> Cooperative Research Projects

Industry/NMFS Partnership Boosts Fishery
SEAFISHER Applies Software, Video and Ingenuity to Better Manage Bycatch
by David Benton

F/T SEAFISHER was selected by MCAF for the research due to its experience with research cruises and crew's reputation for putting the goals of the research first. Crew members and sea samplers (senior NMFS ovservers hired for the research) evaluate the performance of a mechanical drawing device that works the vessel's flow scale to draw stratified random samples. If performance is satisfactory, the device could be used in the regular fishery to free up an observer's time to do more biological sampling such as collecting fish lengths or tissue samples needed for specific scientific anagmeent projects.

(Oct 2006) IN THE MIDDLE OF THE BERING SEA - October can be an unforgiving month for commercial fishermen in the Bering Sea with gathering winter gales that are bitingly cold and whip seas up to 40 feet or more. But it was into these challenging waters that the 211-foot fishing vessel SEAFISHER ventured this past October with its 30-ton trawl nets, fishing for science as much as seafood. "The key to improving fisheries management today is better information and better science," said John Gauvin, a private resource economist who works on Alaska fishing issues. "We want to fish more cleanly and reduce fishing's impact on the ocean ecosystem. That means we need a better way to account for what's being caught which is why we set up this cooperative research program, even in a stormy October." Owned by Cascade Fishing of Seattle, the SEAFISHER mainly fishes for groundfish like yellowfin sole, a small but abundant flatfish with a delicate taste that makes fillet of sole popular on dinner plates in the U.S. and Asia. Its catch in the Bering Sea is third in volume only to pollock and cod, and worth over $30 million annually, but yellowfin share the ocean floor with other flatfish and fishing for it produces a mixed bag of fish.

Called bycatch, this incidental catch includes valuable species like halibut that are closely regulated and reserved for other fishermen. Halibut have to be handled to minimize injury and returned to the sea as soon as possible. Other species like skates may be less desirable but still occupy a place in the ecosystem and need to be protected. The task of sorting out exactly how much of dozens of species are caught is not easy.

"In the past, we've used sampling techniques that estimated the catch of the entire fleet on a week by week basis but as we move toward catch and bycatch accountability for each vessel, there is the demand for much greater precision," said Bill Karp, director of the Alaska Fishery Science Center's Monitoring and Analysis Division. "Without better data we can't move toward systems that promote individual accountability and better utilization of resources that are supported by both fishery managers and the industry."

John Gauvin understood the problem well. "A few years ago, we tried to sample all the fish on a boat using only baskets and platform scales. After a few weeks, one of our samplers said, 'I'd rather be building the Pyramids.' We knew if we're really going to account for all these species, we needed the cooperation of both industry and fishery scientists who could think outside of the box."

Over a year ago, Gauvin started talking with scientists with the National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center. The project attracted the attention of the International Pacific Halibut Commission which has an interest in reducing halibut bycatch. Gauvin enlisted the help of the Marine Conservation Alliance Foundation, a Juneau-based industry association. They referred him to Cascade Fishing, a company known for its innovative and previous scientific work.

Using a research permit and funding from the National Cooperative Research program, Cascade's SEAFISHER ventured out to the Bering Sea in early October with its usual crew of 47 plus two NMFS samplers and three scientists. Fishing east of the Pribilof Islands in stormy weather that occasionally forced them to pull in their nets and ride the weather out, their first challenge was simply getting the fish to where they could be counted, a problem that was mostly mechanical. "This is where industry's ingenuity really shines because they know how to move fish," said Gauvin. "Usually fish come into the holding tank and move forward to where they are headed, gutted and frozen. Cascade came up with a design for conveyor belts that move fish backwards to where they can be sorted and weighed. It was very ingenious. Nobody's ever tried to move fish backwards in a factory before."

A conveyor belt used to transport catch from the vessel's live tanks to the flow scale. Part of the reserach was to evaluate how well video cameras could be used at locations like this to ensure that none of the catch fell off the belt or was removed prior to the flowscale and observer sampling station.

