Article body copy
This story is part of Covering Climate Now, a global journalism collaboration strengthening coverage of the climate story.
Brenda Bernhardt was rowing in a lake near her home on Vancouver Island in British Columbia when she paused at the mouth of the Robertson River, noticing that it had run dry. It was the first day of May 2019, and the spring had been unseasonably warm. Bernhardt, a retired veterinarian, had her two dogs, Ellie and Cody, with her. They were on their way to check out “an eagle’s nest that we always look at. Or I always look at. I don’t know what the dogs are looking at,” she says, her laughter deep and boisterous, even over the phone. “And I thought, ‘Oh, gosh, look, the river is really low.’” Bernhardt climbed out of her rowboat to investigate.
She wandered up the wide, dry gravel riverbed lined with second-growth evergreen trees. Here and there she discovered dozens of disconnected puddles, each teeming with coho salmon only a few weeks old and no bigger than the palm of her hand. Juvenile coho spend a year in the Robertson River putting on weight and muscle before migrating to the Pacific Ocean: they swim north from the river into a little lake called Bear, hang a quick left into the much larger Cowichan Lake, then make a righthand turn into the Cowichan River and follow its flow southeast for over 50 kilometers to Cowichan Bay and beyond.
Two years later, adult coho migrate back to their birthplace, following the scent of the Robertson River through the ocean, bay, and lakes, to spawn and die in the gravel, their bodies feeding their progeny, as well as the river and the forest. But that journey never happens if the puddles dry completely, leaving the fry flopping on land, or if the stress of a shrinking, warming puddle kills them first.
“It’s just a terrible waste,” says Bernhardt, her voice heavy with sadness.
She rowed back home, grabbed a fishing net, and returned to the river. The rest of that day and the rest of the summer, she moved coho into the lake using a net and bucket. She found it impossible to relax at home knowing that, a few kilometers away, an animal might be dying. Her response was instinctive—“my motivation is simply the need: a being needs help, help it”—but she was practicing an obscure, growing branch of conservation called fish salvage: transferring a fish from unsafe to safe habitat. A nearby hatchery, run by Fisheries and Oceans Canada (DFO), operates a similar program and tacitly endorses impromptu salvaging done by people like Bernhardt.
The Robertson River was not the only waterway that went dry on Vancouver Island that summer. The Cowichan River also ran low. While coho flourish in the meandering back eddies of the Robertson, the rushing Cowichan supports a once mighty, now declining chinook salmon stronghold. The local paper mill pumped water in from adjoining Cowichan Lake to keep the river flowing—a local politician called it “mechanical life support.” Using the 2016 Water Sustainability Act for the first time, the provincial government issued a Fish Population Protection Order that prohibited farmers from drawing water out of dangerously low rivers to irrigate forage crops, such as hay and corn.
The wet, rainy winters and snowy mountains of the Pacific Northwest tend to lull locals into a false sense of water security. Then the warmer months arrive, and that security evaporates. The summer of 2019 broke decades of drought records. Bernhardt reckons that until the rain started falling again in the autumn, she moved 26,000 fry down the river. That was a lot of buckets to carry.
Salmon have a tumultuous history in the Pacific Northwest. They’ve survived earthquakes, exploding volcanoes, and drastic swings in temperature, largely thanks to a complex life history. The five houses of Pacific salmon—coho, chinook, chum, pink, and sockeye—cover a lot of real estate over their lives. The details differ, because every population evolved to flourish in a hyper-specific waterway, like the Robertson River. But a general pattern holds true: born and reared in fresh water, salmon migrate to the ocean via rivers and estuaries, grow to adulthood in salt water, and return to spawn and die in their home creek. Salmon cross barriers that other animals can’t, transforming physiologically as they move between fresh and salt water and back again on a mind-blowingly epic migration. Yet these evolutionary feats are not serving them well in the modern age. Since European contact, a third of the 1,400 distinct salmon populations along the west coast have gone extinct. Dams and development hinder migrations. Overfishing in the oceans and rivers cuts into population numbers, too. Now this resilient silver fish is facing a new, possibly insurmountable challenge. A warming climate is drying and disconnecting waterways they’ve traveled for millennia. As far as we know, a salmon has no evolutionary trick for surviving a dried puddle.
