One fish, in the dark.

A Pacific salmon is a keystone: primary prey

NOAA Fisheries

NOAA Fisheries, Southern Resident killer whales — Chinook are the primary prey. Southern Resident killer whale recovery · as of 2026

for the endangered Southern Resident orcas, the center of tribal foodways for millennia, and — when it returns to spawn and dies — the fertilizer that feeds the forest itself. 1978 of them supported a working fishery you could see from shore.

Then it thinned. Washington's salmon fishing fleet fell from 79 vessels (was 3,041)

NOAA Fisheries

NOAA Fisheries, West Coast salmon & steelhead. West Coast salmon status · as of 2022 · scope: Washington State fleet only — label as Washington, not the coastwide fishery.

— from three thousand boats to fewer than eighty in a single working lifetime. Today 28 population groups

NOAA Fisheries

NOAA Fisheries, ESA-listed West Coast salmon & steelhead (LA County to Puget Sound). ESA listings · as of 2026

stretch from Los Angeles County to Puget Sound.

It is not all loss. NOAA projected the 2026 West Coast ocean salmon season at >76 $M

NOAA Fisheries

NOAA Fisheries: the 2026 season 'will support more jobs and more than $76 million in revenue or value' — a ~63% increase in coastwide commercial ex-vessel value vs. 2025. 2026 West Coast ocean salmon season · as of 2026 · scope: Total revenue/value (commercial + recreational), NOT ex-vessel alone and NOT the Alaska-inclusive $2.4B FEUS figure. The commercial ex-vessel change is reported as a ~63% increase vs. 2025.

in revenue and value, a sharp rebound off closed years. And 1.8 $B

NOAA Fisheries

"As of 2023, PCSRF has appropriated $1.8 billion to state and tribal salmon recovery programs and projects." — NOAA Fisheries, Pacific Coastal Salmon Recovery Fund Pacific Coastal Salmon Recovery Fund · as of 2023

has flowed to states and tribes for recovery since 2000.

Start at the top of the scale. The U.S. Fish & Wildlife Service counts 7M+ barriers

U.S. Fish & Wildlife Service

"Across the United States, over seven million dams and road-related barriers are fragmenting rivers, blocking fish migration, and putting communities at higher risk of flooding." — U.S. Fish & Wildlife Service, The National Aquatic Barrier Inventory & Prioritization Tool. Inventoried-dam count (~561k) per the tool's live counter: 561,189 inventoried dams (tool.aquaticbarriers.org). National Aquatic Barrier Inventory · as of 2025

fragmenting the country's rivers — and the inventoried dams are only a small fraction of that count. The overwhelming remainder are not the walls you picture. They are road crossings: undersized pipes buried under a driveway, a logging spur, a state route, each one a small ledge or a too-fast chute a returning fish cannot climb.

Now collapse the scale. Zoom from the country to one state: Washington's wildlife agency has identified 17,000+ culverts

WDFW / The Seattle Times

The Seattle Times (Mar. 10, 2024): WDFW 'has identified more than 17,000 fish-blocking culverts across the state, the vast majority of them not owned by WSDOT.' Statewide fish-passage barriers · as of 2025

across every owner — counties, cities, private landowners, the state — and that is what has been found, not a ceiling. Zoom again to one peninsula: the Wild Salmon Center counts 4,000+ barriers

Wild Salmon Center

"There are over 4,000 documented barriers in coast rivers alone..." — Wild Salmon Center, Cold Water Connection (Coldwater Connection Campaign background). Cold Water Connection · as of 2025

on the Olympic Peninsula alone. Keep going and the scale ends at a single thing: one road culvert, on one creek, blocking one run.

The plus signs are honest on purpose. The statewide and Olympic figures are floors — what agencies and survey crews have actually identified and written down, not a census of every pipe under every road, so the true totals are larger. The injunction count is different in kind: a court has fixed it. We show you which is which rather than rounding them into a single confident wall.

Follow a single fish up a single creek on the Oregon North Coast and you reach a road culvert — a corrugated pipe set too high, running too fast, a wall a salmon cannot climb. Pull it out, drop in a bridge, and the creek is whole again. Do that ninety-three times across six rivers and you get the Salmon SuperHwy: 137 miles (was 180)

Salmon SuperHwy

Salmon SuperHwy (93 projects, 6 rivers, Oregon North Coast). Project progress · as of 2026-01 · 137 / 180 = 76.1%; the org reports "75% complete" — show the org display value with the computed figure and a rounding note.

of blocked habitat reconnected.

