The long-awaited draft EIS for the Columbia River System Operations (CRSO) will soon be completed and released. As a reporter, I want to be prepared for this important document. What’s in the final EIS and subsequent Record of Decision will have a big impact on the future of the region. It may play a role in whether the parties involved in National Wildlife Federation et al. v. National Marine Fisheries Service et al. go back to fighting it out in court, or whether the region takes a step toward resolving these longstanding issues.

As I was pondering the best way to prepare for the EIS, I received a couple of emails from readers pointing to juvenile survival.

And so it dawned on me: this EIS will analyze many things—from the many impacts of tearing out the four lower Snake River dams to how dam operations, spill and total dissolved gas affect fish. But a problem that keeps arising in the Snake River is juvenile survival. It’s a complex and confusing topic with a long history, so I took a deep breath, started researching the issue and began making calls to people who know a lot more about this than I do.

After a few weeks of gathering information, what is most striking to me is that when it comes to evaluating juvenile survival, there’s a lot of noise. Noise, as in unwanted information that can drown out the information you want.

There's noise in figuring out what's causing the mortalities; there’s noise in determining the best way to measure juvenile survival; and there’s noise in the year-to-year comparisons of those measurements. From water volumes and velocities to the timing of spring melt and the number of predators, the natural world that juvenile salmon and steelhead find themselves in is constantly changing.

To complicate things further, operations at all of these dams are also changing from year to year—adding more noise. In 2018, we spilled to spill cap levels for 24 hours a day. In 2019, we spilled to 120 percent total dissolved gas for 16 hours a day. Under this year’s flexible spill agreement, spill will go up to 125 percent TDG at most of the eight dams for 16 hours a day.

Just to be clear, if you’re looking for me to confirm your worldview that the dams should be removed, or that the dams should not be removed, you can stop reading here. This is just a look at juvenile survival in the context of this EIS. There are so many other things to consider when it comes to removing the dams. With my basic understanding of the whole picture, I wouldn’t feel qualified to make that call anyway. What I will say is that both sides make compelling arguments related to juvenile survival, and that’s what I’m exploring.

First, here’s a glimpse of the issues involved with juvenile survival—and my apologies to readers for whom this is basic information.

There are many different ways that scientists measure juvenile survival in the Columbia Basin. NOAA Fisheries provides an annual report on direct juvenile survival, which is a detailed look at how many PIT-tagged fish make it past each dam, and how many of those make it past the next, and the next. The Fish Passage Center also tracks and makes available reams of information through its Juvenile Survival Queries, including an annual report with juvenile survival estimates, and a multiyear reach analysis tool that provides survival estimates and fish travel times.

For a more complete picture on how well each of those PIT-tagged juveniles survived, we have to wait a few years for smolt-to-adult ratios, or SARs, which tell us how many of those juveniles make it back upriver, and presumably to their hatchery or spawning area. A goal in the Northwest Power and Conservation Council’s Fish and Wildlife Program is to achieve smolt-to-adult return rates from 2 to 6 percent, with a minimum of 2 percent and an average of 4 percent, for listed salmon and steelhead. Wild Snake River spring/summer Chinook and steelhead continue to fall short of those objectives.

The SARs are also used to determine whether juvenile fish that were transported downstream by barge returned in greater or lower percentages compared to the fish that remained in the river. Those survival rates are measured in TIRs, or transport-to-inriver ratios.

Each of these measuring devices is completed for each of the Columbia Basin’s 13 endangered or threatened stocks, or, for purposes of the Snake River dam debate—the four ESA-listed salmonids in the Snake River. So a bad survival year for juvenile Snake River sockeye isn’t necessarily a bad year for juvenile Snake River steelhead.

In trying to analyze juvenile mortality in the Columbia Basin, there are two life-cycle modeling studies. The Comparative Survival Study (CSS) is a joint project of the Fish Passage Center, the Columbia River Inter-Tribal Fish Commission, fish agencies in Washington, Oregon and Idaho, and the U.S. Fish and Wildlife Service. COMPASS is NOAA Fisheries’ model, which predicts survival and passage behavior through the hydropower system. NOAA uses separate models for adult returns.

According to the EIS on the Washington Department of Ecology’s proposal to increase spill to 125 percent total dissolved gas, the main differences between the two are that NOAA’s model does not factor in the same assumptions about delayed mortality and makes different conclusions about the relative benefit of transportation as an alternative to spill.

Delayed or latent mortality occurs subsequent to dam passage—in the estuary and the ocean—and is directly related to their hydrosystem experience. The 2019 CSS includes a new chapter that reviews studies supporting the theory and those that don't.

Whether or not you factor in latent mortality is a big deal. The studies supporting latent mortality show that between two-thirds and three-quarters of the juvenile salmon that make it through all eight Snake and Columbia river dams later die from latent mortality. And that’s after losing roughly half of them due to direct injuries and stress of the journey.

But a recent NOAA study found that adult returns depend more on the size of the juvenile than its passage route through hydroelectric dams, and questioning the magnitude of benefit that spill—or dam removal—would provide.

