Seabird bycatch in the Brazilian pelagic longline fishery and a review of capture rates in the southwestern Atlantic Ocean
The southwestern Atlantic Ocean is an important foraging ground throughout the year for several albatross and petrel species. Longline fishing fleets in the region currently pose the main threat for this group of seabirds at sea, and conservation measures are urgently required. We present information on bycatch rates of seabirds in the Brazilian domestic pelagic longline fishery from 2001 to 2007, and review bycatch rates reported for the demersal and pelagic longline fisheries in the southwestern Atlantic Ocean. Overall seabird capture rate for the Brazilian pelagic longline fleet during 63 cruises (656 sets and 788 446 hooks) was 0.229 birds per 1000 hooks, varying from 0 to 0.542 according to season. Capture rates were higher between June and November (cold season) and affected mainly the black-browed albatross Thalassarche melanophris (55% of birds captured), the white-chinned petrel Procellaria aequinoctialis, the spectacled petrel Procellaria conspicillata and the Atlantic yellow-nosed albatross Thalassarche chlororhynchos. Capture rates previously reported in the SW Atlantic varied from 0 to 5.03 birds per 1000 hooks, with those reports based on logbooks or fishermen interviews tending to underestimate capture rates, whereas those based on small numbers of hooks or short time periods tend to greatly overestimate rates in both pelagic and demersal longline fisheries. Previous studies have played an important role in delimiting the seabird bycatch problem, forming a baseline for mitigation actions and serving as a guide for improvements in data collection. However, data collected by onboard seabird-dedicated observers are more reliable, provide a greater range of information relating to bycatch, and form a baseline for more robust analysis and addressing further questions. The current study highlights the stochastic nature of seabird fatalities in longline fisheries and the need for extensive sampling to obtain realistic estimates of capture rates covering different years, seasons, vessels, and the range of fishing gear and practices.