Salmon spawn fierce debate over protecting endangered species, thanks to a single gene
For 40 years, Leaf Hillman, a ceremonial leader of California’s Karuk Tribe, has danced on the banks of the Klamath River. Following the tradition of his ancestors, he implores the salmon that have long sustained his tribe to return from the sea.
Chinook, or king, salmon (Oncorhynchus tshawytscha) arrive in two waves, in spring and fall, to spawn in freshwater. But the Karuk hold the spring arrivals in “special esteem,” Hillman says. The fish leave saltwater in March, having packed enough fat onto their meter-long bodies to sustain them for months, until they mature and spawn far up the river. Fall Chinook spawn lower down in the watershed and mature in the ocean before heading upstream, so they don’t carry as much fat.
The spring runs were historically larger, but dams built on the Klamath between 1912 and 1964 denied these so-called springers access to hundreds of kilometers of spawning habitat in the uppermost tributaries. And other changes, such as water diversions for farming, have affected the spring Chinook more than the fall fish, because springers spend more time in the river. As a result, fewer and fewer spring salmon answer Hillman’s appeal these days.
In 2011, conservationists petitioned federal officials to protect the Klamath’s spring Chinook runs under the Endangered Species Act of 1973 (ESA). The listing was justified, they argued, under an ESA provision that allows the government to protect a “distinct population segment” of vertebrate as though it were a full-fledged species. But officials at the National Oceanic and Atmospheric Administration (NOAA), which oversees salmon and other anadromous fish that spend part of their lives in the ocean, denied the petition. One reason: Researchers had concluded that the Klamath’s spring-run Chinook are genetically similar to fall-run Chinook.
New research findings, however, are forcing scientists and federal officials to revisit that decision. In 2017, researchers announced that they’d identified a single gene that appears to control whether Chinook salmon, as well as steelhead, a closely related species of rainbow trout, migrate upriver before or after reaching sexual maturity. They concluded that the genetic change that produced spring-run Chinook occurred only once in the species’s history. And new data published on 29 April on bioRxiv show that in rivers where spring runs disappeared decades ago, less than 1% of the remaining fish carry a copy of the early migration version of the gene. The scarcity of that gene makes it very unlikely a spring run will reappear once lost.
The findings have prompted the Karuk Tribe to submit a new ESA petition to designate the Klamath’s spring Chinook as threatened or endangered. The research has also sparked a fierce debate among conservation biologists and even the researchers who made the discovery. At its heart is the question of whether a difference in a single gene should be enough to qualify a population for ESA protection.
“We think that the biological reality is that these fish meet the criteria—the legal criteria—for listing,” says Craig Tucker, a biologist in Orleans, California, who works for the Karuk Tribe. But Fred Allendorf, a conservation geneticist at the University of Montana in Missoula, believes “you shouldn’t be describing units of conservation, or any taxonomic group, on the basis of a single gene.” Instead, he favors taking a broader look at the species’s entire genome.
The debate could extend well beyond salmon. The Klamath petition, experts say, could be the leading edge of a new wave of ESA petitions based on detailed genomic analyses that can pinpoint DNA mutations that produced major changes in the biology of an organism.
“It’s not just this one decision,” predicts Michael Miller, a geneticist at the University of California, Davis, who led the latest salmon studies. It could take decades, he says, for scientists and regulators to figure out “how to deal with this.”
The Klamath petition is only the latest twist in a long quest to protect the six species of salmon that spawn in U.S. rivers along the Pacific coast. Salmon return to their birth river to spawn and die, so salmon in different rivers often vary in, for example, size, run timing, and genetics. The first ESA petition to protect a distinct, river-specific population came in 1985, and it was followed by a spate of similar requests for salmon in other rivers.
But NOAA faced a big problem in deciding how to respond, recalls Robin Waples, a conservation geneticist at NOAA’s Northwest Fisheries Science Center in Seattle, Washington. “Nobody knew what a distinct population segment of salmon was,” he says. The ESA didn’t specify how distinct a population needed to be to win protection, so Waples was tasked with drafting a policy.
