It seems a contradiction: A California hatchery charged with ensuring salmon procreate despite the dams blocking their native runs is killing fish by the thousands without allowing them to breed.

Since Sept. 17, the Feather River Hatchery in Oroville has been killing all salmon that enter the hatchery near the base of Oroville Dam. These fish are not being bred to make more salmon, as has been standard practice for decades, but instead are being donated to area food banks.

The reasoning is rooted in new efforts to strengthen the species. The California Department of Fish and Game, which operates the hatchery, says it can't be sure whether the fish entering the hatchery at this time of year are fall- or spring-run salmon. Rather than breed them anyway, as it has in the past, Fish and Game is culling them from the population based on research that shows it's healthier for the species to keep the two runs genetically distinct.

"The operation has been critiqued by academia and conservation groups for quite awhile," said Stafford Lehr, fisheries branch chief at the Department of Fish and Game. "It was clearly pointed out that we needed a clearer separation of the runs, and this is where we start doing it. We're taking this opportunity because it's a good year for (salmon) returns."

The new practice was expected to end Friday.

As of Thursday, about 7,800 salmon had been "culled" from the hatchery without being bred. These fish will go to food banks, including those in Butte County and the Enterprise Rancheria, a branch of the Maidu Indian community. Some may also make their way to Sacramento via the regional Foodlink program.

Losing these fish from the breeding program does not sit well with fishing groups. Their members are eager for every opportunity to boost salmon populations after three historically bad seasons.

"I think it's a bad idea," said Zeke Grader, executive director of the Pacific Coast Federation of Fishermen's Associations. "The hatchery was built to mitigate not only for losses of upstream habitat but also the operation of the (water) projects. It flies in the face of their legal obligations."

The problem is that spring- and fall-run salmon cross paths at the hatchery for a period of time each fall, and it is difficult for hatchery managers to tell them apart. Called "tweeners," this body of fish has historically been bred at the hatchery despite the fact that they have completely different migratory patterns.

The result has been a loss of genetic diversity for both runs, which recent research shows threatens their survival by weakening their ability to survive changes in the environment. It also runs contrary to the Endangered Species Act, which considers the spring run a threatened species.

"We were concerned about how would it be perceived by not utilizing them," Lehr acknowledged. "But it's very clear we need to do everything we can to keep these runs separate."

The two runs cross paths because of Oroville Dam, built in 1968 by the California Department of Water Resources as part of the State Water Project.

Before the dam, spring-run salmon swam far upriver in spring and summer. High flows caused by snowmelt allowed them to reach high-elevation portions of the river, which would become inaccessible by fall. These fish would then wait through the summer in deep pools, and spawn in the fall.

Oroville dam eliminated access to this upstream habitat entirely. The hatchery was built to allow the species to survive despite this obstacle.

The fall run moves upriver in late summer and fall, and because flows are diminished, it could not reach the pools where the spring run waited to spawn. Instead, fall-run salmon would spawn in lower-elevation areas.

Now, spring-run fish swim as far as the hatchery, where they simply wait for the genetic signal to start spawning. By then, the fall-run salmon have arrived behind them to start their own spawning run.

Outdated hatchery practices caused the species to become at least partially interbred. In some cases, remaining wild spawners from both species bred together in the river while queued up below the dam.

One effect is that the spring run has begun to arrive later, and the fall run earlier. This led to the "tweener" period in which hatchery managers can no longer be sure which species they are breeding.

Culling this group is a first step toward distinguishing the runs again.

"If we can get that separation back and have the spring run spawning prior to the fall run coming back, over time this should become less and less needed," said Kathy Hill, a fisheries program manager at Fish and Game.

The practice of culling these fish is new to the Feather River Hatchery but has been a regular practice elsewhere, including at Trinity River Hatchery, also operated by Fish and Game. Nor is it unusual that the fish are being killed. Even in the wild, all salmon die after spawning.

Hill said the hatchery expects to meet spawning targets for both spring and fall runs despite the culling. For spring run, the annual goal is to breed 2,000 fish and produce 2 million juvenile fish, or smolts. For fall run, the goal is to breed 6,500 adults and 6 million smolts.

What happens at the Feather River Hatchery is important because it is the largest producer of spring-run salmon. Although the species still breeds in the wild in Sacramento River tributaries such as Butte, Deer and Mill creeks, the hatchery is critical to sustaining the population.

It's also a major producer of fall-run salmon, which underpins the commercial fishery in the ocean. That's why Grader opposes culling. Until habitat improvements help more salmon breed and survive in the wild, hatchery production is vital, he said.

"I don't think this is how you protect your wild fish stocks," Grader said.

Eliminating the overlap of fall and spring runs was one recommendation in a June report by a science panel that reviewed California hatchery practices. And while culling is a crude practice, several independent fishery experts told The Bee they support it as an interim measure.

Keeping the runs distinct is important, they said, because each has unique adaptations that may help salmon survive future environmental disruptions, such as climate change.

"It's never a good thing when you have to kill fish," said Brad Cavallo, a senior fisheries scientist at Cramer Fish Sciences, a consulting firm. "But in the interim, it makes sense to try to force some separation between the two runs."

© Copyright The Sacramento Bee. All rights reserved.

• Read more articles by Matt Weiser

Order Reprint