A hatchery is a fish farm, and fish farming in the Columbia River Basin has been utilized since the late 1800s to provide salmon and steelhead for harvest, particularly commercial harvest, and to compensate for fish losses caused by human activities such as the destruction of fish habitat and construction of dams.

To its earliest practitioners in the basin, hatchery practices were deceptively simple. Animal husbandry techniques, which successfully produced herds of farm animals, simply were incorporated into producing artificial populations of anadromous fish. By the late 19th century, animal husbandry techniques were being implemented successfully in the livestock industry, and it seemed only logical that the same principles could be applied successfully to anadromous fish. The unique life history of salmon was not understood at the time, and over the ensuing decades it would become clear that salmon are not cows that could be released into the open range of the ocean and collected later as adults. The open-range concept didn’t work for many of the early cattle operations in the West, and it didn’t work for salmon, either.

In 1875, Spencer Baird, the United States Fish Commissioner, advised the commercial fishing industry that artificial propagation of salmon would be so successful it would eliminate the need to regulate harvest. Regulation was a controversial issue at the time, as the salmon runs were being fished heavily for economic gain but without effective regulation, and some scientists already were concerned that overfishing might prove catastrophic to the runs. In response to a request from the Oregon Legislature, Baird outlined the problems he saw for the salmon industry: 1) excessive fishing; 2) dams; and 3) altered habitat. Baird believed each of these problems could be resolved through artificial propagation of fish. That is, sufficient numbers of fish could be produced in hatcheries to satisfy the demand of commercial fishers, hatcheries could be located on tributaries of the Columbia where the fish would not have to pass dams on their way to the ocean as juveniles or back from the ocean as adults, and altered natural habitats would be of minor consequence because so many fish would be spawned artificially at the hatcheries.

It is ironic to note that this wisdom prevailed for more than 100 years, even in the face of declining returns of hatchery salmon. Fisheries scientist and author Jim Lichatowich observes that Baird saw artificial production as a better means of protecting the economically important commercial fishery than attempting to protect naturally spawning fish by regulating harvest. In other words, “protection” favored the fishery, not the fish. Hatcheries would eliminate the need to regulate the fishery. “Baird reached this conclusion just three years after the first hatchery for Pacific salmon was opened on the Sacramento River,” Lichatowich writes in a 1996 report prepared for the Bonneville Power Administration. “Ninety years later the hatcheries began making meaningful contributions to the fishery, but by then most of the original natural productivity of the Pacific salmon in the Columbia River had been destroyed.”

In 1877 Columbia River cannery operators, worried about the decline of the prized spring Chinook runs, organized the Oregon & Washington Fish Propagation Company, raised $21,000 in donations, and built a salmon hatchery on the Clackamas River. It was the first hatchery in the Columbia River Basin. The facility was operated by the U.S. Fish Commission and only partially funded by the cannery operators. Other contributors included the state of Washington. The facility operated until 1881 and then closed, apparently for lack of funding (The state of Oregon reopened it in 1888, then turned its operation over to the U.S. Fish Commission again in 1889).

In 1883 George Brown Goode, U.S. Fish Commissioner, told the World Fisheries Congress that the Columbia River salmon industry was under the complete control of fish culturists. This was surprising, in retrospect, because the only fish hatchery in the Columbia River Basin had been closed for two years. The Commission’s primary management activity was the promotion and operation of hatcheries. Goode told the delegates that Columbia River hatcheries were successful because the United States Congress continued to fund them. His twisted logic was that if the hatcheries were not successful, the government would not continue funding them. Because the facilities continued to receive funding they must be successful, therefore.

In reality, there was no critical examination of the impacts of hatcheries on wild fish. The extent of success or failure simply was not known. Few questioned the opinion of the U.S. Fish Commission, even though it was hardly unbiased on the question and even though, at the time, only two hatcheries were operating on the West Coast. It didn’t matter. Hatcheries were political tools to assuage powerful fishing interests as much as they were fish farms. Hatcheries produced fish; fish produced commercial fishing opportunities; fishing opportunities put people to work. Hatchery fish even were planted in some rivers to win political favor from elected officials.

