The dynamics of pelagic ecosystems is becoming increasingly relevant, yet the contribution of mid-trophic level pelagic predators is typically understudied. Describing the habitat preferences of pelagic fish can improve inputs for ecosystem modelling, as well as help fisheries management strategies adapt to a dynamic and changing ocean. Using a combination of three fishery dependent and independent datasets, we examined the habitat preferences of two contrasting pelagic species, dolphinfish (Coryphaena hippurus) and yellowtail kingfish (Seriola lalandi) off the east Australian coast. The three datasets are spatially confined to the New South Wales coast and contain a total of 16,589 dolphinfish and 15,971 yellowtail kingfish presences collected between 1972 and 2012.

Binomial generalised linear models with a logit link were used for each dataset to assess the probability of fish presence. Predictor variables retained in the model include sea surface temperature, sea surface height, chlorophyll-a, eddy kinetic energy, temperature fronts, latitude, bathymetry and month. The model outputs show complex relationships between predictor variables and fish presence, depending upon the season and bathymetry. Probability distribution maps, with associated confidence intervals, reveal the seasonal habitats of dolphinfish and yellowtail kingfish as well as indicate habitat overlap between the species. 

Determining the distributions for highly migratory species is complex.  The use of fishery-dependent datasets with and without estimates of fishing effort will be discussed.  These habitat models can be used in conjunction with forecasting products to create seasonal (9 month forecast) and long-term (2062-2073 forecast) forecasts of fish distributions.