Are Salmon Endangered? The Status of Wild Populations

Wild salmon populations range from thriving to critically imperiled depending on the species, the river system, and the region of the world. There is no single answer to whether “salmon” are endangered because the word covers multiple species across two oceans, and individual populations within each species can be in radically different shape from one watershed to the next. Some runs have been formally listed under the Endangered Species Act in the United States; others in Alaska and parts of Russia remain large enough to support commercial fisheries. The picture is less about a single species in crisis and more about a patchwork of local declines driven by an overlapping set of pressures that keep getting harder to untangle.

Not One Species, Not One Answer

When people ask whether salmon are endangered, they’re usually thinking of one kind of fish. In reality, “salmon” encompasses at least seven Pacific species (Chinook, coho, sockeye, pink, chum, steelhead, and masu) and one Atlantic species. Their statuses vary enormously. A global assessment of sockeye salmon found the species as a whole is not at risk of extinction, but among 62 populations with enough data for analysis, roughly a quarter were at risk, and five populations were already gone.¹ That pattern repeats across species: the species persists globally while individual populations wink out.

In the Pacific Northwest, dozens of distinct population segments of Chinook, coho, sockeye, chum, and steelhead are listed as either threatened or endangered under the U.S. Endangered Species Act. Atlantic salmon on the North American side have suffered even more dramatic losses; most U.S. populations in New England are federally endangered, and Canadian populations in the inner Bay of Fundy are critically endangered. Meanwhile, Alaskan pink salmon returns recently hit record highs in odd-numbered years. The contrast is stark enough to make any blanket statement about salmon misleading.

Dams and Blocked Migration

Salmon are anadromous, meaning they hatch in freshwater, grow in the ocean, and return to freshwater to spawn. Anything that breaks that loop threatens an entire population. Hydropower dams have been the most obvious barrier for over a century. Even dams outfitted with fish ladders and bypass channels can cause significant mortality and delay. In one study of Atlantic salmon smolts migrating past a hydroelectric dam, only about 65% of the tagged fish present in the area in front of the dam managed to pass it, despite multiple routes being available.² The delay itself compounds the problem: smolts that spend extra time milling around a dam are burning energy reserves they need for ocean entry and becoming easy targets for predators.

Multiply that effect across four, six, or eight dams on a single river system, and cumulative losses can gut a run. The Columbia River basin, where the largest Chinook and sockeye runs in the lower 48 states once thrived, has more than 400 dams. Some tributaries are entirely cut off. Where dams have been removed, however, the results have been encouraging. On the Elwha River in Washington, model simulations after dam removal suggested that the resumption of anadromy by landlocked kokanee (a freshwater form of sockeye) could rapidly boost smolt production, because returning adult sockeye produce roughly seven times as many eggs per female as kokanee.³ Recovery, where the habitat allows it, can begin surprisingly fast once the physical obstruction is gone.

Warming Waters and Changing Oceans

Climate change acts on salmon at almost every stage of their lives. In freshwater, warmer river temperatures during the spawning migration can be lethal; adult Chinook salmon in the subarctic Yukon River have shown declining juvenile production in recent decades as mainstem water temperatures rise and discharge drops.⁴ In the ocean, warmer sea surface temperatures during the coastal phase of juvenile sockeye migration have been linked to lower recruitment two years later, possibly because warmer conditions favor predators or reduce prey availability at a critical time.⁵

Across North American Atlantic salmon populations, researchers found that the widespread declines of the early 1990s were too coherent across regions to be explained by local river-level factors. The pattern pointed toward a marine-phase shift driven by poor feeding conditions and warmer ocean temperatures across their entire habitat area.⁶ When the ocean itself becomes less hospitable, no amount of freshwater restoration can fully compensate.

Ocean acidification adds another dimension. Experiments on juvenile chum salmon exposed to elevated carbon dioxide levels found a three-fold increase in mortality over 25 days, driven by impaired ion regulation that food availability could not offset. Smaller fish were hit hardest.⁷ This is laboratory work at CO₂ levels higher than current averages, but it flags a vulnerability that could matter more as ocean chemistry continues to shift.

