Download the publication
Summary
Leatherback turtles (Dermochelys coriacea) are the most frequently reported turtle bycatch in New Zealand commercial surface longline fisheries. Leatherback turtles are protected under the Wildlife Act 1953. This project updated the fishery captures of leatherbacks in New Zealand waters up to, and including, the 2022–23 fishing year and attempted to identify any temporal changes in fishing practices and/or catch composition associated with changes in leatherback bycatch.
Most surface longline fishing effort targeted southern bluefin tuna (Thunnus maccoyi), followed by bigeye tuna (T. obesus), and swordfish (Xiphias gladius). The surface longline fishery started in northern waters around October and moved south as the season progressed and waters warmed, returning to more northern waters as winter approached. In summer months, the southern bluefin tuna fishery extended to the southeast of the South Island, while the swordfish fishery occurred on the southeast North Island and west coast South Island. Bigeye tuna catches were centred further north, rarely extending beyond the Bay of Plenty and East Cape.
Leatherback captures have been centred in the Bay of Plenty. The leatherback spatial ‘hotspot’ and season (the ‘hotspot’) occurred between latitudes of 36° S and 38°S from January to April. Bigeye tuna, and then swordfish fishing, had the greatest spatial and temporal overlap with the leatherback hotspot. The southern bluefin tuna fishery had very little overlap. About 75% of the bigeye tuna catch, 80% of the swordfish catch, and almost all the southern bluefin tuna catch, ware taken outside of the leatherback hotspot. The greatest leatherback captures were reported from 2021 when fishing effort was relatively low, but more focused on the east coast North Island than usual.
The fisheries characterisation using reported commercial catch and effort data indicated the strongest interaction was between leatherbacks and the fishery targeting bigeye tuna. An alternative analysis based on clustering of catch compositions produced a different result, with the strongest association being between leatherbacks and swordfish. This difference was because the reported target species did not always accurately describe the catch composition. The catch composition analyses did not isolate the Bay of Plenty region as a specific fishery subunit, meaning there was nothing apparently unique about the fishing in and around the leatherback hotspot.
A Generalised Additive Model (GAM) was developed to investigate the potential reasons for trends in leatherback captures in 2021. The GAM analysis was updated to 2023 with additional environmental and fisheries variables included and restricted to the east coast North Island. The updated GAM explained more of the variability in leatherback bycatch probability, but with a different set of predictor variables. The variables used were proximity to steep sea surface temperature (SST) gradients (fronts), mixed-layer depth, water depth, strength of the west-to-east current, number of hooks between floats (an alias for fishing depth), number of light sticks used between floats, and moon phase. Further GAMs predicting leatherback occurrence and fish catch rates from environmental conditions found the closest association was between leatherbacks and swordfish. A close association between swordfish catch rates and leatherback captures is consistent with international leatherback capture mitigation focusing on swordfish target fisheries.
Comparison of vessels reporting and not reporting leatherbacks was made and found the greatest difference was in location fished. Overall, fishing location was the most persistent and important factor determining the likelihood of a vessel capturing a leatherback.
Publication information
Dunn, M.R., Finucci, B., Sutton, P., Pinkerton, M.H. (2024). Characterising surface longline fishing fleet behaviour in relation to leatherback bycatch. NIWA Client Report 2024214WN. 80 p.