As the last block of concrete was pulled from the riverbed, the Elwha River in the Olympic Mountains of Washington State flowed freely for the first time in over 100 years. The river was historically one of the most productive salmon streams for its size in the Pacific Northwest. Four hundred thousand salmon once swam its length each year but, in the century since the dam’s construction, that number had fallen to a few thousand.1 Within months of the dam’s removal, nature has rushed back: over 200 salmon have already returned. The prospect of a river teeming with silverbacked salmon weighing over 45 kilograms each may no longer remain a hazy memory of local Native American tribes.
The Elwha dam removal project stands as one of the first large dams ever removed. The intent of removing the dams is to fully restore the Elwha River ecosystem and its native migratory fish species. In doing so, the Elwha dam project revived the debate of how to balance the conflicting demands of humans for both clean energy and healthy ecosystems. Previously, that debate has been weighted decisively in favor of dam projects. But with a greater understanding of the value of ecosystem services, the Elwha dam project may represent the start of a revolution in how we assess the West’s aging dam infrastructure.2
The Elwha watershed was the traditional homeland of the S’Klallam Tribe, whose culture flourished on salmon from the river, among other natural resources. Against tribal will, construction of the Elwha Dam began in 1910 for the sole purpose of generating the first electricity in the region. The electricity powered several lumber mills and fueled economic development, resulting in construction of a second dam, the Glines Canyon Dam farther upstream, in 1927. The lower Elwha Dam did not have fish passage and the salmon runs declined from 400,000 per year to about 3,000 fish in the lowest eight kilometers of the river. Tributaries in the headwaters of the Elwha River were protected from further development in 1938 with the establishment of Olympic National Park. The impact on the S’Klallam Tribe was devastating for their culture and livelihood. A fishery that could be worth over $10 million was lost. The near disappearance of salmon in the watershed also had a cascading effect on the terrestrial ecosystem, where some 22 species of resident wildlife were affected, and over 90 species of migratory birds. The decomposing salmon carcasses have been shown to significantly contribute to the biomass of the forest itself, accounting for 20–60 percent of riparian biomass.1
The dams also stopped the movement of sediment through the river system, resulting in deposition in the reservoir deltas.1 As a result, the river incised its channel, armored its bed with boulders (instead of sand and gravel where salmon could lay eggs), and reduced the sediment delivery to the coastal environment, causing beach erosion for at least 30 kilometers along the shore.3 Some of the most intense erosion occurred on Ediz Hook, which creates an important lumber shipping port at Port Angeles. From the 1970s through the 2000s, the Army Corps of Engineers spent hundreds of millions of dollars each year to protect the Ediz Hook from erosion.
In 1968 the S’Klallam Tribe and numerous environmental groups tried to stage a comeback by opposing relicensing of the dams by the federal government, citing loss of the salmon fishery, negative environmental impacts within the watershed, and submersion of a tribal sacred site under the reservoir.
But despite the strong case against the dam, the local nontribal community strongly favored relicensing it. The electricity from the dams continued to play an important role in powering the region’s timber-based economy. There were additional challenges in assessing the possible impacts of removing the dam, given the amount of sediment that had built up, and on the potential impact on the City of Port Angeles’ water supply. Even a U.S. Department of the Interior study in 1991, which recommended the removal of the dams, failed to energize foot-dragging state officials and local interest groups. The removal was legislated by Congress with the passage of the Elwha River Ecosystem and Fisheries Restoration Act P.L. 102-495.
However, in the nearly two decades that followed the passage of the 1992 federal law mandating the dams’ removal and the release of funds to begin work, a generational shift took place. Dams—for so long seen as symbols of development and progress—were increasingly being criticized for their social and environmental impacts. The issue came into focus internationally following protests over the forced evictions of hundreds of thousands of people in India and China due to dam projects. In the United States, aging dam infrastructure was pushing local governments to repair or remove many of these structures. The 85,000 large dams in the United States have an average age of 53 years, and over 4,000 of the large dams are considered structurally unsound.4 An additional problem with dams is that, as they age, they fill with sediment, reducing storage capacity.
