ooking locations in sea turtles incidentally captured by artisanal ongline fisheries in the Eastern Pacific Ocean

Bycatch by longline fisheries, especially by artisanal small-scale fisheries, is one of the main conservation problems for some sea turtle populations around the world. Since 2004, a network of professionals under the “Eastern Pacific Regional Sea Turtle Bycatch Program” have been working with artisanal longline fishers in the Eastern Pacific Ocean (EPO) to reduce sea turtle bycatch and related mortality. Trials assessing circle hooks of different sizes and shapes, and different baits, have been conducted to determine the effectiveness in the reduction of sea turtle bycatch and changes in hooking location. In this paper, information from 1823 olive ridley sea turtles incidentally captured in the EPO were analyzed to assess how hook type (J, tuna hooks or circle hooks), hook size, bait type (squid or fish), turtle size and target species (tunas, sharks or mahi-mahi) affect hooking location on sea turtles. This were modeled with a Classification and Regression Tree using hooking location as a multinomial variable response (for 6 categories of hooking locations); and also as a binomial response (swallowed vs. non-swallowed) using a Generalized Linear Mixed Model (GLMM). Hook type and size, plus bait type, were the most important factors affecting hooking location, while turtle size and target species did not have any significant effect. J-hooks and tuna hooks had a much greater probability of being swallowed than circle hooks. In addition, as the hook size increased, the likelihood of swallowing it decreased. The use of fish bait in combination with larger circle hooks tended to produce higher proportions of external hookings. An increase in external or lower mandible hookings is preferred since these locations are assumed to be less dangerous for the animal’s post-release survival, and because hooks and attached gear are easier to remove by well-trained