By the 1890s, salmon runs were declining. Overfishing, habitat destruction and lack of effective regulation — in 1890 it still was legal in Washington to fish with dynamite — were taking a toll. In 1898, the Washington fish commissioner, A.C. Little of Tacoma, alarmed by the declining runs, called for more aggressive artificial production to rebuild and save them, writing:

“. . .if from the cause of a large catch or from obstruction being placed in the streams only a small number reach the spawning beds, the result three or four years hence must inevitably be a short run of fish; and if none are able to ascend the streams, four or five years later, not a salmon will be found in that locality. It is therefore absolutely necessary that some of the fish ascend the streams and spawn, or that the supply be kept up from fry planted, produced in other localities.”

Little reported that he traveled more than 3,000 miles investigating salmon fecundity in the upper Columbia River Basin, noting along the way that salmon runs had dropped precipitously in tributaries including the Tucannon, Yakima and Touchet, a tributary of the Walla Walla River. He estimated that less than 5 percent of the Chinook run in the Little Spokane River was present in 1898, compared to former years. Those fish were significant because the Spokane and its tributaries historically produced some of the biggest salmon in the Columbia basin, the June Hogs as they would be called in later years. He estimated at least 50 million salmon fry would have to be released annually in the upper Columbia tributaries “ . . .to keep up the supply of the May and June run of Chinook to the amount of the last four or five years.”

Little recommended a tax on the commercial fishing industry to pay for increased hatchery production. As salmon became more scarce, he reasoned, the human impacts on the remaining fish — “an increasing amount of sewage and other pollution cast into the waters” — would intensify. This would “. . .make it more difficult each year to stem the current of destruction.”

Little noted the “existing skepticism” among the public about the usefulness of hatcheries, but nonetheless recommended that hatchery production be tripled in the Columbia basin and that a new hatchery be constructed on the Wenatchee River because it was known to produce large runs of Chinook salmon and summer steelhead. Little wrote that Washington’s salmon fisheries had been developed “. . .to the highest point attainable unless radical measures are taken towards keeping up the supply. In no way can this be done successfully but by artificial propagation.” In effect, he argued that the way to beat the salmon decline was to ignore its obvious causes and simply produce more fish.

This was a prevalent attitude among fish experts. In 1903, the U.S. Fish Commissioner told Pacific Fisherman magazine that “the successful system of hatcheries” in Oregon and Washington would assure “the permanency of the fishing industry. . .in the Columbia River District. . .for all time to come:” Artificial production solved the problem of restocking the river, he claimed. Yet in 1908, when the annual release of salmon fry into the Columbia River reached 34 million, Oregon’s Fish Commissioners were alarmed at the continuing decline of the spring Chinook runs.

Hatchery fish faced many of the same obstacles to survival as fish spawned in the wild — water pollution, degraded stream habitat, high water temperatures, the effects of predator fish and birds and overfishing. One problem was unique to hatcheries and the survival of hatchery fish. This was the impact of what was known as stock transfers.

Stock transfer means transferring eggs from one part of the basin to a hatchery, where they are incubated and hatched, and then transferring them to another part of the basin for release. The theory was that the hatchery could serve as a nursery to assist in rebuilding salmon and steelhead runs in streams where they were depleted. Plentiful fish from one part of the basin would repopulate other parts of the basin where runs were in decline. In 1909, the state of Oregon constructed the Central Hatchery, which later was renamed Bonneville Hatchery, on Tanner Creek on the lower Columbia River just upstream from the present-day Bonneville Dam. The Central Hatchery served as a place to incubate salmon eggs brought in from other hatcheries.

The resulting fry were shipped to streams all over the Northwest, not to their home streams. Some were released directly into the Columbia at the hatchery. Sometimes eggs from as far away as Alaska were incubated at the Central Hatchery and then released into the Columbia. In this way, scientists of the time believed, fisheries science would rebuild the depleted Columbia salmon runs with fish from other rivers. But in hindsight, stock transfers created the biological equivalent of hash, a melting pot of fish genes that diluted native stocks. Transferring salmon from other river basins to the Columbia disrupted the process of natural selection and adaptation that made Columbia River salmon uniquely suited to their home streams. As a consequence, the fry that were incubated at the hatchery and then released into streams generally did not survive well.

Faced with these failures, scientists reasoned the problem might be that the fish simply were too young and fragile to survive away from the protective environment of the hatchery. So beginning in 1910, hatcheries in Oregon experimented with holding juvenile fish longer before releasing them, on the theory that larger fish would be better able to avoid predation. The salmon harvest increased in 1914 and continued to improve through 1918. In 1919, the Oregon Fish and Game Commission reported approvingly, “. . .this improved method has now passed the experimental stage, and. . .the Columbia River as a salmon producer has ‘come back.’” But in the 1920s, a growing understanding of the life history of salmon caused biologists to question whether hatcheries actually might do more harm than good through overproduction, harvests based on high-volume production, and the subsequent decline of the wild component of the runs. At the time, hatcheries continued to release large amounts of juvenile fish, but the number of adult fish returning to spawn, while large, was not increasing. A U.S. Bureau of Fisheries report published in 1920 noted that “the hatcheries probably inflicted as much, or more, damage to the salmon runs than they had service of value.”