This has led salvagers, like Bernhardt, to make a last-ditch effort to keep a run alive. The practice is mostly ad hoc and volunteer-run, with government agencies giving implicit permission by practicing it themselves. A few wildlife agencies on the west coast of the United States have run fish salvages since the 1950s, like the California Department of Fish and Wildlife’s network of concrete collection facilities and fleet of oxygenated trucks that move endangered fish around water pumps in the Sacramento-San Joaquin River delta. As the climate warms and threatens freshwater habitat, more governments will likely adopt the practice.
Normally, this might be a success story, except no one really knows if salvage succeeds. Is a salvaged fish truly saved? And, in response to a warming climate that is drying out Pacific Northwest rivers, a new and more extreme version of saving fish is taking root: fish rescue. In this iteration, wild fish are moved into holding tanks and released when habitat improves. Neither practice is well studied, the terminology is still slippery, and the science is murky on whether such efforts help or hurt. But that hasn’t held back salvagers and rescuers from wading through streams, nets and buckets in hand, saving salmon while they still can.
People help animals all the time in ways big and small: moving a worm off the hot pavement, shooing a deer off a dark street as a car comes barreling down the road. This altruistic human impulse has accelerated in the age of the Anthropocene. Climate change, pollution, and habitat loss are setting off a wave of mass extinctions, and human intervention is becoming bigger, bolder, and costlier than ever. Chytrid fungus has wiped out dozens of Central American amphibian populations, but a network of global captive breeding programs has preserved some frog species from outright extinction. In Australia, hundreds of slow-moving koalas went up in smoke during the bush fires of 2019—people dashed into burning thickets to pull survivors from eucalyptus trees. Preventive efforts, like wildlife crossings, have sprung up to shepherd animals around expanding development. Fish salvage and rescue do the same for aquatic environments, by bridging the gaps in drying rivers as the climate warms.
It’s hard to pinpoint where exactly fish salvage began. A few early 20th-century examples exist, like in 1914 after the Hell’s Gate landslide blocked British Columbia’s Fraser River, a major salmon migration route. Local Indigenous groups moved salmon with dip nets around the slide until fish ladders were installed decades later. In the United States, the 1966 Endangered Species Preservation Act—a precursor to the 1973 Endangered Species Act (ESA)—codified the protection of endangered animals into law. In 1982, the act was revised to include an experimental category for moving threatened animals. This encouraged piecemeal salvage efforts to expand in earnest. “Salvage is just ubiquitous,” says Jonathan Armstrong, an assistant professor in the Department of Fisheries and Wildlife at Oregon State University. Armstrong works in Corvallis, Oregon, on the Willamette River, where fish salvage is particularly popular among the public and where state governments move fish when a river is obstructed for construction or irrigation.
Salvages can look wildly different depending on who’s doing them and why. There are individuals, like Bernhardt working the puddles of Robertson River. There are also official salvages, like the intergovernmental effort in British Columbia during spring 2019 that deployed a helicopter and a 68-person crew to fly salmon around a landslide on the Fraser River.
Salvage is not only intended for short-term workarounds or emergencies. For over 60 years, the California Department of Fish and Wildlife has run a continuous daily salvage operation near San Francisco Bay, where the federally listed, endangered delta smelt gets trapped at a water diversion facility. The smelt are trucked in oxygenated tanks away from the facility and released back into the bay.
When it comes to saving wildlife, humans are the heroes and the villains: we are saving animals from a mess of our own making. Anthropogenic disruptions to the waterways have caused over 80 percent of studied fish strandings. Another study found that humans caused 66.9 percent of surveyed mass fish die-offs, like in Australia’s Murray-Darling Basin, where a record-setting three-year drought started in 2017 and then collided with years of over-extracting water for agriculture, leading to a toxic algal bloom that killed a million fish at once.
The uncomfortable reality is that the salvager is shielded from the fish’s final end after releasing the animal. The salvaged fish might swim off only to encounter the same danger right around the riverbend. This is the problem with sending a struggling animal back into a compromised habitat. Over the last two decades, Armstrong noticed a new tactic emerging: fish rescue. By holding onto an animal until habitat improves, rescuers hope to better the animal’s chances in the wild.
