Near real-time spatial management based on habitat predictions for a longline bycatch species

Abstract  Southern bluefin tuna (SBT), Thunnus maccoyii (Castelnau), is a quota-managed species that makes annual winter migrations to the Tasman Sea off south-eastern Australia. During this period it interacts with a year-round tropical tuna longline fishery (Eastern Tuna and Billfish Fishery, ETBF). ETBF managers seek to minimise the bycatch of SBT by commercial ETBF longline fishers with limited or no SBT quota through spatial restrictions. Access to areas where SBT are believed to be present is restricted to fishers holding SBT quota. A temperature-based SBT habitat model was developed to provide managers with an estimate of tuna distribution upon which to base their decisions about placement of management boundaries. Adult SBT temperature preferences were determined using pop-up satellite archival tags. The near real-time predicted location of SBT was determined by matching temperature preferences to satellite sea surface temperature data and vertical temperature data from an oceanographic model. Regular reports detailing the location of temperature-based SBT habitat were produced during the period of the ETBF fishing season when interactions with SBT occur. The SBT habitat model included: (i) predictions based on the current vertical structure of the ocean; (ii) seasonally adjusted temperature preference data for the 60 calendar days centred on the prediction date; and (iii) development of a temperature-based SBT habitat climatology that allowed visualisation of the expected change in the distribution of the SBT habitat zones throughout the season. At the conclusion of the fishing season an automated method for placing management boundaries was compared with the subjective approach used by managers. Applying this automated procedure to the habitat predictions enabled an investigation of the effects of setting management boundaries using old data and updating management boundaries infrequently. Direct comparison with the management boundaries allowed an evaluation of the efficiency and biases produced by this aspect of the fishery management process. Near real-time fishery management continues to be a realistic prospect that new scientific approaches using novel tools can support and advance.