In 1922, biologist Willis Rich reported there was no evidence that artificial propagation of salmon had conserved the runs. From his investigation of the commercial harvest and the production of the canneries, Rich concluded that “… the popular conception, that the maintenance of the pack on the Columbia River is due to hatchery operations, is not justified by the available science.” Similarly, in 1923 the U.S. Fish Commission, which existed primarily to promote hatchery production, reported that without better knowledge of unique salmon life histories, the artificial production of salmon and efforts to conserve them by restricting and reducing harvests “may prove wasteful and ineffective while at the same time imposing futile obstacles to the development of a legitimate and essential industry.”

Increasingly, scientists could find no correlation between fish abundance and hatchery production. There was no scientific basis for the hatchery programs, and no regular or ongoing evaluation of their successes and failures. In a 1930 report, the dean of the School of Fisheries at the University of Washington, John Cobb, called hatcheries a threat to salmon fisheries and chastised the lack of critical evaluation of hatchery programs.

A ten-year government-funded study in British Columbia, which concluded in 1936, could not find a significant difference between the efficiency of natural salmon production and artificial salmon production. As a result, the government could not justify spending money on hatcheries and ordered them closed. In the United States, where there was strong ideological and political support for hatcheries, the results of the Canadian study were ignored.

By the 1940s, it was clear that artificial production of fish needed a better basis in fish science. The primary purposes of hatcheries continued to be the production of fish to compensate for habitat destruction and lost natural production, and also to produce large numbers of fish for commercial harvest. Scientific research accelerated as fish managers sought a better understanding of the complex salmon life cycle. Nutritional improvements boosted salmon production, as did improved understanding of the optimal times to release fish from hatcheries. But not until the latter decades of the 20th century did fish biologists begin to understand the impacts that hatchery fish can have on fish that spawn in the wild. For decades, fish managers in their optimism assumed that hatcheries could compensate for lost spawning and rearing habitat by simply outproducing nature. It was becoming clear that hatcheries could not do that, and, moreover, were contributing to the decline of wild fish by introducing fish into streams that looked the same as their wild counterparts but behaved much differently.

Scientists began to understand that the hatchery environment acclimates fish to artificial conditions, and that fish carry this learning into the natural environment. In the hatcheries, fish were raised by the hundreds of thousands in open concrete tanks called raceways and were fed pellets at regular intervals. The fish learned to grab the pellets as they fell through the water. They learned to dash for food when it appeared, and they had no awareness or fear of predators. Released into streams, they behaved the same way. Fish spawned in the natural environment, on the other hand, learned to conserve energy, avoid predators, eat when food is available and disperse to places in the aquatic environment where food, shelter and appropriate water temperatures are available. In the stream environment, hatchery-bred fish out-competed naturally spawned fish for food and were easy targets for predators.

Clearly it was time for a management change. Through the 1980s and into the 1990s fisheries management began to undergo a transition from a fish-production focus to an ecosystem-management focus, in which fish and wildlife are managed as part of the environment that supports them. This approach “places artificial production in the basin in a very different role than that employed in the past,” a panel of independent scientists, the Scientific Review Team of the Independent Scientific Advisory Board, reported in a 1998 review commissioned by the Northwest Power and Conservation Council. The Council’s Columbia River Basin Fish and Wildlife Program directs funding to hatcheries operated by Indian tribes that are experimenting with new fish production techniques. At the time, Endangered Species Act listings of salmon and steelhead, as well as resident fish, increased the management emphasis on conserving wild and naturally spawning stocks. The scientists noted this would increase concern over the potential for hatchery-bred fish to spawn with wild fish and cause genetic and ecological damage. However, the shift in management emphasis also could create a new and positive role for hatcheries to assist in species recovery, they suggested.