Sea Lice and Disease From Aquaculture

Open-net salmon farms, concentrated in Norway, Scotland, Chile, and western Canada, sit directly in the migration corridors wild salmon use. Two farm-related pressures stand out: sea lice and disease spillover.

Sea lice are naturally occurring parasites, but aquaculture creates unnaturally dense hosts for them. In Norway, a comprehensive analysis found that wild Atlantic salmon catches declined significantly as concentrations of adult female sea lice in nearby farming areas increased. When farms exceeded a threshold of 0.1 adult female lice per farmed fish, the risk of below-average wild catches jumped by roughly 47%.⁸ A separate study of rivers categorized by lice exposure found that high-lice years reduced adult returns by an average of about a third compared to low-lice years.⁹

Disease transmission is the less visible but potentially larger concern. Piscine orthoreovirus (PRV-1) was traced through phylogeographic analysis back to European Atlantic salmon aquaculture, which facilitated its spread to both the North and South Pacific.¹⁰ In British Columbia, wild salmon sampled close to farms tested PRV-positive at rates of 37 to 50%, compared to just 5% in regions farthest from farms.¹¹ And escaped farmed fish are swimming disease vectors: testing of 567 farmed Atlantic salmon escapees captured in a Norwegian river over five years found that more than 90% carried at least one virus, with PRV-1 the most common at about 76%.¹²

Urban Stormwater and a Chemical From Tires

One of the more surprising recent discoveries in salmon conservation involves car tires. Coho salmon returning to spawn in urban creeks of the Pacific Northwest were dying suddenly for decades before anyone identified the cause. Researchers eventually traced the mass die-offs to 6PPD-quinone, a transformation product of a chemical used as an antioxidant in tire rubber. Roadway runoff carried it into streams at concentrations well above its lethal threshold, with the median lethal concentration for coho coming in around 0.8 micrograms per liter, a level frequently exceeded in stormwater-affected creeks.¹³

The threat extends to earlier life stages as well. Coho embryos survived exposure at concentrations that kill juveniles and adults, but growth was inhibited, and newly hatched fish proved vulnerable to lethal effects immediately after emerging from the egg. This creates a large window of sensitivity from hatch through smoltification.¹⁴ The chemical is globally ubiquitous in tire rubber, which means any creek receiving urban runoff is potentially affected. Bioretention swales and rain gardens can filter some of it out, but retrofitting entire urban watersheds is a generational project.

Marine Mammals and Shifting Predation Pressure

Pinniped populations in the Pacific Northwest have rebounded strongly since protections under the Marine Mammal Protection Act took effect in the 1970s. That recovery has created a growing predation burden on already-stressed salmon populations. In the lower Columbia River, increasing sea lion abundance was linked to an additional 21% mortality for the earliest-migrating Chinook populations during high sea lion years, compared to baseline years. Even the latest-migrating populations experienced roughly 10% additional mortality.¹⁵ A study in an Oregon estuary found that harbor seals consumed an estimated 21% of the prespawning coho salmon population during a single fall period, with a small fraction of specialist seals doing most of the damage.¹⁶

This sets up a conservation conflict: marine mammal protection and salmon recovery are both legal mandates, and at certain bottleneck sites they work against each other. Lethal removal of individual sea lions at specific dams has been authorized in some cases, but it remains politically contentious and biologically limited in scope.

Competition at Sea With Pink Salmon

Not all threats come from outside the salmonid family. Pink salmon in Puget Sound have reached record-high odd-year abundances, and this boom appears to hurt Chinook salmon. Since 1997, Puget Sound Chinook abundance has averaged about 34% less in odd-numbered years (when pinks are present in force) than in even-numbered years. The pattern has rippled further up the food chain: Southern Resident killer whales, which depend heavily on Chinook as prey, have shown biennial patterns in mortality and births that track the Chinook-pink cycle.¹⁷ The irony is that pink salmon are themselves native and ecologically important. The problem isn’t that pinks exist but that hatchery-supplemented pink runs have reached scales that shift competition dynamics against already-depleted Chinook stocks.

