Summary of bycatch in WCPFC longline fisheries at a regional scale, 2003–2021

The Western and Central Pacific Fisheries Commission (WCPFC) has a responsibility to assess the impact of fishing on non-target species. In this report, we estimate the bycatch of the longline fishery operating in the WCPFC Convention Area for the period 2003 to 2021 . The estimates cover the full range of finfish, billfish, shark and ray, marine mammal and sea turtle species that have been recorded in longline observer data. The estimates do not cover domestic longline fisheries in the west-tropical sector of the WCPFC Convention Area, or former shark-targeted fisheries in the EEZs of Papua New Guinea and Solomon Islands. It is difficult to obtain reliable estimates of WCPO longline catches from observer data, given the low levels and imbalanced nature of observer coverage, and additionally the low coverage of available aggregate effort data disaggregated by hooks between floats in the mid-2000s. Observer coverage has been particularly low in the north west Pacific. As such, the catch estimates for the region north of 10°N, and consequently the catch estimates for the WCPFC Convention Area as a whole, are unlikely to be reliable and should be viewed in that context. The catch rate models do not appear to adequately capture targeting behaviour, or spatial variation in catch rates more generally. There may be sufficient observer data to consider explicitly capturing spatial variation in catch rate models in the next iteration of this work, given the recent increases in spatial coverage of available observer data.

The Scientific Committee is invited to: • Note that hooks between floats (HBF) is now estimated for reported effort with no HBF information. This allows uncertainty in estimated HBF to propagate through to uncertainty in estimated catches; • Note the difficulties in robust estimation of longline catches from observer data, particularly for rarely caught species, given the low levels and imbalanced nature of observer coverage, and for some years the low coverage of available L BEST HBF data; • Note that earlier work suggests that the trends in estimated catch rates are more reliable than the magnitudes of the estimated catches; • Note that there should be sufficient available observer data to enable a substantive revision of the catch rate models for the next iteration of this work, assuming a timely return of observer coverage to pre-COVID levels; • Note that enhancement of the level and spatial coverage of observers through human and electronic monitoring approaches would improve the estimation of the catch rate models and catches.