Michele DeHart, manager of the Fish Passage Center, told the Southern Resident Orca Task Force last year that the 2017 CSS looked at the potential of removing the four dams, and concluded that the highest adult returns would come if those four dams were removed, and the four lower Columbia River dams spilled to 125 percent TDG. Under that scenario, Snake River Chinook SARs would increase to two percent under poor ocean conditions, and to 11.3 percent when ocean conditions are good.

Jim Faulkner, a research statistician at NOAA’s Northwest Fisheries Science Center, said NOAA models also predict increased inriver survival if the four dams are breached, due to decreased travel times, potentially decreased river temperatures, and a reduction in the direct mortality. That modeling was done for the EIS, which has not yet been released.

The two models, and their corresponding theories, may lead us to two very different conclusions about whether or not removing the dams is a good idea—just from the perspective of juvenile survival.

Jim Faulkner, research statistician at NOAA's Northwest Fisheries Science Center, defended the COMPASS model as much more detailed than the CSS, noting it is the one accepted by the region for predicting salmon passage.

Ed Bowles, Fish Division administrator for the Oregon Department of Fish and Wildlife, said there's a broad consensus of scientists behind the Comparative Survival Study—including tribes, the states of Washington, Oregon and Idaho, and the U.S. Fish and Wildlife Service.

"Even though those states and entities may disagree on policy, this is still their scientists' view of the world and the ocean," he said.

He told me one way to understand why latent mortality makes sense, even if the precise cause is not understood, is to compare adult fish returns (known as smolt-to-adult return ratios, or SARs) to Lower Granite Dam (the fourth dam on the Snake River) with adult returns to the John Day system in Oregon (the third dam up the Columbia River main stem).

Spring Chinook from both watersheds experience the same estuary and the same ocean conditions, he noted. Bowles said that prior to completion of the Snake River dams, those two stocks had similar life cycle or SARs survival rates. But once the Snake River dams were completed, the interior stocks' SARs became two or three times lower than the John Day stocks. Both stocks experienced declines in years following poor ocean conditions, he said, but the percentage of adults returning to the Snake River is still two or three times lower than those returning to the John Day system.

Bowles said that when it comes to in-river survival and the issue of transporting fish—because all fish can't be transported—it's important to optimize their river conditions. TIRs (transport/inriver ratio) are a measurement of how well transported fish survive and return as adults compared with juveniles that migrated within the river system.

"In the past, when we did a lot more transportation, we actually never tried to optimize conditions through spill or even flows," he said, so it makes sense transported fish were doing better. In addition, he said, TIRs compare transported fish to fish going through the bypass system, or through turbines, since there are no PIT-tag reading devices on spillways. "So those are biased in terms of true in-river fish that never experience the powerhouse," he said.

Perhaps most importantly, Bowles said, "Neither one of those are currently working." And if neither method is enough to sustain the population, it's time to find a method of downstream survival that will.

Scott Levy, director of, was the author of one of the two emails I received that got me thinking about juvenile survival. He shared a graph of NOAA Fisheries' juvenile survival data showing that yearling Chinook—hatchery and wild combined—on the lower Snake River has not improved overall since 1995. It also showed survival has been particularly poor the past two years. Ocean conditions or a multitude of other factors farther downstream do not influence these numbers, he noted.

Levy said spill is the only major change he can think of to account for the recent drop in juvenile survival. The spill in 2018 and 2019 increased total dissolved gases for the entire spring to between 115 and 120 percent—levels that have not been modeled. He said he's especially worried about the flexible spill, which was implemented in 2019. He believes seiches—or oscillations in the water similar to back-and-forth sloshing in a bathtub—are becoming a serious problem, and juveniles may be getting lost in the confusing flows and starving while seeking a course downstream.

Levy said that the percent of total dissolved gas was not used in the Fish Passage Center's models suggesting that increased spill will benefit juvenile migrants. He suggests the models should be redone before this year's spring spill begins, using the past two years of data. "Millions of juvenile salmon are perishing in the lower Snake River and we should not be going into this next spill season unnecessarily blind," he said.

Darryll Olsen, board representative for the Columbia-Snake River Irrigators Association, is the other person who emailed me about juvenile survival. Olsen created a graph showing the relationship between in-river juvenile survival and transport-to-inriver survival ratios. When I called him to ask what the graph means and why he pulled it together, he said, "On the question of dam breaching or drawdowns, it always comes down to, what is the scientific crux of the argument here? Forget the politics. What do the numbers say?"

To Olsen, the science has focused on juvenile survival, and the real crux of the issue is whether more juveniles survive and return as adults if they're transported past the dams compared to those staying in the river. At one point, he noted, the policy for salmon managers was to "spread the risk." That means leaving roughly half of the juveniles in the river to make their way to the ocean through a series of dams, and transporting the other half around the dams, to be released below Bonneville Dam. But higher spill has forced that policy to go by the wayside.

From 2006 through 2016—the most recent comparable year for which adult returns are available—a higher percentage of fish survived through transport than they did by remaining in the river in eight of the 11 years—significantly higher in three of those years, Olsen noted.

