, since the transitions made by fishers are the outcome of economic behaviour and decision making in fishing, the variation in transition probabilities and the fishing patterns realised by commercial fishers are explained in the thesis using the multinomial logit (MNL) approach. To capture fishers' response to fishery policy, the ordinary Markov model is enhanced using estimates from the MNL model. This enriched Markov model, referred to as the MNL Markov, requires data on policy variables and characteristics of fishers and fishing grounds. Although not primarily policyoriented, the effects of shortening the fishing season and closing selected fishing grounds are illustrated by inspecting the marginal effects of policy variables from the MNL Markov and evaluating the limiting distribution of transition probabilities in the ordinary Markov model. The results obtained in the thesis suggest that there is scope for fishery managers to use such models to forecast changes in commercial fishing patterns in the NPF that result from management changes such as shortening the fishing season and closing selected fishing grounds.