The Klamath River salmon die-off was tragic. Was it predictable?

A recent large die-off of young salmon released into the Klamath River shocked and dismayed state biologists, reinforcing that human efforts to restore nature and undo damage can be unpredictable and difficult  to control. 

The tiny Chinook salmon turned up dead downriver just two days after they were released from the California Department of Fish and Wildlife’s brand new Fall Creek Fish Hatchery, built to supply the Klamath River as it undergoes the largest dam removal in history.  

The $35 million state hatchery, on a tributary just upstream of Iron Gate dam in Siskiyou County, was constructed to help the river’s threatened coho and dwindling fall-run chinook salmon, a mainstay of commercial and recreational fishing and tribal food supplies.

The hatchery’s first release ended with an unknown number of the 830,000 young Chinook salmon found dead, their eyes bulging, in a federal sampling trap about 9 miles below the dam.

State officials called it “a large mortality,” but said there’s no official count yet and released no additional details about the size of the die-off.

California’s fish and wildlife officials said they suspect “gas bubble disease,” a condition similar to decompression sickness in scuba divers, is to blame — likely caused when the salmon traveled through a 9-foot-wide tunnel out of Iron Gate dam to reconnect with the Klamath downriver. 

Gas bubble disease in fish is caused by “environmental or physical trauma often associated with severe pressure change,” officials said.

Jason Roberts, inland fisheries program manager with the state agency, said it’s an outcome that state, federal, and tribal scientists involved in the decision didn’t anticipate.

“The basin co-managers made the best decision they could with the information that they had, and unfortunately, it did not go well,” Roberts said. “I don’t think anyone thought water going through this tunnel would cause gas bubble disease, or we obviously wouldn’t have done it.”

“You see gas bubble disease at hatcheries when there’s flood flows and there’s tons of water,” he added. “I think everyone assumed, given Iron Gate reservoir was basically drained, that there wouldn’t be a problem.”

“I don’t think anyone thought water going through this tunnel would cause gas bubble disease, or we obviously wouldn’t have done it.”

Jason Roberts, california department of fish and wildlife

It was a tragic, heartbreaking event for the fish biologists and a setback for a costly and high-profile project: The state had hatched and raised the salmon, then released them into a place where they died almost immediately. 

“I feel really bad for the fish. And I feel really bad for my staff that spend all their time taking care of these fish, and for all the partners and stakeholders that are counting on us,” Roberts said.

Ecologists say that taking steps to restore nature or undo environmental damage — such as constructing new wetlands to replace bulldozed ones, building passages for wildlife and providing new habitat for endangered species — can be unpredictable and have unintended consequences. Failure is always a possibility.

Toz Soto, fisheries manager for the Karuk Tribe, commended the efforts to remove the dams and restore the river. The Karuk Tribe’s ancestral territory covers hundreds of miles along the middle of the Klamath River, and salmon are a key part of their traditional diet.

“While there is a level of uncertainty in anything you do, especially a massive dam removal project like this, they’ve done a really good job of predicting what’s going to happen, especially with sediment movement and transport, poor water quality,” Soto said. Despite the tragedy, he said, “I’m overall really happy with the outcome so far.”

No wild salmon were harmed. And the consequences aren’t expected to be catastrophic for the Klamath hatchery project.

“This is just a sad reminder of the harm that these dams have caused, and continue to cause.”

Robert Lusardi, Uc davis Department of Wildlife, Fish, and Conservation Biology 

The hatchery is still raising 3.27 million healthy chinook salmon, more than the state’s annual goal of 3.25 million, to be released into the river. From now on, officials plan to bypass the dam and truck the remaining fish downriver.

“We believe the effects to salmon populations will therefore be minimal,” said Michael Belchik, senior water policy analyst with the Yurok Tribe, the largest tribe in California with a reservation that spans a 45-mile stretch of the lower Klamath River. 

And by the end of the year, Iron Gate dam should be gone. 

“Iron Gate’s killed its last salmon,” Belchik said. 

The die-off came as four aging hydroelectric dams are being cleared from the Klamath River’s path, reconnecting a river that has been divided for over a century. The reservoirs have been drained, and the river is reclaiming its track. One of the dams is already gone. 

California tribes and conservation groups have been fighting for dam removal for decades. Reconnecting hundreds of miles of habitat and restoring more natural flows to the river is expected to help revive the area’s salmon. 

