The effect of American and Spanish longline systems on target and non-target species in the eastern South Pacific swordfish fishery

摘要

Catch compositions obtained using two different types of commercial surface longlines for swordfish (Xiphias gladius) in the eastern South Pacific were compared. The American (''Florida style'') system uses polyamide monofilament nylon for the mainline and gangions, whereas the Spanish system consists of polyethylene multifilament with wire in the distal part of the gangion. A total of 37 fishing sets were carried out using both longline types. The amount and diversity of species caught, mainly swordfishes, billfishes, and tunas (Thunnus spp.), was highest with the monofilament longline; however, shark catches were higher with the multifilament longline. The type of longline significantly affected total catch, swordfish catches, and shark catches; total catch and swordfish catch rates were higher with the monofilament longline and shark catch rates higher with the multifilament longline. Moreover, a positive correlation was found between total and swordfish catch rates and the soak time with the monofilament longline, and between shark catch rates and the soak time with the multifilament longline. Significant differences were found in the average length of blue sharks according to the type of fishing gear used: larger specimens were caught on the multifilament longlines. A selectivity effect was also observed, with the multifilament system retaining more specimens exceeding 200cm in fork length. Generalized additive models were developed to explain swordfish, shark, and teleost catches as a function of seven operational (longline system, soak time, depth, number of light-sticks), spatial (zone), and environmental (sea surface temperature, wind velocity) variables. The final fits of the models explained 68% of the swordfish catches, 38% of the shark catches, and 81% of the teleost catches. The results showed that two operational variables - longline system, soak time - were best able to explain the catches in all the models. The effect of the environmental variables significantly explained swordfish and teleost catches, whereas only wind velocity significantly added to the sharks model. These results show that the use of monofilament leaders in the surface longlines might be an effective way to reduce shark by-catches in the eastern South Pacific swordfish fishery.