The scientists were not encouraged by the past inability of hatchery programs to respond to changing realities and changing scientific knowledge, but they were hopeful that hatchery management would change to allow experiments that might help solve past problems that contributed to the salmon decline, including inbreeding, domestication, overemphasis on a few types of fish and producing more fish than the receiving streams and rivers could support. The primary experimental fish production technique is called supplementation. This is the practice of raising fish in the artificial environment — which sometimes is made more natural by shading the water and placing woody debris, rocks and sand in the raceways where the fish rear — and then releasing the fish into streams. The theory is that the fish will return to spawn in the streams, not the hatchery, and thus rebuild naturally spawning runs. Fish raised in supplementation facilities, such as the one operated by the Yakama Nation in the Yakima River Basin, come from indigenous stocks. This increases the probability that the fish will adapt to the new environment.

Supplementation facilities have had successes in terms of adult fish returning to spawn, but the technique remains controversial among some scientists. In June 1999 the Independent Scientific Review Panel (ISRP), a group of 11 scientists who analyze projects proposed for funding through the Northwest Power and Conservation Council’s Columbia River Basin Fish and Wildlife Program, issued a report that questioned the use of supplementation. The ISRP expressed concern about unforeseen genetic interactions that could dilute the gene pool of the remaining wild fish, even though the brood fish are carefully selected from indigenous stocks.

Nonetheless, and with that caveat, and others, in mind, a gradual shift in hatchery management is under way in the Columbia River Basin, from the old-school focus on mass production in the lower parts of the basin to supplementation and increased production for harvest in the upper parts of the basin. Producing fish for commercial harvest in the lower river will remain an important focus, but hatcheries also will pay more attention to producing fish consistent with local goals and objectives for environmental restoration, harvest and species recovery in the upper basin. Experimental “terminal fisheries” in the lower Columbia are providing fish for commercial harvest in bays and sloughs along the estuary shorelines of Washington and Oregon. For these fisheries, salmon are raised in floating net pens, released to go to the ocean and then return to the vicinity of the net pens as adults, where they are harvested. In this way, fishing pressure is reduced in the mainstem of the river, where the weaker wild stocks are migrating to upriver spawning areas.

The new paradigm of incorporating hatcheries into the environment and reducing the number of fish produced is a sea-change in thinking about the role of hatcheries. How might the necessary changes be made, to which facilities, over what time period, and regarding which species? In theory, at least, the increasing emphasis on artificial production of fish as a component of ecosystem management has promise to rebuild not only the diversity of salmon and steelhead in the Columbia River Basin, but also the number of fish returning to spawn. But in practice, wrenching change can be difficult.

To better understand the necessary changes in hatchery practice, in 2007 Congress authorized an independent review of hatcheries in the Columbia River Basin, and in 2009 the Hatchery Scientific Review Group (HSRG) issued a report with recommendations that had as its foundation the concept that salmon conservation goals need to be met for key naturally spawning populations while at the same time maximizing salmon harvest. In order for hatchery actions to effectively address conservation goals, harvest reforms are also necessary, according to the HSRG, which previously studied hatcheries and recommended reforms for facilities in Puget Sound.

To organize an ecosystem-level review of such a large landscape as the Columbia River Basin, the 14-member HSRG divided the basin into more than a dozen regions based largely on earlier definitions developed by the Northwest Power and Conservation Council. The regions then were grouped into four areas: Lower Columbia, Mid Columbia, Upper Columbia, and Snake River. The scientists analyzed every state, federal, and tribal salmon and steelhead hatchery program in the Columbia River watershed, covering 178 hatchery programs and 351 salmon and steelhead populations. The resulting report, issued in 2009, included scientific principles of reform, hatchery-specific recommendations, new analytical tools, and procedures that provide a foundation for managing hatcheries more effectively into the future.

Here is a brief review of the HSRG’s general principles for hatchery reform in the Columbia River Basin and corresponding generalrecommendations:

Principle 1: Develop clear, specific, quantifiable harvest and conservation goals for natural and hatchery populations within an “all H” (hydropower, hatcheries, harvest, habitat) context

  • Recommendation 1: Express conservation goals in terms of a population’s biological significance (primary, contributing, stabilizing) and viability (natural-origin spawning abundance and productivity).
  • Recommendation 2: Express harvest goals in terms of a population’s contribution to specific fisheries.
  • Recommendation 3: Ensure that goals for individual populations are coordinated and compatible with those for other populations in the Columbia River Basin.