With its backwards conveyors in place within the confines of the SEAFISHER's hull, the next challenge was to make sure they could get a truly random sample of the catch. This proved to be mainly a software issue.

"Having an electronic flow scale was absolutely fundamental," said Karp. "The scale is built into the conveyor and fish are automatically weighed as they pass along. NMFS and Cascade worked with the software manufacturer to modify its system to select random samples from each catch that could be diverted to the observer."

On a keypad, the scientist only needed to punch in the estimated weight of the tow and the number of samples needed and computer automatically diverted the right amount of fish. But did these samples give an accurate estimate of the actual catch of target and non-target catches?

Crew members and sea samplers sort the catch by species during the research cruise. One goal of the work was to evaluate the increase in accuracy from drawing samples mechanically. This required weighing the principle components of the catch by species and comparing them to the estimated species composition from basket samples.

"To answer that, we needed to count and weigh everything, a whole-haul sample," said Todd Loomis, NMFS' chief scientist on the project. "We sorted through every single fish as they ran by on the conveyor." Sorting through 30 tons of fish three or four times a day couldn't be done under normal conditions but was the only way to confirm they were getting a statistically valid sample of the catch. Lastly, there's the "trust but verify" question: how do you make sure nobody is presorting the catch before it gets to the sampler? Video technology, now used to monitor everything from traffic jams to volcanoes, offered a possible solution. Video cameras were focused on the SEAFISHER's fish deck, mounted in the fish bins and along the conveyors, and placed in the processing room. In all, nine cameras followed the fish from when they were hauled on board until they went into the freezer.

"A large part of an observer's work is to monitor compliance with the regulations and that's a difficult position to be in aboard a fishing vessel," Karp said. "Designing a system like this makes it whole lot easier to observe what's going on and since everybody knows the monitoring is taking place, it lends itself to more cooperation and consistency."

After sorting through some 37 thirty-ton tows, the results of the two-week SEAFISHER experiment seem to be success.

"The automatic sampling mechanism worked," said John Gauvin. "They had a couple of glitches at first but once they were ironed out, it delivered a sample of the catch right to the observer. Eliminating their need to lug baskets of fish around means more of their time can be spent on getting better scientific data. That's their real job and the fact that the system worked so reliably shows a lot of promise."

The exhaustive whole haul count showed the sample was indeed representative. "Using this technology, can we eliminate sampling subjectivity and minimize bias," said Bill Karp. "This kind of research is required to understand the relationship between sampling and actual catch composition, address questions of uncertainty, and ultimately implement the management strategies that research suggests."

"With the video, we looked to see whether we could monitor all the points of the process for presorting or mechanical sorting prior to sampling," said Todd Loomis. "We tagged halibut with plastic zip ties to see if we could track these fish and our initial analysis is, yes, video is a tool that you can use to track the flow of a prohibited species."

"It removes doubt on both sides," added Cascade Fishing's Phil Dang who served as the company's scientific liaison on the trip. "If there's ever an issue we can always go back and check the video."

With increasing demand for better bycatch accountability on individual vessels, the SEAFISHER test was a win-win for fishery managers and industry alike.

"Our livelihood depends on accuracy," said Dang. "I really believe this new system is going to help us. It's easier for both us and the scientists to work with. With better accuracy in sampling, NMFS will be better managing the fisheries for us."

"And what's really important in my mind is the proof of concept," added Bill Karp. "If we can bring NMFS, the Halibut Commission and the fishing industry to work collectively on solving problems like this, you can expect to see more of these projects in the future."

David Benton is Executive Director of the Marine Conservation Alliance that represents harvesters, processors, and communities involved in Alaska's groundfish and shellfish fisheries. Previously, Benton served as Deputy Commissioner of the Alaska Department of Fish and Game and chair of the North Pacific Fishery Management Council.