After the Elwha Dam blockage was out of the way, starting in April 2012, biologists were able to confirm the movement of anadromous salmon and trout by looking for fish and taking count of their nests, or redds, within the gravel.
Pink salmon were first seen above the former Elwha Dam site in September 2012, following the arrival of winter steelhead the previous June and Chinook in August, all on their expected migration schedules. Coho soon followed in November, also right on time, but it would be another two years before chum redds were seen upstream from where the lower dam had stood.
Beyond the upper dam, coho and pink salmon were not so quick to recolonize. After the Glines Canyon Dam came down in August 2014, a rockfall that October created an obstruction to fish in the river. Crews blasted out the rockfall a year later, but it would be one more year — October 2016 — before coho and their redds were seen upstream.
“It was exciting to see coho above Glines Canyon Dam,” said Heidi Connor, fisheries technician for Olympic National Park who was involved in the early surveys. “We actually saw the fish in Boulder Creek (about 1.5 miles above the dam site) on Halloween in 2016.”
Although visibility was low, Connor spotted something moving in the creek. Stepping closer, she spotted two adult coho. It was the first actual sighting of coho above the upper dam.
“It seems they had just arrived there, because their spawning coloration had not changed,” she said. “That was pretty cool to see.”
The summer and fall of 2016 became the beginning of an upstream push among anadromous fish, including summer steelhead, Chinook and coho, Conor recalled. “The door was open, and they were coming,” she said.

Timeline showing the dates of dam removal and first observations of adult migratory fish upstream of Elwha Dam (orange lines) and Glines Canyon Dam (blue lines). Observations were compiled from ongoing field studies indicated by superscripts: (1) snorkeling surveys; (2) redd surveys; (3) smolt trapping; (4) tangle netting; and (5) radio-telemetry. Graphic: Duda et al. 2021 with permission
Pink salmon weren’t spotted until a snorkeling survey in 2019. And, so far, the only sign of chum in the upper watershed was trace amounts of DNA found in water samples during environmental DNA (eDNA) analysis. Pink and chum salmon are known to spawn mostly in the lower river, scientists say, so a slow recolonization farther upstream was not unexpected.

Pink salmon. Photo: John McMillan NOAA/NWFSC (CC BY-NC-ND 2.0)
Both pink and chum populations remain critically low in the Elwha River, as they were before dam removal. The demolition project, which dislodged large amounts of sediment and deposited them downstream, seemed to create more problems for the depressed populations, causing pink and chum numbers to sink even lower. Knowing that both species were abundant at one time, restoration managers remain hopeful that the populations will rebound as the fish take advantage of newly accessible habitat.
Limited genetic information is available for current populations of pink and chum salmon in the Elwha. Before dam removal, an early and a late run of pinks were identified in the Elwha and compared to similar populations in the nearby Dungeness River. Elwha and Dungeness stocks were found to be “genetically different but connected through historical gene flow or recent common ancestry,” according to a report by Maureen Small and fellow geneticists at the Washington Department of Fish and Wildlife.
Two runs of fall chum have been identified in the Elwha, one returning in October-November and one in December, according to McHenry. The earlier run has been associated with the native Elwha chum. The later run has been associated with a large number of hatchery fish imported from Hood Canal from 1975 to 1985. Since dam removal, observers are noticing what could be the beginning of an even earlier chum run, described as “summer chum,” he said.

