Dinámica y manejo de la captura incidental de peces en la pesquería del langostino patagónico (Pleoticus muelleri)
This work characterizes the composition and dynamics of the incidental catch in the patagonian shrimp (Pleoticus muelleri) fishery, with special emphasis on the bycatch of common hake (Merlucius hubbsi). The Observer Program coordinated by the Chubut Province (Argentina) and used to monitor the fishery is evaluated. The factors that explain the variability of hake bycatch rates are analyzed, different statistical estimators are evaluated and possible strategies for controlling bycatch are discussed. The study is centered on the double-otter-trawl freezers, responsible for 75 percent of the Argentine shrimp landings. This fleet operates in San Jorge Gulf, in waters under the jurisdiction of the provinces of Chubut and Santa Cruz, and in waters of the platform near the gulf, in the area that is currently closed to protect hake juveniles. The incidental catch was composed of 81 species of fish: two species of mixins, 25 species of cartilaginous fishes and 54 of bone fishes, belonging to 50 families. Of these 81 species, 21 taxa had not been previously reported for central Patagonia (SW Atlantic). Two different fish assemblages were identified based on the analysis of qualitative abundance and frequency of occurrence: one in the region of Santa Cruz and the other in the northem coastal zone of San Jorge Gulf. Life-history characteristics, species abundance and status of conservation according to IUCN categories were used to identify nine species for which the bycatch in the shrimp fishery may not be sustainable, three of which are threatened and three vulnerable; the impact on two species could be sustainable but monitoring would be needed, 32 species appear in low frequency so the impact would be insignificant, and the level of impact is unknown for 39 species. Hake was the dominant species in the bycatch, in terms of both abundance and frequency. Analysis conducted using generalized mixed-effects models suggest that hake catch per set (CPUE) reported by the observers is markedly affected by the fishing area, the observer, the vessel, and the interaction between area and two-month period, area-year and vessel-year. Although substantial differences were encountered between areas, with some zones characterized by generally higher bycatch rates than others, there was no single area that could be identified as especially problematic in a consistent manner over the years. The observer effect is attributed to the methodology used to estimate the catch, whose precision depends on the observer expertise. The vessel was one of the main factors that explained the variability in hake CPUE, an effect that was apparently associated with the captain rather than with structural features of the vessel. The performance of different estimators of hake bycatch was compared by applying them to real and simulated data. Bycatch estimates were very sensible to the estimator employed. Estimators based on CPUE sample mean and based on mixed-effects models had better properties than those based on the hake-to-shrimp catch ratio. Finally, different strategies for the control of bycatch in the shrimp fishery are discussed in the light of the results, including selectivity devices, closures in space and time and bycatch quotas.