In the fall of 2016, Armstrong’s lab joined forces with the Washington Department of Fish and Wildlife to assess a long-running fish rescue initiative on the East Fork Lewis River just north of Vancouver, Washington. A nonprofit organization, Northwest Wild Fish Rescue, captures threatened lower Columbia River wild coho and steelhead, raises them, and then after summer droughts have passed, releases them. Armstrong wondered whether this new way of saving fish might go too far. When humans insert themselves into a wild animal’s life cycle, the results are unpredictable. A little bit of help might lower fish fitness by impairing the animal’s ability to compete in the wild. If that compromised fish survives to spawn and then its progeny is also rescued and released, the genes could continue to pass down the line. The Washington Department of Fish and Wildlife wanted to know whether adding fish rescue to its arsenal for saving threatened coho was a winning strategy. Armstrong and graduate student Brittany Beebe went to find out.
Over 15 years earlier, Dave Brown stumbled on a nearly identical scene as Bernhardt did in the Robertson River: a hot summer, a drying creek, young salmon on the verge of dying in a disconnected pool. The hobby fisherman was out for a walk on his property near the East Fork Lewis River when he came upon the scene, and he acted with the same reflexive gusto as Bernhardt, moving the coho to safer water. The following year, he built a spring-fed pond in his backyard to raise the coho until they came of age.
Brown founded and now runs Northwest Wild Fish Rescue. Every year, he and his volunteers rescue and release between 15,000 and 30,000 wild coho—all with the blessing of the Washington Department of Fish and Wildlife. The hope is to revive an ESA-listed population of wild coho in the Columbia River. Before European contact and ensuing development, up to 16 million salmon returned to the Columbia River Basin every year. Today, many wild runs in the region have as little as two percent of their historical abundance.
Sweeping up the west coast of North America, a pattern emerges. The Pacific salmon in the hotter, drier, more developed south are swimming in seriously threatened straits compared with their counterparts in the cooler, wetter, and less populated north. Coho are especially at risk, and turn up consistently in salvage and rescue programs, thanks to a quirk in their life history: they spend their first year in rivers high up in the watershed that dry like clockwork every summer. Fish rescue attempts to circumvent these systemic issues by adding healthy fish to the river during ideal environmental conditions.
Beebe ran a modeling study to look at the outcome of rescuing and holding fish across various seasonal conditions and life stages. Her results revealed that Northwest Wild Fish Rescue boosted fish numbers temporarily. Then the numbers dropped as soon as humans stopped rescuing. “The second you take your foot off the gas, the benefits are just going to go right back down because you haven’t actually changed the capacity for the environment to support fish,” Armstrong explains.
The result is a salmon run dependent on humans to survive, the same problem that dogs fish hatcheries, the widespread but controversial practice of releasing artificially reared fish in the wild. Fish rescue, however, does not equal a salmon hatchery in one way: intention. David Brown built salmon tanks on his property to preserve wild salmon, not breed them. Still, the outcome was largely the same, as Beebe found. After Beebe published her thesis in 2019, Brown, who didn’t return requests for an interview, was disappointed, according to Armstrong. “Yeah, he said Brittany got an F,” Armstrong says with a rueful laugh. He understands Brown’s reaction. “If somebody evaluated my parenting and told me about all the potential things I’m doing wrong,” he says, “I probably wouldn’t be thrilled.”
For now, fish rescue has largely avoided the scrutiny that hatcheries attract, simply by dint of obscurity. “You start talking about hatcheries and people are gonna start getting red in the face,” Armstrong says about the negative backlash he’s seen against hatcheries. “But then somebody can scoop up an ESA-listed coho, bring them into a hatchery-like environment for a year, and it escapes controversy? I think we’re going to see [fish rescue] become a bigger issue.”
Before beginning the study, Beebe dug into the field, searching for examples of both rescue and salvage. Salvage is far more common and well known. Google “fish rescue” and you’re likely to end up on Ohio Fish Rescue’s YouTube channel watching Shaquille O’Neal adopt some gigantic koi. But a few official fish rescues are beginning to pop up in California and Australia—arid landscapes that portend the climate disasters awaiting the rest of the planet.
Gabriel Rossi, a postdoctoral ecologist at the University of California, Berkeley, echoes Armstrong and Beebe’s concerns. Rossi, who studies water-stressed streams in Northern California where fish salvages occur, had never heard of this new way of saving salmon practiced at Northwest Wild Fish Rescue. Confusingly, in California, fish salvage is sometimes called fish rescue. The scientific literature has not settled on a unifying term or definition either—an indication of just how understudied the field is. Fish salvage can be called translocation, trap and haul, or assisted migration, while fish rescue might be called an ark strategy or captive rearing. When I explain the captivity stage of fish rescue to Rossi, he has an immediate gut reaction: “I would be worried about that. It seems like a much more heavy-handed management approach,” he says, “but I’m not saying there’s not a place for it.”