Up the coast in Washington, the Wild Salmon Center runs the same play at a different scale: the Cold Water Connection, a ten-year push to reconnect 125 miles

Wild Salmon Center

Wild Salmon Center, Cold Water Connection project tracker: "125 total river miles reconnected across the Washington Coast" (10-year goal); "32 total miles of wild fish habitat opened"; "reconnect 43 remaining river miles"; "WSDOT restores fish passage to 50 additional miles of habitat." 32 + 43 + 50 = 125. Cold Water Connection · as of 2025 · 32 opened + 43 WSC-remaining + 50 WSDOT-separate = 125. The WSDOT 50 is the term the PRD never named — render it.

of the Olympic Peninsula's coastal rivers. That round number only closes once you name every hand doing the work.

Two coastal projects could be cherry-picked. So step back to the country: the U.S. Fish & Wildlife Service's National Fish Passage Program has taken out 3500+ barriers

U.S. Fish & Wildlife Service

National Fish Passage Program has "removed or bypassed more than 3,500 barriers to fish passage" and "reopened access to more than 64,000 miles of upstream habitat for fish" (cumulative since 1999). — U.S. Fish & Wildlife Service, National Fish Passage Program. National Fish Passage Program · as of 2025

— reopening tens of thousands of river miles nationwide. The pattern holds far beyond two creeks. And it shows up in the fish: most of 28

NOAA Fisheries

NOAA Fisheries, "Most Threatened and Endangered Pacific Coast Salmon Populations Increased After Listings" (Dec 4, 2025): "Most Pacific Coast salmon and steelhead listed under the Endangered Species Act have increased in abundance over the past 25 years, arresting earlier declines... examined trends in 28 population groups... some—such as Snake River fall-run Chinook, Hood Canal summer chum, and Oregon Coast coho—have increased dramatically since their listing." West Coast salmon recovery research · as of 2025

have arrested their decline over about twenty-five years. Snake River fall Chinook, Hood Canal summer chum, and Oregon Coast coho rose sharply.

The work upstream is real. So is the bill. Washington is under a 90 % habitat by 2030

U.S. District Court / 9th Circuit

WSDOT (US v. Washington / Federal court injunction page): "Additional funding is needed to open 90% of potential habitat according to the permanent federal injunction." WSDOT Gray Notebook: the "March 2013 federal injunction ... requires WSDOT to restore access to 90% of blocked habitat within the area by 2030." United States v. Washington (culvert case) · as of 2025

— a 2013 federal injunction, upheld by the 9th Circuit in 2016 and left standing by a 4–4 Supreme Court in 2018. It orders the state to reopen 90% of the salmon habitat blocked by its own road culverts in the western-Washington injunction area by 2030. The state has told the court it will miss that date.

3.8 $B

WSDOT

The Seattle Times (Times Watchdog, Mike Reicher, Mar. 10, 2024): "By the end of 2024, the department will have committed all $3.8 billion previously pegged for culverts, but that amount will only open up 80% of the habitat." Fish Passage Program / Gray Notebook · as of 2024

opened roughly the first 80% of that blocked habitat. Closing the rest is where the arithmetic turns: the state estimates 4 $B more

Seattle Times

The Seattle Times (Mar. 10, 2024): "Now the state needs $3.5 billion to $4 billion more than the Legislature previously allocated." And: "To reach the last 10% — replacing about 100 culverts — WSDOT estimates it could cost up to $4 billion more." Culvert case investigation · as of 2024

to reach 90% — that money buys the 80→90% band, about the last hundred injunction culverts, not "zero to the last 10% of everything." At its peak the program was spending 1 $M/day

Seattle Times

The Seattle Times (Mar. 10, 2024): "Washington is now spending about $1 million a day to meet the deadline." (Headline: "Removing WA salmon barriers surges to $1M a day, but results are murky.") Culvert case investigation · as of 2024

.

That is the cost story. Here is the state's answer to it: the injunction culverts it has actually corrected, and the stream miles reopened, as two real yearly counts.

What critics say — and what the state answers

Three criticisms recur in the reporting on this program. Each is attributed reported opinion, not Passage's own verdict, and each is shown next to WSDOT's own numbers. We do not print a critique without the rebuttal.