He started with 2006 because that's when spill and transport protocols changed. Olsen said his graph shows that, if the lower Snake River dams are breached, in-river survival from the Lower Granite Dam pool to the Bonneville tailrace must be at least 56 percent or greater to equal transport survival. And according to numbers in NOAA's preliminary juvenile survival estimates for yearling Chinook this year, that's only happened three times since 1997.

If the dams are breached, Olsen noted, the option is lost of transporting juvenile salmon, which currently offers higher survival rates, and may be the best shot at saving salmon runs as rivers continue to warm.

Jim Litchfield, who represents Montana on the Columbia River Technical Management Team, is also worried about the impact of spill on juvenile survival. He said looking at the last two years of spill, the direct survival showed no improvement, and some reaches had worse survival. "It's not clear why, but I don't see how more spill helps. Increased TDG levels are clearly not good, and we're going to expose them to 400 miles of really high TDG. It feels like we're really playing with them," he told me.

Litchfield said that failing to maximize juvenile survival is a big frustration. At one point, he said, the U.S. Army Corps of Engineers and NOAA Fisheries conducted multiple tests at each dam to fine-tune juvenile passage to maximize survival. They used acoustic and balloon tagging, and measured direct survival at each dam, determining the best spill patterns at each.

At some dams, he said—such as Bonneville Dam—fish have a higher survival rate by going through the corner collector bypass rather than over the spillway. But the results of those tests were never put to use because a requirement to spill to certain levels was spelled out in the biological opinions, and later in court orders.

Litchfield said if NOAA's research is correct, the two models—COMPASS and CSS—are not measuring overall juvenile survival anyway, so comparing results doesn't clarify anything. NOAA's study showed larger fish tend to go over the spillway, while smaller fish go through the bypass system. "The fish going over the spillway survive better because they're bigger," he said, noting it's long been known larger smolts have a much better chance of survival.

Litchfield said every time a megawatt of would-be energy is spilled, another megawatt has to be generated by the power system to match demand. Some might be replaced by wind or solar, but some will be replaced with a thermal power plant, running on coal or gas. "If you think climate change is underlying this problem we're in, you ought to be pretty thoughtful about spilling carbon-free energy," he said.

Noting that anadromous fish runs up and down the West Coast are not doing well—including those returning to undammed rivers—he added, "I think continually keeping the spotlight on the dams is something a lot of folks want to do, but it may be diverting us from what we really need to be focused on."

Jim Faulkner, NOAA's Northwest Fisheries Science Center researcher who was lead scientist for the study on smaller fish tending to favor bypass systems, said it's tough to draw conclusions about the impact of spill on juvenile survival—yet.

He said NOAA's annual juvenile survival report is mostly a monitoring effort to ensure nothing seriously wrong is happening. "There are so many other things in the system, it's really hard to use this information to directly get at those questions," he said. Even drawing conclusions by looking at the last few years of higher spill is risky, he said, adding, "I think there are things we can't or aren't measuring that may be more influential. Things like water temperature, velocities, turbidity and the impact of predators in any given year may be driving a lot of the differences between years, he said.

Faulkner said NOAA scientists are still investigating why juvenile survival was lower during the last two years with higher spill. "It's hard to tell, but with the Snake River trap, you have to consider there could be some things you just don't know about. There could be some tagging related mortality, or changes in tagging practices," he said. In general, he said, one has to be careful about interpreting the data. "There are so many details that end up going unknown, and all kinds of little things happening at the dams," he said, adding, "I honestly think that ocean survival is really the big issue."

To Joseph Bogaard, executive director of Save Our Wild Salmon, the growing pressures on salmon due to climate change are very much a part of the issue. He suggested that focusing on juvenile survival may offer too few solutions to the problem of recovering salmon stocks. "If we get wrapped around the axle on the ins and outs of juvenile survival and all the details of the EIS, we are also asking the wrong question," he said. "What we really need is a conversation and a dialogue with key stakeholders, relevant sovereigns and policymakers that serves the need of our region." That, he said, will be the best way to find a solution that leads to healthy and harvestable salmon populations, a healthy, reliable and affordable energy system, and more certainty for communities.

Bogaard said he doesn't think the EIS will provide that solution. "If we continue to stay focused there, I think we just jump back on the merry-go-round with what has not served our region well for many years."

Bowles said when it comes to the upcoming Columbia River System Operations EIS, he'll be looking for a thorough and comprehensive analysis of the multiple objectives that is fair and honest and looks at all sides of the issues. It should be true to the National Environmental Policy Act, with the necessary ESA consultations, and with consequences of different alternatives outlined, and mitigation measures included for impacted sectors.

From Oregon's perspective, Bowles said, the flexible spill agreement is a bridge through the EIS process during which everyone could learn a little and stay out of court, but it was not designed to be the long-term solution. "We anticipate that a long-term solution may not be clarified" in the EIS, he said.

However, Bowles said, he's encouraged by several processes underway to help find solutions that benefit all.

K.C. Mehaffey covers fish issues for Clearing Up, and is editor of the NW Fishletter. She joined the NewsData writing team in February 2018. From lawsuits to scientific studies, she is enjoying the deep dive into the Columbia Basin's many fish topics.