The Hatchery Dilemma

Hatcheries were originally built to offset the salmon losses caused by dams and overharvest. They can boost raw numbers, but decades of evidence show they also create genetic and ecological risks for wild fish. The core tension is straightforward: fish bred in captivity adapt to captivity, and when they or their offspring interbreed with wild fish, the resulting population can lose fitness traits honed for natural conditions.

Modeling of Alaska pink salmon found that hatchery strays on wild spawning grounds could bolster total population size toward carrying capacity but reduced the variation in adult return timing by up to 20%, effectively narrowing a population’s life-history portfolio.¹⁸ That diversity in timing is exactly what buffers populations against unpredictable conditions. Simulations comparing different hatchery management approaches found that “integrated” programs, which regularly bring wild-origin fish into the broodstock, posed less genetic risk to wild populations than “segregated” programs that try to keep hatchery and wild lines completely separate, largely because real-world implementation is never perfect and some mixing always occurs.¹⁹

The upshot: hatcheries aren’t going away, but the way they’re managed matters enormously. A well-designed program with genuine integration of wild genetics and strict limits on straying can coexist with wild recovery goals. A poorly managed one can undermine the very populations it was supposed to help.

Mixed-Stock Fisheries and the Harvest Problem

Commercial and recreational harvest adds another layer of complexity because salmon from different populations often migrate and feed in the same waters. Highly mixed-stock fisheries are a problem when the aggregate includes both stocks large enough for commercial harvest and stocks of critical conservation concern. Harvest rates set for the aggregate may be sustainable for the abundant stocks while simultaneously depleting less productive ones toward local extinction.²⁰ In the Fraser River system, for instance, management groups salmon into four run-timing groups, meaning that harvest effort can be dialed back for a timing group but not for specific stocks within it. Abundant and endangered stocks sharing the same timing window get caught together.

Improving stock selectivity, through terminal fisheries closer to spawning grounds, genetic stock identification, or time-area closures, can help. But each approach involves trade-offs in cost, efficiency, and the livelihoods of fishing communities, so progress has been incremental.

What Wild Salmon Feed Beyond the River

The conservation stakes extend well beyond the fish themselves. Salmon function as nutrient conveyor belts, hauling marine-derived nitrogen and phosphorus into freshwater and terrestrial ecosystems when they return to spawn and die. In one Alaskan stream, spawning salmon increased water nitrogen concentrations by an average of 190% and phosphorus by 390% compared to pre-salmon levels, and primary producers incorporated those nutrients into their tissues.²¹ The pulse of nutrients extends well past the streambank. On the coast of British Columbia, bears and scavengers drag carcasses into the forest, depositing nitrogen that shows up in the foliage of mosses, huckleberry, salal, and western hemlock trees growing within ten meters of spawning channels.²²

At the other end of the food web, Southern Resident killer whales depend on Chinook salmon as their primary prey year-round.²³ Estimates suggest this small endangered orca population would need roughly 240,000 Chinook annually to meet its energetic requirements if eating nothing else.²⁴ The fate of two endangered species is yoked together by a single dietary link.

Why Life-History Diversity Is the Real Insurance

Researchers studying Bristol Bay sockeye salmon in Alaska found that populations with diverse life-history strategies, meaning variation in age at migration, run timing, and spawning habitat, were more resilient to unfavorable conditions like habitat loss and climate variability.²⁵ The analogy often used is a financial portfolio: a diversified one weathers downturns that would ruin a concentrated bet. Bristol Bay’s sockeye remain robust partly because the system still supports hundreds of distinct spawning populations, each adapted to slightly different conditions. When one run fails, others pick up the slack.

This insight reframes the conservation challenge. Saving “salmon” in the abstract isn’t the goal. Preserving the full range of local adaptations, run timings, and life-history variants is what keeps the species capable of absorbing shocks. Every time a small, genetically distinct population blinks out, the overall portfolio shrinks, even if total numbers hold steady for a while. That erosion of diversity doesn’t show up on a harvest chart until conditions change and the buffering capacity isn’t there anymore.