For over a hundred years, four dams stood across the Klamath, built between 1918 and 1962 to make power, not to pass fish. Below them, a river that was once among the great salmon producers of the West Coast. Above them, hundreds of miles of spawning water no salmon could reach. The people who fought hardest to take the dams down — the Yurok, Karuk, and Klamath Tribes — were not bystanders to this. They were the architects

American Rivers / public record

Public record of the Klamath removal — the tribes as architects, not beneficiaries. Klamath removal coalition · as of 2024

of the coalition that spent decades forcing it through.

Between 2023 and 2024 the dams came out: 4 dams

American Rivers

American Rivers, Klamath dam removal. Klamath dam removal · as of 2024-10

in all, a roughly half-billion-dollar project and the largest dam removal and salmon-restoration effort the country has ever attempted. They fell in order — Copco 2 first, then J.C. Boyle and Copco 1, and Iron Gate last, its final in-water work finished in October 2024.

What happened next was not a wall of water crashing down. It was a reservoir draining. As the impoundments emptied, the drowned historic channel surfaced for the first time in a century — then a hundred years of trapped sediment let go, and the river ran an ugly ochre before it ran clear. On the newly opened water, the surge: fish moving up ground no living salmon had ever seen.

Within about ten days of that final work, more than 6,000 Chinook

ASCE / American Rivers

ASCE Civil Engineering Source. Klamath recovery · as of 2024

moved upstream, over roughly two weeks. Salmon reached the Upper Klamath Basin — the Williamson, Sprague, and Wood rivers — for the first time in 100+ years, some 360+ river miles

Oregon Dept. of Fish & Wildlife

CalTrout, "One Year After Klamath Dam Removal, Salmon Surge into Newly Opened Habitat" (Dec 2025): salmon reached "more than 360 river miles from the ocean into the Upper Klamath Basin for the first time in over a century." ODFW (myodfw.com) corroborates radio-tagged fall Chinook detected in the Williamson River (Oct 10) and Sprague River (Oct 14), with spawning documented in the Wood, Williamson and Sprague rivers. Klamath salmon return · as of 2024

from the ocean. The counts come from sonar enhanced by AI, about 98.4 % near-bank accuracy

CalTrout / CDFW

CalTrout, Klamath dam-removal monitoring update (Dec 2025): the Fisheye Project's machine-learning system analyzing sonar footage achieved "98.4% accuracy on the near bank (within 16 meters)." Klamath monitoring · as of 2025

, alongside PIT tags, video weirs, and spawner surveys. Precise, but not a census: these are estimates, and they carry error.

The 2025 fall run came in at 51,400 adults

CDFW / PFMC

National Fisherman (reporting CDFW, Feb 25 2026 Salmon Information Meeting): ~51,400 Klamath fall Chinook returned, ~180% of the 28,600 forecast. Klamath fall Chinook run estimate · as of 2025

. That single number can be told two honest ways, and the temptation is to pick the flattering one. We won't. Flip between them:

The 2025 run, told two honest ways

WLD
UNL

~180% of projectionvs. pre-season forecast (~28,600)~61% of averagevs. 48-year long-term average

51,4002025 fall run

above the pre-season forecast (~28,600)

The bar is the same 51,400 adults in both views. Only the line you measure it against moves.

Both are true~180% of projection vs. pre-season forecast (~28,600) · ~61% of average vs. 48-year long-term average

Source: CDFW / PFMC — National Fisherman (reporting CDFW, Feb 25 2026 Salmon Information Meeting): ~51,400 Klamath fall Chinook returned, ~180% of the 28,600 forecast. · as of 2025. The run is one number; the frame is a choice.

And in 2,026

Pacific Fishery Management Council

PFMC News Release, "Pacific Fishery Management Council Adopts 2026 West Coast Ocean Salmon Seasons" (April 12, 2026): increased Sacramento and Klamath River fall Chinook forecasts "allowed for broader fishing opportunities relative to prior seasons, including the first troll fishery openings off the coast of California since 2022"; the release also cites "the first commercial opening since 2022." 2026 ocean salmon fishery · as of 2026 · scope: California — NOT framed as the payoff to Washington's fleet collapse.

