Training fish for restocking: refuge and predator training in the hatchery has limited benefits for a marine fish

Large numbers of hatchery-reared fish are released in stocking programmes; however, success is limited by high mortality. Predation is seen as the main cause of deaths but might be reduced by training fish before release to avoid predators and/or use refuge. In this study on a potential restocking species, yellowfin bream Acanthopagrus australis, the effects of predator training and refuge on the behaviour of fish in the hatchery were tested. In the first experiment, juvenile bream were exposed to predatory mangrove jack (Lutjanus argentimaculatus) fed exclusively on bream flesh while housed in tanks with and without refuge. Predator training altered fish behaviour when fish were re-exposed to predators, but the effects were subtle and varied between groups of fish. In contrast, refuge created strong and consistent changes in behaviour, significantly slowing down the amount of time that fish took to consume food. A second experiment focused on the effects of refuge. Bream were trained to use artificial seagrass or house bricks as refuge and then exposed to mangrove jacks in a laboratory predation experiment. When refuge was available, fish significantly slowed down their feeding rate. There was a small, transient increase in survival for fish given seagrass refuges, but this was irrespective of whether the bream were trained to use refuge. The results of this study indicate that the use of refuge may be innate and the benefits of refuge may be available to naive hatchery-reared fish or fish trained to use refuge shortly before release. This suggests that there is potential to improve post-release survival of fingerlings without time-consuming and expensive hatchery training.