In 1991, he published a definition that holds legal weight to this day: A salmon population is distinct enough to qualify for listing if its members rarely mate with fish from other populations. It also should represent an important piece of the species’s evolutionary legacy—a piece that, if it were lost, could not be found elsewhere. Nine U.S. populations of Chinook currently meet those criteria and are listed as threatened or endangered.
But in rivers where spring- and fall-run fish aren’t genetically distinct, petitions to list spring Chinook separately haven’t gotten far. A key sticking point is that genetic study after genetic study found the same pattern: Spring Chinook generally look like fall Chinook in the same river—and not like spring Chinook in other rivers. That pattern led scientists to conclude that up and down the west coast, each river had evolved its own spring run. And it supported an assumption that if the springers became extinct, they could re-evolve from fall-run fish.
Is run timing important enough for salmon that it would justify making a big exception?
Robin Waples, NOAA’s Northwest Fisheries Science Center
Still, some researchers were curious about which genes were responsible for determining when a salmon leaves the ocean. Miller decided to figure it out. A native of rural western Oregon, he grew up fishing for spring-run salmon. “It’s a big part of what my culture is,” he says. “I’ve always been tremendously fascinated by salmon.”
He and his colleagues scanned a huge swath of the genomes of spring and fall Chinook taken from rivers in Washington, Oregon, and California. “In 99.99% of the genes, it showed the same old pattern” of spring fish looking like fall fish, says Waples, who wasn’t involved in the research. But when Miller’s team looked more closely, they detected a handful of spots where the genomes of spring and fall fish differed. Still, the researchers didn’t know how many genes were involved because their method—which entailed chopping the genome into tiny pieces and looking at a subset—couldn’t pinpoint where the DNA changes were located on various chromosomes. “We were working blind,” Miller says.
A breakthrough came when they got their hands on a genome of the rainbow trout, a Chinook relative. To their surprise, when they matched the intriguing snippets of Chinook DNA to the full rainbow trout genome, they saw that all the DNA changes associated with run timing lined up in the same region: in and around a gene called GREB1L. The gene doesn’t have a known function in salmon, but may interact with sex hormone to influence the expression of other genes. And its apparent importance to run timing was surprising, because researchers had assumed that migration would be governed by many genes, not just one. “Once we started to get things pieced together, I realized, ‘Wow, this is a single [spot in the genome],’” Miller says. “This is the premature migration gene!”
That conclusion—published in 2017 in Science Advances—was supported by another finding: Miller’s team and a second group found that GREB1L is also linked to run timing in steelhead, an anadromous form of rainbow trout that has summer and winter runs. After comparing DNA changes in GREB1L in different populations of Chinook and steelhead, the researchers also concluded that the genetic change that created seasonal runs occurred just once in Chinook and just once in steelhead, sometime in the 10 million to 15 million years since the two species split from a common ancestor. That challenged the conventional wisdom that the evolution of spring-run fish had occurred many times within each species.
The findings are rewriting the evolutionary history of early run salmon, scientists say. But they came as little surprise to the Karuk Tribe, Hillman says. “This is what we’ve always known, that the spring Chinook is not the same animal as the fall Chinook.”
Now, the question is whether NOAA will agree. Officials have started to sort through public comments on the Karuk Tribe’s petition, but it could be a year or more before the agency announces its decision. If NOAA decides to list spring Chinook, the ruling could trigger changes to water regulations on the Klamath, such as curtailing diversions for irrigation.
The issue has led to some feisty disagreements among researchers—including those directly involved in the Chinook studies. Five scientists collaborating with Miller, for instance, dropped off the 2017 Science Advances paper after a preprint was posted online, but before the paper was published. “One group really wanted to push ahead with more of a conservation emphasis for the paper; the other group thought the results were exciting in and of themselves without pushing that kind of angle,” says Andrew Kinziger, a conservation geneticist at Humboldt State University in Arcata, California, and one of the scientists whose name was removed. He believes the data are too “hot off the presses,” and that more work is needed before they should be used to influence listing decisions.
Allendorf worries that if NOAA approves the petition, it could open the door to splitting species into ever-finer groups, each with a genetic claim to legal protection. It’s a slippery slope, he says, and could become bureaucratically impractical. “What happens if we find another gene that’s really important?” he asks. “Does that mean we have to split [a population] even further?”