Chum salmon. Photo: Roger Tabor/USFWS (CC BY-NC 2.0)
Fall chum sampled in 2016 showed interesting genetic connections to chum runs in Canada’s Georgia Strait, to those on Washington’s coast, and to those in Hood Canal, as well as other runs in the Strait of Juan de Fuca, according to a new report by Small and her collaborators. The findings reveal an ongoing genetic influence of past hatchery imports, effects of possible straying from other streams, and genetic changes brought about by the mixing of different stocks as they adapt to new conditions after dam removal.
As for Elwha coho, it appears that most of these fish spend their first year in the watershed, sometimes hanging out in marshes, side channels and slow-moving tributaries where aquatic insects abound. Other coho begin their migration to sea soon after they emerge from the gravel.
At the time of dam removal, more than 90 percent of the coho returning to the Elwha were survivors reared in downstream hatcheries before their release. A few were the offspring of wild or hatchery fish that spawned in the limited habitat below the Elwha Dam. Aware of the vast habitat above the dams, longtime observers were hoping for a large resurgence of naturally spawned coho once the dams were removed. Those hopes were largely fulfilled, thanks in part to human hands that transported more than 3,000 coho to upstream waters, jump-starting the upstream migration.
Under an approved recovery plan, biologists moved returning adult coho from downstream hatcheries to five tributaries that flow into the Elwha River above the lower dam. This coho-relocation effort started in 2011, while demolition work was still underway on the Elwha Dam. The streams provided a safe haven from sediments unleashed by the removal of both dams.
For the first few years, the relocation focused on the two largest tributaries above the Elwha Dam, Little River and Indian Creek. Success of the relocations could be seen immediately in the number of young coho leaving the two streams and heading downstream, some going out to sea. But a more telling sign came in 2016, when more than half the adult fish were seen to be missing a Floy tag — a small plastic tube attached beneath their dorsal fins to show that they were transplanted fish. The number of adults showing up without a tag provided strong evidence that these were offspring of the transplanted fish. They were ready to produce their own offspring, further expanding the coho population in upstream waters.
The relocation effort in Little River and Indian Creek came with an added bonus: an experimental opportunity to measure how coho respond to entirely different habitat conditions. Although the two streams enter the Elwha at nearly the same location — Little River flowing from the east and Indian Creek from the west — they are very different streams. Little River flows down from the Olympic Mountains through a steep, mountainous channel, while Indian Creek meanders gently through the lowlands from Lake Sutherland and nearby wetlands. Little River is a cold stream with three waterfalls that impede or block the movement of coho. Indian Creek is a relatively warm stream due to high temperatures in Lake Sutherland, and the fish have access to extensive spawning and rearing habitat. It’s a tale of two tributaries.
The study, which used multiple techniques to trap, count and measure young fish, found that the warmer Indian Creek produced fish that emerged from the gravel earlier and grew faster. As juvenile coho prepare to migrate to sea, they begin to lose their stripes (parr marks) and grow silvery.
During 10 years of study, the number of smolts coming out of Indian Creek each year ranged from 5,000 to 35,000, while Little River produced between 500 and 1,500 during the same time period. For a given mile of stream habitat, Indian Creek produced more than four times as many smolts, on average, as Little River.
On the other hand, parr migrants — those still wearing their camouflage stripes — have been seen leaving Little Creek in far greater numbers than for Indian Creek. Although their destinations remain unknown, some of those less-mature juveniles probably find habitat where they can grow for another year. Others head on out of the Elwha River and into the brackish waters of the estuary.

Juvenile coho salmon in a small spring-fed channel in the lower reach of the Elwha River. Photo: Roger Tabor/USFWS (CC BY-NC 2.0)
Why coho parr leave Little River in large numbers may be related to the colder waters that delay their growth, said Mike McHenry, habitat manager for the Lower Elwha Klallam Tribe. Maybe they are looking for slightly warmer conditions. The out-migration may also be the result of degraded habitat left over from historic logging practices, he said. Thanks to recent habitat restoration, researchers may soon learn whether the parr in Little River are influenced solely by temperature or if they will stay in the stream when provided with better living conditions, such as low-velocity side channels and pools.
In 2018, the relocation effort was discontinued as native spawners continued to outnumber the relocated fish and the production of native juveniles continued to rise. By 2019, the number of spawners was estimated at 3,140, rising to 5,250 in 2020 and 6,200 in 2021 (preliminary estimate). The number of subyearling coho captured in the fish traps — one measure of productivity — multiplied rapidly over the past five years: 1,567 in 2016, 2,570 in 2017, 12,216 in 2018, 16,180 in 2019, 20,470 in 2020 and 55,630 in 2021.
Since dam removal, Lake Sutherland has been shown to be a highly productive rearing habitat for coho. In 2017, a lakefront property owner filmed six coho spawning in Falls Creek, a small stream that flows into the lake. Three were transplants, as shown by their Floy tags, and three made it on their own, as shown by the absence of tags. Other coho may have spawned in the lake itself.
The result has been not only more coho leaving Lake Sutherland and other habitat along Indian Creek but also larger ones, as shown by captures in a smolt trap installed at the lake’s outlet in 2018. This past year, the Lower Elwha Klallam Tribe estimated that 234 coho smolts had left the lake, some as large as 12 inches long. That is more than twice the size of typical coho in less productive systems.
Studies now underway include an examination of the genetic makeup of the Elwha coho population and its relationship to other coho populations in the region. Based on genetic samples taken in 2013, Elwha coho appear to be different enough to tell them apart from other coho populations in the region, according to Todd Seamons, director of the Molecular Genetics Laboratory at WDFW. More work is needed to determine the precise kinship to other coho stocks.
In studies conducted before dam removal, researchers produced conflicting results about whether Elwha coho were more closely related to Puget Sound stocks or to Washington’s coastal stocks. The geneticists agreed that Elwha coho salmon, like Elwha pink salmon, were most closely related to those from the nearby Dungeness River. The pre-dam-removal analyses were conducted at a time when nearly all the Elwha coho were from hatcheries. Now, with the reopening of extensive coho habitat upstream, conditions are dramatically different. As wild coho evolve in the watershed, their genetic makeup may continue to change.
The primary goal of Seamons’ study — conducted in collaboration with researchers from five agencies in the U.S. and Canada — is to produce a “reference baseline” of genetic signatures for a vast array of coho populations in the Northwest. This upcoming baseline is funded by the Pacific Salmon Commission, which coordinates fishing treaties between the U.S. and Canada. Among other things, the baseline can be used to help identify the stream of origin for coho harvested in various sport and commercial fisheries.