For these reasons, hundreds of small dams (less than 7.5 meters in height) have been removed in recent decades. With smaller structures, there is little question that rivers can return to their pre-dam flow characteristics.5,6 Successful dam removals and ecosystem recovery have been seen with the Edwards Dam in Maine, and the Marmot and Condit Dams in Oregon. But as the Elwha dam saga continued to rumble into the new millennium, the question remained whether dam removal would be successful for large structures.
A major concern when removing a dam is managing the remobilized sediments from the lake delta, which are now exposed to flowing water. The Elwha River dams have accumulated over 34 million cubic meters of sediment. The reservoirs of the Glines Canyon and the Elwha Dams had no drains and were too large for a single, explosive removal. Each dam had unique characteristics and required its own removal plans, time frames for safety concerns, and strategies for managing the massive amounts of sediment in the reservoirs. Careless removal of the dams could cause large amounts of sediment in the reservoir deltas to flow down the river.7 Even though the Elwha Dam is the older of the two dams, the majority of sediment lay trapped upstream behind Glines Canyon Dam (GCD). Thus the GCD removal progressed slowly in order to manage the sediment through the river system. Further complicating matters, sediment behind the GCD was located in a federally designated wilderness area of Olympic National Park, where machinery is banned.
In September 2011 work began on removing both dams. The Elwha Dam was structurally unsound, due to poor construction, and required a complex removal process to avoid a catastrophic failure. First, a cofferdam moved the river to the left side of the dam; an artificial channel cut through the bedrock where the dry right spillway is located. Then a second cofferdam directed the river into the artificial channel for removing water from behind the main dam, which was subsequently removed by jackhammering and blasting. By contrast, the Glines Canyon Dam was structurally sound, allowing for a large jackhammer to be used directly on the dam face. Both dams were removed within 13 months. However, the Elwha Dam was removed first, and muddy sediment poured down the river and into the ocean for the first time in over 100 years. Beaches near the river’s mouth experienced immediate growth, even faster than expected. After the Glines Canyon dam was fully breached, even larger amounts of sediment began moving through the river. Logs and other floating debris created logjams, causing the river to erode its banks and migrate across its floodplain as it had before the dams were installed. Sediment levels are expected to return to normal in one to three years.
Of course, removing the dam was only the first part of the first step of the real goal of restoring salmon fisheries. During the decade prior to removal, scientists surveyed fish populations in the river to inventory populations of native and migrating fish species.7 Fisheries biologists also captured Elwha River fish stock for transport to hatcheries and nearby streams for rearing in order to preserve genetic diversity. The schedule of work during the process of removing dams was periodically halted to protect fish during their seasonal runs, and provided windows for their capture and transport to safe rearing sites. Fish stocks will recover following complete deconstruction of the dams, stabilization of sediment transport, and the recovery of the ecosystem food chain that provides food for juvenile salmon that will grow in the Elwha River before migrating to the ocean. However, even in the short time (less than six months) following removal of the Elwha Dam, some wild salmon found the new habitat and spawned, in spite of turbid water. The premature appearance of the native salmon population is a positive signal forecasting the recovery of the salmon population.
The exposed reservoir lakebed represents another restoration problem. Without vegetation cover, the soft lake sediments are subject to erosion during the rainy season. For approximately 10 years, the Park Service has been collecting and saving native seeds, and rearing plants to revegetate the lakebed. As soon as lake levels were drawn down, crews were planting seeds and seedlings in order to head off invasive species. Luckily, this winter has been mild with few erosive rainstorms, and the seed bank in the sediments produced a nearly continuous cover of vegetation. The invasive species will be monitored in coming years, but the soil was more stable than expected during this first critical year.
The removal of dams on the Elwha River offers a unique opportunity to evaluate the effects of large dam removal and subsequent recovery of formerly productive aquatic ecosystems that supported large populations of salmon and a related complex ecosystem.8 Although intentional dam removal of this magnitude is unique, it could become more common as those in the United States and other nations manage an aging system of dams. An essential step in removing both small and large dams is assessing watershed scale features before and after dam removal. A comprehensive plan designed to evaluate the effects of dam removal on existing fish populations, food webs and habitats, sediment flow, and many other factors is essential before removing dams. Now, we are well positioned to see exactly how the system responds.