Fish rescue may have a place in conservation by buying time for critically endangered fish while a waterway is replenished or restored, says Beebe. Solving big systemic issues, like repairing a damaged habitat or regulating an over-tapped water table takes years. But the danger of introducing easier stopgap solutions is that we come to rely on them and delay fixing the larger problems because they’re more difficult.
After Vancouver Island’s record-breaking drought in the summer of 2019, the winter followed with record-breaking rain and snow. In January, the biggest snowstorm in decades struck the region. Then the rain kept pouring down all the way into April. By May and June 2020, Bernhardt began checking on the Robertson again. She was relieved to find it gushing with winter rain. By late July, the pebbly bottom broke through the stream and by early August, she was back to salvaging fry. On a warm Saturday morning in mid-August, Bernhardt headed out once again, packing a fishing net, a few buckets, and her two dogs into her Dodge Caravan.
“There’s nothing funner, being a little kid again, hanging out in the puddles,” she says, narrating her day in a laid-back vernacular: lots of “bummer” and “get this” and “ya know?” It’s not yet noon as she cuts through an overgrown path to the river, but the air is already hot and heavy with the sweet scent of broom. Mosquitoes and bees wander in and out of the sunlight that streams through the tall Douglas firs overhead. Her dogs run ahead, bursting onto the dry riverbed where Ellie, a big German shepherd, tramples through the first puddle. “Ellie, c’mon!” Bernhardt groans. “We were going to work that one.”
The Robertson River has a long history of fish salvage, Ted Burns tells me later over a phone call. Burns, a retired biologist who once worked for DFO, came across drawers of documents tracing salvage in the region back to the 1930s. The Robertson is one of 12 major tributaries that feed into the Cowichan River system. During droughts, the Robertson was already prone to drying out, but a legacy of clearcutting made it worse. “The forests here have just been brutalized,” says Burns. When the rain fell on the stripped hillsides, they eroded more quickly, raising the Robertson’s streambed and interrupting the flow. Back then, federal fisheries officers salvaged fish in hand-built wooden tanks as big as coffins. When logging techniques improved later in the 20th century, development ratcheted up the destruction of freshwater habitat. Lakeside development clawed back the shoreline, while all the fixtures that come with sprawling urbanization—roads, parking lots—chipped away at the watershed. Still, a sizable population of coho continued to return to the 15-kilometer-long river.
Despite all the intrusions, the coho thrive in the Robertson’s lower reaches, particularly the last 10 kilometers before it runs into Bear Lake. Among the 20 or so tributaries that Burns has studied in the Cowichan River system, he says that the Robertson is special: “Some of [the tributaries] only produce a hundred fry or a few thousand, but the Robertson River typically produces several thousand, if not more. Maybe a hundred thousand in a big year.”
By the 1970s, the government salvages had lapsed, leading a local furnace repairman, Leo Nelson, to take over the effort and found the Cowichan Lake Salmonid Enhancement and Hatchery Society. “The Robertson River was always on Leo’s list,” says Burns, who joined the society and its salvages of wild runs on the Robertson. Sometimes, when the distance between pools was too far, Nelson drove his truck along the gravelly riverbed, destroying the truck’s undercarriage, but saving the fish. Around 1985, a logging crew took a lunch break on the banks of a Robertson River tributary called Easy Creek and noticed tens of thousands of salmon stranded in a pool. The crew called Nelson, and the Cowichan Enhancement Society came down with their nets and saved 85,000 fry that day. The construction crew joined in, chasing fry with their hard hats and saving another 5,000.
Bernhardt, who moved to nearby Mesachie Lake six years ago, learned about the society when she was on the hunt for a better net. The society loaned her a specialized salvage net and a bubbler to aerate the buckets of water. In return, Bernhardt reports her salvage numbers to the society, which routes the report over to DFO.
Wading into the pool up to her ankles, she maneuvers the net into position. The fish skitter to the opposite end. The coho do not want to be saved. She stalks closer and opens the net, which is draped like a sheet between two wooden poles and weighted at the front with a heavy chain. She keeps that front edge pinned to the river bottom to catch any breakaways. “You gotta get the front end up real fast, because as soon as they know they’re trapped, they’ll sprint,” she explains. In one swift movement, she raises the net from the stream, and dozens of silver-brown fry appear, wriggling in the sunlight.