, California reopened its commercial ocean salmon fishery after three straight closed years. Read that carefully: this is California's fishery, off California's coast — not the payoff to Washington's collapsed fleet from the start of this story. Recovery is regional, and it does not settle old debts.

Scene 5 ends on a held breath: the Klamath's dams are gone, but does removal actually last? The Elwha already found out. Two dams came down between 2 dams ✎

NPS Olympic National Park

NPS Olympic National Park, Elwha ecosystem restoration. Elwha River restoration · as of 2014

— the Elwha and Glines Canyon, on the Olympic Peninsula. Forget the superlative you may have read: this was the largest until the 2024 Klamath ✎

NPS / USGS

NPS Olympic National Park; Klamath completed 2024. Elwha River restoration · as of 2024

, so we are done ranking it by size. Rank it by time. It has been dam-free for a decade, which makes it the only U.S. megaproject with real long-term outcome data — the river that answers the question the Klamath can't yet.

The decade-later scorecard

A decade on, the honest answer is species by species, not a single verdict. Some runs rebounded. One re-emerged and stayed fragile. The flagship lags. And for four species it is simply too soon: they have reached or are passing the old dam sites, but nothing yet counts as recovered.

The rebounds are real. Winter steelhead climbed to 2,519 adults (was 175)

Pess et al. 2024, Frontiers

Pess et al. 2024, Frontiers in Ecology and Evolution (CC BY 4.0). Initial responses of Chinook and steelhead to Elwha dam removal · as of 2022

, near its goal and largely natural-origin. Bull trout ran up 2–4×

Pess et al. 2024, Frontiers / USGS

Pess et al. (2021), Front. Ecol. Evol.: 'The total number of Bull Trout observed increased by 241% over the time period during which we conducted our snorkeling surveys'; counts rose from 117 (2007) and 86 (2008) before removal to 264 (2018) and 399 (2019) after — a ~2–4× increase. Brenkman et al. (2019), N. Am. J. Fish. Manage. 39:560–573: 'Bull Trout migrated between the river and its estuary (up to 168 km).' Elwha bull trout response · as of 2023

and reached the headwaters among the first fish back. Summer steelhead, nearly lost, 74–318 adults (peak)

Pess et al. 2024, Frontiers

Pess et al. (2024), Front. Ecol. Evol., Table 5 — adult summer steelhead observed on snorkel surveys of the Middle/Upper Elwha: 74 (2017), 216 (2018), 318 (2019), 92 (2020); text: 'the number of adult summer steelhead... ranged between 74 and 318 between 2017 and 2020,' with <1% hatchery-origin. First re-observed in 2013 (1 adult) and 2016 (6 adults). Elwha steelhead response · as of 2022

— back, but low and variable.

The Chinook reveal

Then there is Chinook, and it is the whole thesis in one chart. Watch the juvenile line first. Production rocketed to 324,000 juveniles/yr (was 44,000)

Pess et al. 2024, Frontiers

Pess et al. (2024), Front. Ecol. Evol.: 'The number of natural-origin subyearling Chinook salmon from the Elwha River averaged 43,828 (± 47,932), 46,973 (± 39,798), and 323,764 (± 407,976), before, during, and after dam removal, respectively... over 500,000 and almost 1 million subyearlings were produced in 2019 and 2020.' Elwha Chinook juvenile production · as of 2020

, peaking at over half a million subyearlings in a single year. It reads as a triumph.

Now the adult line overlays it, and stays flat. Adult returns sit at 4,000 adults (goal 10,000)

Pess et al. 2024, Frontiers

Pess et al. (2024), Front. Ecol. Evol. 12:1241028, 'Initial responses of Chinook salmon (Oncorhynchus tshawytscha) and steelhead (Oncorhynchus mykiss) to removal of two dams on the Elwha River': 'During (2011–2014) and following (2015–2020) dam removal, SONAR data showed average annual returns of 3,444 (± 1,125) and 4,734 (± 2,409) Chinook salmon respectively.' Naturally spawning adults before/during/after dam removal averaged '1,393 (± 1,218), 1,930 (± 747), and 3,523 fish (± 1,949), respectively.' Historical run ~10,000–30,000 adults is attributed to a separate source (Dept. of Interior 1996; cited in Front. Ecol. Evol. 2023, DOI 10.3389/fevo.2023.1240987), not the pinned paper. Phase framework (Preservation → Recolonization → Local Adaptation → Viable Natural Population) attributed to Front. Environ. Sci. 2024, DOI 10.3389/fenvs.2024.1291265. Elwha Chinook adult returns · as of 2022

, measured against a recovery goal of more than ten thousand, still in the Preservation phase, one of four. The resolution is the uncomfortable part — 96 % hatchery-origin

Pess et al. 2024, Frontiers

Pess et al. 2024 (median 95.9%, 2009–2020, CC BY 4.0). Elwha Chinook responses · as of 2020

, and wild reproduction runs below replacement. The fish are being made, in a hatchery, not made anew in the river. A real biological surge that has not become self-sustaining recovery.