Waples, a veteran of endangered species debates, agrees that single-gene differences shouldn’t become a routine justification for protecting a population. But he thinks it might make sense if the gene codes for a trait that’s vital to the ecology and evolution of a species. “Is run timing important enough for salmon that it would justify making a big exception? … I’d like to see more debate on that.”
Nina Overgaard Therkildsen, a fish geneticist at Cornell University, believes that “early migrators should receive special protection.” They’re a unique part of the evolutionary legacy of the species, she says—a part we’d ignore if we overlooked what that single gene tells us about the evolutionary history of Chinook.
Hillman hopes such thinking carries the day at NOAA. An ESA listing could provide the Klamath’s spring Chinook with “a little bit of interim relief” before engineers begin to dismantle the Klamath’s four dams in 2020. That project—the largest dam removal project ever attempted—will reopen vast stretches of spawning grounds.
In the meantime, Hillman will continue to dance on the banks of the Klamath—hoping that his generation hasn’t seen the last of spring-run salmon.
Spring Chinook restoration might be easier said than done
When California’s Iron Gate Dam—the last of four dams constructed on the Upper Klamath River—was completed in 1964, it was a “nail in the coffin for the spring salmon run,” says Craig Tucker, a biologist in Orleans, California, who works for the Karuk Tribe. Many of the Klamath’s spring-run Chinook salmon migrated to the river’s uppermost tributaries to spawn, so they were more affected by the dams than fall-run Chinook were, which tend to spawn in lower stretches of the river. Similar dams have become “a problem for springers all up and down the [Pacific] Northwest,” Tucker says—leading to controversial efforts to protect some spring salmon runs under the federal Endangered Species Act of 1973 (ESA).
Now, with the Iron Gate Dam and three other Klamath dams slated to be removed, starting in 2020, salmon biologists face some pressing questions: Will the fish return on their own? And, if they don’t, should they try to restart the runs by replanting fish from elsewhere?
Of particular concern are the Klamath’s spring-run Chinook salmon. Biologists expect fall Chinook to recolonize the Upper Klamath, which spans northern California and southern Oregon, because some of the fish already spawn near the base of the lowermost dam. Those fish will push farther upstream once the obstacles are removed, they predict. But they are skeptical that the springers will return on their own. In large part, that’s because the remaining spring Chinook tend to spawn in Klamath tributaries more than 100 kilometers downstream from the dams.
To get around that problem, managers are thinking about an intervention: actively moving spring-run stock to the upstream reaches, says Kevin Goodson, a conservation planning coordinator with Oregon Department of Fish and Wildlife in Corvallis. Plans are still in the works, but they’d likely move salmon in an early life stage to the upstream reaches. There, the youngsters would undergo a process called imprinting, which encodes the river’s chemical signature in a salmon’s brain before it heads downstream and out to sea for 1 to 6 years. The imprinting allows the fish to return to spawn in the same waters where it grew up. Such encoding has also led, over millions of years of evolution, to salmon that are uniquely adapted to their home rivers.
Such nuances mean that although the transplantation strategy may sound simple, it’s not clear which fish biologists should use. The Klamath’s remaining spring runs, for instance, are likely adapted to spawn in rivers with different water flows and temperatures from those in the Upper Klamath. And the remaining fish are already at risk of extinction, so it’s hard for some conservationists to stomach the idea of using stock from those struggling populations. Another option is to use stock from a nearby river, such as Oregon’s Rogue River to the north. It has a more sizable spring salmon population, and somewhat similar hydrological characteristics.
But there’s a hitch: Chinook in the Rogue are genetically distinct from Chinook in the Klamath, to the point where, technically, they are considered separate “species” under the ESA. That could complicate management, especially if Chinook in either watershed are given federal protection. Implementing a transplantation scheme would likely involve years of study and acquiring numerous permits—and the risk that the fish might not adapt to their new home.
“The challenges that we’re facing now with the restoration of spring Chinook in the Klamath are really just a mild preview of what we expect in other locations,” says Michael Miller, a geneticist at the University of California, Davis. “It’s just going to continue to get worse as these [early run] populations continue to blink out.”