A study using satellite data from tagged leatherback turtles has identified possible "by-catch hotspots" in the Pacific Ocean.
By tracking 135 turtles, researchers highlighted areas where the critically endangered animals were likely to come into contact with fishing vessels.
The authors hoped the findings could be used to help cut the number of turtles accidentally killed as by-catch.
The paper appears in the Proceedings of the Royal Society B journal.
The species, scientifically known as Dermochelys coriacea, is globally listed as vulnerable by the International Union for Conservation of Nature's Red List of Threatened Species.
However, conservationists have identified a number of distinct populations in the world's various oceans. Within the Pacific, the two populations - eastern and western - are considered to be critically endangered.
For this study, data was collated from numerous projects between 1992 and 2008, explained lead author John Roe from the University of North Carolina's Pembroke campus.
"To really get an idea about where the leatherbacks go, you have got to have turtles tagged from multiple locations," he said.
"So it took getting just about everyone who has put a satellite tag on a leatherback for other research purposes to collaborate to get a sample size large enough to allow us to answer that question."
Dr Roe and his colleagues from a range of US and Costa Rican institutes then integrated the turtle location data with similar data of where the longline fishing activity in the Pacific Ocean was highest.
"We used that data to overlay with the data of the areas the turtles were using in order to figure out where the turtle hotspots matched with the fisheries hotspots to identify the areas where by-catch was most likely to occur," he told BBC News.
Writing in their paper, the team observed: "For East Pacific nesters, an area of potential risk occurs along the primary leatherback migration corridor between Costa Rica and the Galapagos Islands.
"For the West Pacific population, ...the greatest by-catch was predicted to occur adjacent to nesting beaches in north-west New Guinea."
They added that the analysis represented the largest compilation of data of its kind.
However, Dr Roe said that more tagging data was required in order to establish a more detailed picture.
"We need to target these areas to see if the turtle persistently use these areas over and over again," he said.
"That would provide really useful information in the management of by-catch because the fisheries authorities would have that knowledge and adjust their fishing efforts accordingly."
He added: "We are just trying to highlight some areas to make it easier to look for these needles in a haystack.
"The idea now is to try and get some more refined information in those targeted areas but we now know where and when to look, and where to concentrate our efforts."