177 fish

The Lower Elwha Klallam fought hardest for removal, and then paid for it first: they voluntarily stopped fishing their own river for roughly twelve years, from 2011 through 2023. A generation grew up never fishing it. The first fishery back was ceremonial, coho only, the lower three miles: 177 coho

NPS Olympic National Park

NPS Olympic National Park; Northwest Treaty Tribes. Elwha tribal fishery · as of 2023-10

. Symbolically enormous, materially tiny — which is the true pace of recovery, in the tribe's own framing, not a stock-photo comeback.

We came the whole way upstream: past the ocean's stakes, the wall of barriers, the culvert bill that keeps climbing, the Klamath running free, and the Elwha's decade of receipts. This is where it settles. A returning salmon reaches the gravel, spawns, and dies — and that death is not the end of the accounting: carcasses feed the watershed

USGS / peer-reviewed ecology

"Pacific salmon transport large amounts of marine nutrients to freshwater and forest ecosystems when they migrate from the ocean, spawn, and die." — U.S. Geological Survey Marine-derived nutrients · as of 2025

. A spawned-out carcass is a delivery of ocean nutrients into the headwaters that raised it; the forest and the next generation of smolts are fed by fish that made it home.

Then the wait. The eggs in that redd won't return as adults for two to five years, into an ocean we can't predict. So the honest way to close this is not a triumph reel. It is to stop moving and show you the live count, with everything that complicates it left in.

The count you can read live — and what it will not tell you

Every adult that climbs the ladder at Bonneville is counted. Here is the running season-to-date total, straight from DART, updated as the Corps posts it. It is the one place in this story you can drive: a real number, refreshed, with its provenance attached. Read it as a calm instrument, not a scoreboard.

The hatchery-vs-wild reckoning, in plain words

A large share of the fish in that Bonneville count are hatchery-origin. The ladder counts every body that passes; it does not ask where the fish was born. That matters more than any single number, because a rising total can mask a falling wild run. When you see the count climb, you are seeing passage — not proof that self-sustaining wild populations are recovering. DART does not split the headline Chinook count into wild and hatchery, so this project won't pretend to. We show the total, name what it contains, and stop there.

Recovering, and not

Across the listed populations the picture refuses to resolve into one word; in NOAA's own status reviews it is recovering & not

NOAA Fisheries

"Under the Endangered Species Act, NOAA Fisheries periodically reviews the status of each of 28 West Coast salmon and steelhead species listed as threatened or endangered." — NOAA Fisheries, "Report Card on Recovery: Reviews Assess 28 Salmon and Steelhead Species Returning to West Coast Rivers" (page last updated 12/11/2024) 5-year status reviews · as of 2025

. Some are climbing; some are still declining; none has yet recovered enough to come off the endangered-species list. Both halves are true at once, and the tension is the point. Recovery is real and partial. Progress on one river does not average out loss on another.

The climate overhang

Beneath every count runs the ocean itself: the shadow over all of it

NOAA Physical Sciences Laboratory

NOAA PSL, PDO index (1854–present, monthly CSV). Pacific Decadal Oscillation index · as of 2026

. Ocean regime shifts and warming change how many of those eggs ever come back, and they do it on a clock no restoration budget controls. We plot this as honest correlation, never causation: the PDO index moves alongside returns, but it does not, by itself, explain them. When a run comes back thin, we can rarely say how much was the river we fixed and how much was the ocean we can't. That uncertainty is not a hole in the story. It is the most honest thing in it.

Here's who's actually doing this

No donation wall, no email capture, no ranking. These are the organizations and sovereign tribal nations doing the restoration work behind the numbers in this story. Listed without order of importance. If any of it moved you, go read theirs.