The Science of Salmon Recovery in the Columbia River Basin

The Impact of Hydroelectric Dams on Salmon Populations

These huge dams, and many more built later on the Columbia River and the Snake River, dramatically intensified the problems involved in restoring anadromous fish runs. Grand Coulee was constructed without any fish passage capabilities. The rational was that hatcheries constructed downstream at Wenatchee and Chelan would mitigate for lost wild salmon reproduction. Plans for Bonneville Dam construction were finalized without allowance for adult fish passage. However, upstream fish interests forced the Corps of Engineers to install fish ladders four years after initial construction began (Petersen 1995). The Corps viewed this as a major concession to up-river fish interests; the fish ladders probably would not have been built without the intense criticism from these upstream groups. The Corps’ original intention was simply to concentrate all fish propagation downstream from Bonneville Dam. Their reasoning was based not only on a highly unrealistic optimism concerning the possibility of wholly replacing natural propagation of the fish with hatchery production, but also on the belief that they were only accountable to down river harvesters of fish.

During the 1940s the Army Corps of Engineers also struggled with the realization that they were accountable not only for enabling adult fish to migrate over the dams safely, but also for the safe passage of seaward-migrating smolts past dams. Willingness to accept responsibility for the safe passage of smolts down river came grudgingly, however. Some Corps officials claimed that they had not been informed by fishery scientists that they had to worry about downstream migration of juveniles. This claim was made despite Corps dam-planning reports that showed otherwise (Petersen 1995). In 1947, the Corps refused to publish research findings, from one of its own scientists, which identified a 15% smolt loss during passage over Bonneville Dam (Petersen 1995). These findings were dismissed by Corps administrators as inconclusive, a response to which they increasingly resorted. All action in response to these findings and criticism was stalled in favor of conducting further studies and seeking easier solutions, such as increased hatchery production.

The truth is that during the 40 year period of their intense preoccupation with building dams, from approximately the mid 1930s to the mid 1970s, the Corps of Engineers gave only grudging and token consideration to the impacts their dams would have on anadromous fish runs. Many of their attitudes were similar to other industrial and logging interests and probably reflected the dominant social and political values of the time. What is perhaps less understandable, and even less excusable, is that they were assisted in this neglect of the fish runs by many of the fishery scientists of the time. Partly out of desperation in dealing with the enormously destructive impacts to salmon populations, but largely as a consequence of their unbridled optimism that science could solve any problem it seriously applied itself to, fishery scientists issued two claims: First, science would improve dam technology and provide safe passage for migrating juvenile salmon as well as adults. Second, increased development of hatchery technology was all that was needed to make hatcheries work (Croker and Reed 1963). Importantly, this optimism was cited frequently by the Corps of Engineers in their justifications for the construction of each successive dam. Stronger opposition to the dams from fishery scientists with more realistic assessments of their own capacities to mitigate the destructive effects of the dams on fish runs might possibly have prevented the building of some of them. Obviously, the current salmon crisis has proven their optimistic assurances to be unjustified.

While the development of fish ladders was technically accomplishable, although very expensive, and all dams built after Grand Coulee on the mainstem of the Columbia were built with adult fish passage facilities, the downriver passage of smolts proved more difficult. Passage of seaward migrating smolts has remained a problem that both engineers and fishery scientists have been unable to resolve. Research in the late 1980s and early 1990s makes this evident, indicating an 15-45% smolt mortality during downstream migration at each lower Snake and Columbia River dam and associated reservior (Raymond 1988). Adult up-river migration past dams and smolt down-river migration past dams were each extremely complicated, both on an engineering and biological level. Accordingly they were very expensive problems to solve. Dam-builders and fishery scientists were understandably ready to fall back on hatcheries as the magical cure. Again, some fishery scientists insisted that hatcheries could provide the unlimited numbers of fish to compensate any and all destructive forces; they simply had to learn more about the artificial propagation of salmon (Croker and Reed 1963).