The die-off also came amidst increased scrutiny of the dam removal, which opponents in Siskiyou County have called “the largest, most devastating dam removal experiment in modern history.” They warn that it is creating a “river of death” that will unleash sediment and other materials into the river — jeopardizing the fish.

But wildlife officials say the river’s sediment and dissolved oxygen levels were suitable for releasing the fish and weren’t a cause of the die-off. They say it’s likely that conditions within the Iron Gate dam’s tunnel killed the salmon — reinforcing the danger of these vestigial structures. 

“This is just a sad reminder of the harm that these dams have caused, and continue to cause,” said Robert Lusardi, assistant professor of freshwater ecology at the University of California, Davis. “It does speak to the need to immediately remove the dam infrastructure from the Klamath River.” 

Mark Bransom, CEO of the Klamath River Renewal Corporation, the nonprofit formed to oversee the removal effort, said it also speaks to the challenge of undoing decades of human environmental interference. 

“We cannot unwind over 100 years of environmental damage overnight,” Bransom said. “We really do have to take a little bit longer view of what restoration means when we are trying to unspool a century’s-worth of impacts to the environment and to the communities who depend on a healthy river.”

From healthy hatchery newborns to dead in the river

The salmon were healthy when they left the hatchery on Feb. 26. Only months old and an inch or two long, they were released as scheduled to give the remaining fish in the hatchery ample room to grow.

The salmon fry traveled down Fall Creek and into the Klamath River as it snakes through the footprint of the now-drained Iron Gate reservoir, before swimming through the roughly 200-foot long, 9 foot-wide tunnel through Iron Gate dam and back into the Klamath River. 

They began showing up dead two days later in a U.S. Fish and Wildlife Service sampling trap miles downstream. 

Roberts said their bulging eyeballs were a telltale sign of gas bubble disease. A definitive confirmation, though, is unlikely from carcasses that traveled so far.  

Much like with the bends, gas bubble disease occurs when air bubbles form in blood and tissue. In fish, it can be caused by severe pressure changes or in water containing excess levels of gases. 

Bransom said pressure conditions were normal where the river flows into the tunnel, which was flowing only about half full, he said.  

“There has been no definitive conclusion that this was gas bubble disease and that the tunnel was responsible,” Bransom said. “It’s possible there was some hydraulic condition inside of the tunnel that produced a high pressure zone, meaning higher than atmospheric pressure, but again we don’t have any evidence of that, and we don’t know what the level of mortality actually was.” 

Roberts said it’s “near impossible” to put a number on how many fry died at this point. 

“It might not ever be a firm number that anyone knows,” he said. “The majority of the fish in the trap that were fry-sized were deceased, and they had visual signs of gas bubble disease. Therefore, it’s a safe assumption to think that if those fish swam through the tunnel, other fish that swim through the tunnel likely experienced a high rate of mortality as well.”

He added, “thankfully we have an extra 850,000 fish on hand so that will help offset the loss of these fish — from a production standpoint.” 

“So far, we’ve been finding normal fish, healthy fish. So that’s really good — that the wild fish are not affected by this.”

Toz Soto, Karuk Tribe Fisheries

Still, the loss comes at a delicate time for the Klamath dam removal effort, which began in earnest last summer when the first dam was demolished.

“While the project’s going through a lot of scrutiny, this is just another thing for naysayers to pile on to,” said Curtis Knight, executive director of California Trout, a conservation organization. “But I think it’s important to note the dam is not going to be there much longer — that’s what caused these fish to die, and conditions in the river below are improving.”

The dam removal nonprofit group reported on Feb. 15 that 5 million cubic yards of sediment is expected to be released during the months-long drawdown process, and that there had been a significant die-off in the reservoirs during the drawdown. Bransom said they were non-native fish that had thrived in the warm water and that the deaths were expected during the drawdown.

But officials said sediment levels and dissolved oxygen were reading at suitable levels when the fish were released. Also, older fish that had traveled from downstream of the dam and were snagged in the same sampling trap showed no signs of the disease, which suggested that river conditions were not to blame. 

And conditions are improving, even over the last month, said Soto of the Karuk Tribal Fisheries Program.

“Our crews have been out sampling the main stem below the dam because we have concerns with sediment transport and poor water quality, so we really wanted to know how wild salmon juveniles are doing,” Soto said.  “And so far, we’ve been finding normal fish, healthy fish. So that’s really good — that the wild fish are not affected by this.”