Every creature—insects included—and every plant Bernhardt encounters during the day receives a pronoun: he, she, they, whatever fits, as though she’s on a first-name basis with nature. She bends down to pick something from a puddle and holds up what looks like a dusty cobweb, a look of wonder on her face. It is actually the desiccated sac of a caddisfly larva, which coho fry like to eat: “There are 14,500 species of caddisfly larvae!” she exclaims. Her open, expressive face cycles through dozens of emotions in as many seconds. The most prevailing mood is an eager, infectious excitement about nature.
She walks farther down the dry riverbed, the rocks crunching under her practical outdoorsy sandals. “They want to stay local,” she says, pointing out a channel she dug to connect two pools and give the trapped fry more water. And yet the fish did not want to budge. They pick a safe hidey-hole in the rocks and stay put. This reminded me of a 2014 study that Armstrong sent me. The study showed that Oregon coho fry stay close to where they hatch. Those results, Armstrong pointed out, could make a case for salvagers who disperse fish to new, underused habitat within the same waterway.
“We were always worried about how valuable our efforts really were,” says Burns of the Cowichan Enhancement Society. “So for a few years we marked the fry.” They found that about 10 percent reached the smolt stage—a decent survival rate in the wild. But other studies report less rosy results. A 2012 study found that not one of 26 chinook relocated around a dam in a tributary of the Sacramento River survived to spawn. Another 2017 study picked through two decades of extensive conservation efforts on California’s Carmel River where ocean-going steelhead continue to disappear. Both rescue and salvage took place there, sometimes as many as 20,000 salvages in a year, and it would be natural to expect some return on the investment. The researcher’s best conclusion found too many young fish were added to the river, stymieing their growth and migration to the ocean.
Salmon may not need as much help as we think. A seven-year study tracked four tributaries of a California river stocked with hatchery coho, which are less robust than their wild counterparts. Even during the extreme droughts of 2014 and 2015, and without any human intervention, nearly 50 percent of the stranded salmon survived until the rain raised the river once again. Stories passed down through Indigenous communities along the Columbia show the Pacific salmon’s resilience to disaster. Over 500 years ago, a landslide blocked the Columbia River just east of Portland, Oregon, but the salmon regained their foothold above the blockage when the landslide naturally cleared.
Then again, the world is a very different place today than it was 500 years ago. Wildlife is navigating a natural world manipulated by humans on a massive scale. Throw in evaporated habitat, and that might be the final stumbling block for the salmon populations that have persisted this long. In that case, would it be unethical to stand back and let nature take its course? “As fish managers, we never fully answer what we ought to do,” says Rossi with a sigh.
Sometimes Bernhardt becomes so engrossed chasing fry in puddles that the ones she’s already saved slip her mind for a moment. When the fry become sluggish and move to the surface of the warming buckets, it’s time to move. They need to be released back into cool waters, and fast. Hoisting the bucket in her powerful arms, she starts walking down the dry riverbed, hopping over ankle-twisting rocks with surprising agility. She has never dropped a bucket, she says proudly, or lost a single fry in the heat.
Despite the hazy terminology and unclear utility of fish salvage, more and more environmental agencies seem to be embracing and funding it. After the 2019 drought that killed so many fish in Australia, the government announced an AU $10-million rescue and relocation program for future emergencies. A water activist and local tourism operator said that amount won’t begin to pay for the damage inflicted by years of overextracting water and for the loss of threatened keystone species, like the Murray cod that take over a half-century to recover. Another ongoing fight over water in California’s Yuba River shows that small, targeted volunteer efforts, like Bernhardt’s, are probably more sustainable than building infrastructure to support long-term, state-sanctioned salvage.
