Predicting interactions of sperm and killer whales with industrial fisheries in the Southern Ocean: a spatiotemporal modelling approach for conflict mitigation

Toothed whale depredation of fish caught on fishing gear raises socioeconomic and conservation concerns. It can lead to substantial losses for fishers and impacts on depredating species, but effective solutions remain limited. In this study, we implemented a spatiotemporal modelling approach to predict depredation occurrence and intensity, based on natural distribution of predators involved and fishing practices, to support mitigation strategies. Using 11 years of data from the Patagonian toothfish ( Dissostichus eleginoides ) longline fisheries operating around Crozet and Kerguelen islands, and generalized additive models (GAMs), we assessed the environmental and operational factors influencing depredation by sperm whales ( Physeter macrocephalus ) and two killer whale ( Orcinus orca ) ecotypes: Crozet and Type D. All models indicated strong seasonal patterns in depredation, particularly for sperm whales, whose presence decreased in winter and was primarily driven by high abundance of large toothfish. Crozet type killer whales were associated with shallow, low-slope areas near the continental shelf, whereas Type D killer whales were more frequent in deeper waters and near seamounts, suggesting a more offshore distribution. Longer soak times and line lengths increased killer whale depredation, likely by increasing gear detectability. Crucially, vessels that moved more than 70 km after a depredation event significantly reduced the likelihood of further interactions with both predator types. The results suggest spatial overlap between fishing grounds and whale-preferred habitats, but highlight clear depredation hotspots within that overlap. Avoiding these areas provides fishers and managers with easy-to-implement, cost-effective options for mitigating depredation while maintaining the socio-economic viability of the activity.