Influence of upwelling on movement of southern bluefin tuna (Thunnus maccoyii) in the Great Australian Bight

Large pelagic predators move quickly in and out of local ecosystems that may be separated by long distances: their trophic effects are determined by their behaviour while present. To investigate movement and local residence times of one such predator we implanted 29 acoustic tags into juvenile southern bluefin tuna (Thunnus maccoyii) (SBT) in the Great Australian Bight. We used acoustic detectors at a reef known to attract tuna and detected fifteen SBT on 941 occasions over 62 days. SBT were tagged at the reef, 40 km, and 120 km distant. A total of 100% of local and 60% of SBT tagged 40 km away were subsequently recorded at the reef. Presence and absence was related to an upwelling event. Water temperature decreased just after SBT departure from the monitoring region. The immediate area was aerially surveyed 22 times for SBT schools during the experiment. We combined aerial survey observations with computer simulation, calibrated against field studies of SBT movement, to test the hypothesis that tuna could be well simulated by a correlated random walk throughout the area of known occurrence. The most plausible explanation for the observed behaviour was short-term (hours) fidelity to schools combined with medium-term (weeks) fidelity to bathymetric features. The present study illustrates how dynamic models aid interpretation of experiments designed to understand trophic effects of large pelagic predators. © CSIRO 2007.