The Yuba River drains the snowy western slope and foothills of the Sierra Nevada in central California. Since the early 20th century, hydroelectric dams and irrigation siphoned off water for the growing agricultural communities nearby. The Yuba River, which feeds into the larger Feather River, was once a chinook stronghold, but in 2017 the population hit a record low. The population is mostly propped up by strays from nearby hatcheries in Feather River and elsewhere. In 2015, the Yuba Salmon Partnership Initiative, a coalition of government agencies and environmental groups, proposed spending up to US $500-million on trucking wild salmon around two dams on the Yuba River for the next 50 years. The proposal has stalled since then, but biologist Matt Stoecker and Patagonia founder Yvon Chouinard argued in an op-ed that fish salvage “would put even more concrete in the river in the form of collection facilities and create an overly complex workaround powered by diesel fuel.” Both men advocate in their 2014 documentary DamNation for removing the dams and restoring the river to a more natural state. After the largest dam removal in history revitalized the Elwha River on the Washington side of the Juan de Fuca Strait, the river welcomed back impressive numbers of steelhead and chinook salmon. Dam removal has gained traction with the broader public, perhaps because it evokes a romantic (and also convenient) idea that nature can heal itself better than humans can.
As a veterinarian, Bernhardt has an eye trained to spot suffering in animals. Back at the Robertson River, she grabs a handful of foliage from the bushes and covers the top of every bucket of coho with leaves and pine needles. “Why do you do that?” I ask her later, on a phone call. She noticed that the fry swam frantically in stressed circles in the buckets, so she covered them with leaves to mimic their hidey-holes in the wild. After a while, they relax into a lazy, gentle swim. Seeing how much effort and care Bernhardt brings to saving fish, I feel guilty questioning the value of all her hard work.
“What would you say to people,” I ask carefully, couching my question in hypotheticals, “who say, you know, that we shouldn’t interfere with nature? Who say, ‘What world would this be if we had to save wild fish?’”
“It would be a compassionate world,” she shoots back, her voice catching. “If we wanted to not interfere with nature, we better start getting a grip on how to live in harmony with nature. Because we’re creating all these problems, so it’s our responsibility to show up.”
Brown likely feels the same about the hundreds of thousands of coho he’s rescued over the years in Washington State. Both salvagers and rescuers make a sincere attempt to save life, but Armstrong sees a line dividing the two. Salvage, says Armstrong, “is lowest-risk, lowest-reward.” The damage to wildlife appears to be minimal, and there’s a chance it might help—more coordinated research is still needed. Fish rescue is harder to defend. Decades of research have already shown that hatcheries are not restoring, and may even be harming, wild runs.
The modern world can feel like a trap sometimes. A seemingly simple act, like saving a fish, can backfire. But the opposite approach—doing nothing and leaving nature to its own devices—is no solution either. As the climate warms, fish habitat is bound to clash with society’s growing need for water. If freshwater fish have a hope of surviving in the wild, humans will have to play some supporting role. On the west coast of North America, with its recurring droughts and dwindling water supply, that means managing watersheds more rigorously to ensure that fish are never left high and dry.
Oddly, the vilified dams of the past could help with that. “In this new Anthropocene era that we’ve entered into, ironically enough, some of the watersheds with dams are going to do better because they’re human-managed,” says Tanis Gower, a biologist at the Vancouver, BC–based Watershed Watch Salmon Society. In some watersheds, dams holding water back in times of plenty will ensure a consistent flow during times of droughts. This will be critical for freshwater fish to survive in British Columbia where 63 percent of people live in water-stressed regions with groundwater supplies already maxed out, a 2019 Watershed Watch report found. In the face of a dwindling water supply, the nonprofit advocates for setting minimum flow requirements across BC waterways and managing each watershed on a case-by-case basis, rather than keeping the blanket laws in effect now. This individualized approach feels like a step forward from the past. It recognizes that every river and its resident salmon are unique and have their own story to tell.
Bernhardt stands up to her knees in the cool water of Robertson River. Instead of hauling the coho to her van a kilometer down the road and then into Bear Lake, she decides to release them in a deep pool that has collected around the pilings of a bridge near the lake. The autumn rain is only a few days away; the river will rise soon and release the coho from their puddles. Thankfully, 2020 was a better, wetter summer for the coho of Robertson River.
Bernhardt doesn’t just dump the coho in the water and leave. It’s a slow release, lasting about 10 minutes. First she tips the bucket to dribble in more and more water, acclimating the fish to the pool’s cool temperature, until at last she turns the bucket horizontal and allows the fish to swim out on their own. A few coho hang back in the bucket, hiding like they did in the gravel riverbed. “Go find a new home,” Bernhardt says, encouragingly. She’s in no rush. Watching the fish swim away is part of the reward.
Field reporting for this story by Jude Isabella and Vanessa Minke-Martin.