Principle 2: Design and operate hatchery programs in a scientifically defensible manner

  • Recommendation 4: Identify the purpose of the hatchery program (i.e., conservation, harvest, or both).
  • Recommendation 5: Explicitly state the scientific assumptions under which a program contributes to meeting the stated goals.
  • Recommendation 6: Select an integrated or segregated broodstock management strategy based on population goals and hatchery program purpose.
  • Recommendation 7: Size hatchery programs based on population goals and as part of an “all H” strategy.
  • Recommendation 8: Manage harvest, hatchery broodstock, and natural spawning escapement to meet HSRG standards appropriate to the affected natural population’s designation.
  • Recommendation 9: Manage the harvest to achieve full use of hatchery-origin fish.
  • Recommendation 10: Ensure that all hatchery programs have self-sustaining broodstocks.
  • Recommendation 11: Coordinate hatchery programs within the Columbia River Basin ecosystem to account for the effects of all hatchery programs on each natural population and each hatchery program on all natural populations.
  • Recommendation 12: Assure that facilities are constructed and operated in compliance with environmental laws and regulations.
  • Recommendation 13: Maximize survival of hatchery fish consistent with conservation goals.

Principle 3: Monitor, evaluate and adaptively manage hatchery programs

  • Recommendation 14: Regularly review goals and performance of hatchery programs in a transparent, regional, “all-H” context.
  • Recommendation 15: Place a priority on research that develops solutions to potential problems and quantifies factors affecting relative reproductive success and long-term fitness of populations influenced by hatcheries.
  • Recommendation 16: Design and operate hatcheries and hatchery programs with the flexibility to respond to changing conditions.
  • Recommendation 17: Discontinue or modify programs if risks outweigh the benefits.

Many current hatchery programs have been operated in a manner that disrupts the natural selection for population characteristics that are tailored to local conditions in the natural environment. One of the most significant problems associated with hatcheries is the unknown genetic effects of interbreeding between hatchery and natural-origin fish, particularly in the natural environment. In fact, there is mounting evidence of the negative effects and reduced level of productivity that results from domesticated hatchery fish breeding with wild fish in the natural environment.

Throughout the Columbia River Basin, harvest is a primary goal of most hatchery programs, but conservation of unique genetic material is becoming increasingly important. In order for hatcheries to operate effectively for either purpose, interactions with naturally spawning fish populations need to be addressed. The HSRG identified two strategies for limiting the adverse effects of hatchery fish on natural populations: 1) sufficiently isolate hatchery fish from wild fish; and 2) assure that the hatchery fish are as similar to wild fish as possible. To this end, hatchery programs should be managed as either genetically integrated with, or segregated from, the natural populations they most directly affect, according to the HSRG. The intent in either case should be to not allow hatchery fish on the spawning grounds with wild fish, as the hatchery-origin fish pose a risk to natural fish.

The HSRG also recommended that hatchery fish from harvest programs and from most conservation programs should have an external mark, such as a clipped adipose fin, so that hatchery-origin and natural-origin fish can be effectively managed to meet standards for composition and abundance of natural-spawning escapement and hatchery broodstock.

HSRG recommendations have been adopted, or are being considered, by hatchery managers throughout the Columbia River Basin. For example in 2010, the U.S. Fish and Wildlife Service issued hatchery-reform reports for its hatcheries in the Columbia River Basin. The state of Washington, where salmon hatcheries have been operating since the 1800s, is changing its hatchery policies in light of new information and science. Today, state hatchery managers understand that hatchery fish can pose risks to naturally spawning populations, and hatchery reform partly informed by the HSRG recommendations is under way. The goal is to reduce the risks posed by hatchery fish while continuing to provide fish for harvest. In 2012, more than 60 percent of the Washington Department of Fish and Wildlife hatchery programs meet HSRG standards, and that work is continuing. Forty-five percent of the state’s broodstock standards meet HSRG standards, and the state is on track to hit 90 percent by 2015. In addition, consistent with the HSRG recommendation to mark more hatchery fish so they can be identified when harvested, state-run hatcheries now mark 100 percent of the coho and steelhead they produce and 95 percent of the Chinook salmon.

Through the work of the HSRG and the region’s state, federal, and tribal hatchery managers, a better understanding is evolving about the risks that hatchery practices may have on long-term productivity and fitness of natural populations. Coordination and information sharing is improving among hatchery managers, as exemplified by the development of the Anadromous Salmonid Monitoring Strategy (ASMS), the Power Council’s Monitoring Evaluation Research and Reporting Plan (MERR), and the development of high-level indicators to help report and track biological and physical indicators (Viable Salmonid Population indicators, for example) of progress in improving fish populations.