Surface longline targeting mainly swordfish is considered a major threat for loggerhead turtle populations in the western Mediterranean Sea. For this reason, it is necessary to analyse the relationship between the characteristics of longline sets and loggerhead turtle by-catch. The objective of this study is to look for differential spatial distribution of loggerhead turtle and swordfish captures along the longline to gain insights that might be useful to diminish unintended, incidental catch of turtles without affecting the captures of the target species.

Longlines are divided into transects separated by two consecutives buoys. During the 1999 fishing period of the traditional (without roller) Spanish surface drifting longline fleet (August to November), 59 technically homogeneous fishing operations, composed of 23 sections each, were observed. Chi-square test was used to compare loggerhead turtle and swordfish distributions among transects with those expected at random. Significant differences were obtained only for loggerhead turtles, with 93% of the specimens caught in the second half of the longline transects, which were retrieved from the sea during daytime. Significant differences were not found for the swordfish along the whole longline or for loggerhead turtle captures retrieved during daytime. We conclude that swordfish captures are independent of retrieval time whereas the incidental catch of loggerhead turtles occurs mainly during daytime, probably because loggerhead turtles use vision to locate baits.