Keeping shark populations healthy is an important part of ecosystem-based management for sustainable fisheries and resilient fishing communities. But getting the information needed to effectively manage Pacific sleeper sharks, salmon sharks, and other large Alaska sharks is challenging.
“Data on these sharks are hard to get. They’re so big! It’s not always possible to safely land or transport them back to shore for sampling,” said Cindy Tribuzio, NOAA Fisheries Alaska Fisheries Science Center. “There is so much unknown. Every bit of new knowledge we can obtain makes a big difference to how we manage them.”
Alaskan fishermen are stepping up to help collect the data needed to better understand large sharks in Alaska waters. A collaborative project led by Tribuzio is innovating the use of electronic monitoring (EM) technology and machine learning to collect information from sharks incidentally caught by longline and trawl fishing vessels. As part of that research, fishermen are voluntarily taking biological measurements and samples.
“The information they are collecting will contribute to more accurate assessments and more effective management of shark populations,” Tribuzio said. “Ultimately, that means more stable, better-managed fisheries overall.”
“When Dr. Tribuzio asked that pollock fishery EM participants measure all sharks before discarding them, our vessel operators and crews agreed to try. Measuring the sharks, converting the lengths to weights, and recording the data in the vessel logbook was extra time and effort for all involved—but a big improvement to data quality. There was concern by some that it could present a safety hazard handling and measuring each shark versus just discarding it. Not only did they do this, but they also made the effort to actively collect additional biological data that couldn't be collected otherwise,” said Julie Bonney, Executive Director, Alaska Groundfish Data Bank. “Fishermen are more committed to maintaining data quality and contributing to research that maintains sustainable fisheries than many people think. They rely on the ocean for their livelihoods and they are invested in maintaining it.”
The Power of Collaboration
The cooperative relationship among fishermen and scientists is dramatically increasing our knowledge of Alaska sharks. It’s also expanding EM capabilities and helping develop cost-effective technology to rapidly and efficiently analyze data.
Expanded Coverage over Time and Space
“We are now collecting estimates from the largest U.S. fishery—the Alaska pollock fleet,” said trawl EM project director Josh Keaton, NOAA Fisheries Alaska Regional Office. “Thanks to the willingness of industry to collect data not usually available to us, shark sampling now covers vast areas of Alaska waters nearly year round.”
“This is a perfect example of mutually beneficial cooperation where data gaps are being filled while fishermen are able to continue their operations,” said Ruth Christiansen, Science Advisor/Policy Analyst, United Catcher Boats.
Better Size Estimates for More Accurate Assessments
The data EM fishermen are collecting is both reducing a sampling bias toward small sharks and improving the accuracy of size estimates.
“We strongly believe in the power of collaborative research and leveraging fishing vessels platforms when appropriate. The shark research is a unique opportunity, where fishermen are often seeing larger, possibly older, more mature sharks than what can be delivered to plants for sampling,” said Charlotte S. Levy, Natural Resources Assistant Director, Aleutians East Borough. “This is a great example of how fishermen can be partners in research and provide data that otherwise may not be collected or would be difficult to collect.”
“The measurements fishermen are providing have drastically improved precision and accuracy of shark size estimates over previous visual estimates,” Keaton said. “We can see them taking measurements and converting them to weight correctly on EM video.”
New Information on Pacific Sleeper Shark Growth, Longevity, and Vulnerability
Pacific sleeper sharks are the most data deficient shark species in Alaska. They are also the most vulnerable to climate change and overfishing. This is likely because they have long lifespans—generation time could be 40 or 80 years. Our time series of catch rates are not long enough to see growth and population trends. To do that, we need to know how old they are.
“We never had age information for these sharks,” said Tribuzio. “Because we’re working so closely with industry, I’ve been able to ask them to collect photos of mature sharks and eyeballs that we hope to use to determine age. It’s totally voluntary, and they are providing samples I never had before.”
“No one wants to catch these animals, and when they do they try to release them alive,” Keaton said. “But when they can’t be avoided, they can provide valuable information. Fishermen are collecting data that has never before been available to science.”
Machine Learning Enables More efficient, Cost-effective Analysis
Tribuzio is working with partners at the Alaska Pacific University and the University of Washington to develop machine learning to review EM videos. This project is funded by the NOAA Catch Shares program, and was originally intended to focus on the EM video from longline vessels. However, the opportunity to collaborate with the trawl EM vessels has greatly expanded the data available for the project.
The eventual goal is to use machine learning to take measurements instead of asking fishermen to do it. To help achieve that, the data and photos fishermen provide are helping to “teach” machine learning to recognize shark species and categorize them by size. This will be a giant advancement for machine learning, which is already saving an enormous amount of time and labor.
Enhanced Electronic Monitoring Capabilities
Thanks to the participation of numerous partners and fishery vessels, we are increasing our knowledge of these species while simultaneously expanding the field of EM fisheries applications. As NOAA Fisheries transitions into EM, collaboration with industry will help mitigate data loss from boats that no longer have observers.
“Our project demonstrates that cooperative research works really well, and can take many forms,” said Tribuzio. “We need to be open to out-of-the-box ideas.”
“In a world of dwindling resources but increasing need for data and information, the fishing industry and federal fishery managers should strive to work cooperatively when and where they can,” Christiansen said. “Industry and managers are partners in the sustainability and optimized use of our fishery resources. Working together simply makes sense to ensure the continuity of both.”
This research is a collaborative effort among NOAA Fisheries Alaska Fisheries Science Center, Alaska Regional Office, Alaska Pacific University, Pacific States Marine Fisheries Commission, the University of Washington, and many fishing industry participants.
