How Seabirds Can Help Us Predict the Size of Fish Stocks

The scientists who study terns, puffins and other birds are trying to get fisheries managers to heed their warnings

Puffin Holding Fish
Seabirds, such as puffins, have a well-earned reputation as sentinels of change in marine environments. Siddharth Sachar / Alamy Stock Photo

In 2015, fisheries managers monitoring Atlantic herring in the waters off the United States and Canada made a grave miscalculation. Things were looking good for the multimillion-dollar fishery: Adult herring in the Gulf of Maine and the Bay of Fundy were more plentiful than they had been in a decade, and mathematical models estimated that one-year-old herring were also abundant—a promising sign for future catches. The managers set high fishing quotas for the next three years. Then, the herring population went into free fall.

One person who wasn’t surprised by the downturn was Lauren Scopel, then a graduate student at Canada's University of New Brunswick studying terns on Machias Seal Island, on the U.S.-Canada border.

Every summer, on islands throughout the Gulf of Maine, ornithologists hunker inside weathered plywood structures and watch as terns and puffins fly in from the ocean with beakfuls of fish, delivering thousands of meals to hungry chicks. Researchers have been tracking seabird diets on the islands for decades. “We were very aware that there were not a lot of herring” between 2012 and 2014, Scopel says.

This should have been a red flag to herring managers. Terns and puffins are expert samplers of juvenile herring, the little 1-year-old fish that will grow big enough to be caught by fishers in a couple of years. Fisheries managers, however, have no dependable way to count juvenile herring and instead rely on the uncertain estimates generated by mathematical models.

Seabirds are highly sensitive to changing ocean conditions and prey availability, earning them a reputation as sentinels of the sea. But ornithologists worldwide struggle to get fisheries managers to pay attention. Enriqueta Velarde, a seabird ecologist at the University of Veracruz in Mexico, has been trying since the 1990s to convince fisheries biologists in the Gulf of California to heed warning signals from seabirds, which predict local collapses of sardines and anchovies. “I was telling them what they didn’t want to listen to,” Velarde says. “So now I’m not allowed to go to meetings.”

Back in 2015 in the Gulf of Maine, seabird biologists were also struggling to get fisheries managers to consider their data. “How can we be so far apart in terms of what we’re seeing?” Scopel remembers wondering during a fisheries management meeting that year. “How can I watch all these chicks slowly starve to death [while] we’re talking about how there’s more herring than there’s ever been?”

In the wake of the herring crash, instead of saying, “I told you so,” Scopel published a paper that she hoped would convince fisheries managers to adopt seabird data. Her 2018 study showed that the amount of herring in the diets of tern chicks can help predict the future size of the herring stock in the Gulf of Maine and the Bay of Fundy.

But it took the Covid-19 pandemic to give Scopel’s research its big break. In 2020, with fish surveys interrupted, “we went from having very little data on those [juvenile] fish to having none for a couple of years,” says Jon Deroba, a fisheries biologist with the US National Oceanic and Atmospheric Administration (NOAA), who is responsible for assessing New England’s herring stock. NOAA’s herring model “broke down on me because of missing data,” he says. “So I started scrambling, looking for something to fill that data gap.”

Seizing on Scopel’s study, Deroba compiled information on the amount of 1-year-old herring in the diets of seabirds on 12 different islands going back to 1988. Then he started testing the data in the model that predicts the size of the Gulf of Maine herring stock. If the work passes a gauntlet of peer reviewers and makes it into the final version of the model, seabird data will, for the first time, begin directly influencing herring quotas.

“That’s the state of the art,” says Bill Sydeman, a marine scientist and president of the Farallon Institute, a California-based nonprofit focused on ocean research.

Fisheries biologists on the Pacific coast of North America have a history of stronger collaboration with seabird scientists than their counterparts in the Atlantic and Gulf of California regions. Yet, even on the West Coast, the use of seabird data is generally limited to ecosystem assessments. Getting seabird diet data into a model used to determine fishing quotas “would be a huge step,” Sydeman says.

One reason fisheries managers are so reluctant to take advantage of seabird data is inertia, Deroba says. Usually, the models focus on a single target fish species and don’t consider information from other species or the environment.

“It’s hard to go do something new until something breaks [and] really makes it necessary,” says Deroba.

And some fisheries management institutions are just resistant to collaborating with seabird researchers. Tony Diamond, a now-retired bird ecologist with the University of New Brunswick, began studying seabirds on Machias Seal Island in 1995 and was a co-author of Scopel’s study. “[Scopel] tried particularly hard—and so did I—to get some interest from [Fisheries and Oceans Canada], and we failed,” Diamond says.

During a recent overhaul of the model for Bay of Fundy herring, a Fisheries and Oceans Canada (DFO) team—led by research scientist Timothy Barrett—rejected seabird data as a source. The group cited concerns that the seabird monitoring programs fall outside the control of DFO and its industry partners.

But, according to former DFO herring biologist Michael Power, the seabird diet data would be a boon for Canada’s fishery. The herring population has been below the critical level for years, yet management often prioritizes fishing industry demands over science, he explains. With herring at such low numbers, Power says, existing models can’t accurately estimate how many juveniles there are—making it all but impossible to predict the future adult population.

The seabird monitoring data provides an obvious opportunity to see what’s coming for herring, Power says, and DFO scientists should be using it right now. “They should grab the information right out of the paper and plug it into their models and see how it works.”

This article is from Hakai Magazine, an online publication about science and society in coastal ecosystems. Read more stories like this at

Get the latest Science stories in your inbox.