Distributions of large pelagic fish, such as tuna, billfish and shark, are related to water temperature and a range of other climatic variables and biotic interactions. Having an understanding of how these large open ocean predators are distributed in relation to climatic variables and the potential affected of projected climate change on their future distributions is of ecological and economic significance (Chapter 1). A variety of modelling approaches have been used to predict species current and future distributions in relation to climate variables. Species distribution (or environmental niche) models have been widely applied to numerous terrestrial species, but there are far fewer marine applications. I evaluate the theoretical (and practical) limitations and advantages of applying these models in the marine realm (Chapter 2). One practical limitation identified, in applying species distribution models to fisheries catch data on large pelagic fish species, was sampling bias. I tested how sampling bias in catch data affects climatic bias and model predictions by comparing data and model predictions with habitat suitability models that incorporate temperature preference data derived from fish fitted with electronic tags (Chapter 3). This sampling bias evaluation showed habitat suitability models were likely to be less biased than species distribution models, but both models may be hindered by some degree of bias and thus bias correction could not be properly evaluated. However, there was greater agreement between the two independent models of most species targeted by the fishery (i.e. yellowfin tuna, striped marlin, bigeye tuna and albacore tuna) relative to non-target species (i.e. blue shark and southern bluefin tuna).
Off the east coast of Australia, climate models predict the ocean will warm on average by 0.6 degrees by 2030 and 1.5 degrees by 2070 (Chapter 4). Based on these temperature projections, and the assumption that species will maintain their existing thermal preferences, species distribution models and habitat suitability models were used to predict how the potential distribution of nine pelagic fish species may respond to increasing water temperatures (Chapter 4). Baseline and future projections of species potential distributions were made using sea surface and subsurface temperatures from eleven General Circulation Models (GCMs) used in the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4). Projected shifts in species ranges were compared at different points in the range (i.e. edges and centres), seasons (summer and winter) and in the near term (2030) and long term (2070) under medium (a1b) and high (a2) future emission scenarios. Projections showed trailing edges consistently shifted faster than leading edge, all species shifted south faster in 2070 than 2030, and there was little difference in projected shifts among species and between seasons (Chapter 4). Projected changes in the distributions of five tuna and billfish species were used to calculate changes in the future catch of the Eastern Tuna and Billfish Fishery (ETBF) (Chapter 5). This information was used in economic fisher location choice models to project the economic impact on fishers that operate at the tropical northern edge of this fishery (Chapter 5). Southward shifts of target species generally resulted in declines in the profitability of fishers due to reductions in the catch and fishers opting not to fish more frequently. These projected economic impacts can be used to guide the development of potential adaptation strategies that may in turn minimise future profit loss and could be relevant to other fisheries in tropical regions that are at risk of potential declines in catch and profit under climate change.