With the passage of the Mitchell Act in 1938, which allocated funds to mitigate against salmon declines in the Columbia River basin, fishery scientists were given the funds to construct additional hatcheries and to develop the hatchery technology they insisted was needed (Cone 1995). Hatchery technology developed during this period, and became the preoccupation of much of the fishery research completed. Unfortunately, this often occurred at the expense of research focused on habitat restoration or changes to dams to make them more fish friendly. An example of this preoccupation with hatchery research and renewed interest in hatcheries as the ultimate cure can be found in the, Report of Second Governor’s Conference on Pacific Salmon (Croker and Reed (Editors) 1963). This report is a collection of research findings from 1962-1963 and proposed research projects from leading Northwest universities and state and federal fisheries departments. Notably, 23 of the 40 studies described in this report focus on research on supplements and substitutes for natural reproduction. Only 6 of the studies describe research focused on maintenance and improvement of natural production. Fishery scientists reported that: "never has artificial propagation of salmon and steelhead been in such dire need and fortunately, never has the future of this science been so bright." While there were some reservations expressed, the primary need, many maintained, was for even more research on hatchery technology.

During the 1950s, 1960s, and 1970s some hatcheries did experience success. Improvements made in the husbandry science side of fish hatcheries included: better feeds, disease control, water quality, and record keeping techniques. These improvements led to high survivability of eggs and juveniles produced in such hatcheries (80-90% survivability from egg to juvenile release), and in some instances to good returns of hatchery fish (Croker and Reed 1963). But as the science of hatchery technology improved, the desire to understand other important elements of fishery science lessened. This was true both among the vested interests who did not want to spend money on research focused on improving dams or reducing habitat alterations, and among many fishery scientists who became convinced that hatcheries were the cure all.

This mindset of many fishery scientists of the time foretold problems. Some talked naively of improving salmon through selective breeding, expressed in management guidelines that: "the largest, best-shaped and best-colored females should be selected from stocks available for egg production…" and that "grading in the usual manner will cull out those fish in the select lots which are not able to compete or grow at optimum rate" (Croker and Reed 1963). Some even advocated the intentional goal of adapting fish to requirements of hatchery rearing, stating that "the most successful of these hatcheries seem to be the ones which have developed stocks adapted to the hatchery practices and conditions of release" (Croker and Reed 1963). Many shared a preoccupation with quantity of production, almost to the exclusion of quality, believing that the most important element of successful supplementation was massive production and release of smolts. In fact, successful hatchery programs were defined as being those with the highest survivability of eggs and larval fish, rather than return of adult fish. Finally, the arrogance of many fishery scientists and hatchery managers of the 1960s and 1970s led them to even more ambitious objectives. These included introducing excess fish into rivers historically not supporting fish runs and to improving salmon populations to above historic levels for the benefit of the commercial and sport fisheries.

Not all fishery scientists during this period believed that improved hatchery technology gained through research should be used prior to defining the proper role of hatcheries: "basic research must be conducted in an atmosphere free of encumbrance of the demand for quick answers and immediate applicability" (Croker and Reed 1963). But many scientists did. They, along with all of the vested interests who were concerned only with a quick fix and avoiding interference in their own destructive practices, kept on proclaiming that hatcheries were the magical cure. Thus, they continued to push for increased construction of new and improved hatcheries.

The 1980s finally produced a wealth of new, researched-based criticism directed at hatcheries. Considerable evidence was developed indicating that the selective breeding many claimed would lead to "improved salmon" had led instead to the domestication of salmon. The outcome was fish that were less disease resistant, less efficient natural feeders, more vulnerable to predation, and poorly adapted in general to competing, surviving, and reproducing in the wild (Meffe 1991; Hilborn 1992; Stickney 1994).

Many studies also suggested that the policy of transplanting fish from one drainage to another had not only resulted in many unsuccessful introductions, but also had very negative effects on native wild stocks of salmon. In fact, the most important research of this period, finally focused on the tremendously important, and previously ignored, question of how hatchery fish affect wild fish populations. Numerous studies established that hatchery introductions led to predation of wild fish by hatchery fish, spread of disease, and competition between wild and hatchery fish within streams for inadequate food supplies (Hilborn 1992). Studies also reported that the priority hatcheries had given to increasing the harvest of fish over preserving the genetic viability of salmon populations, had led to genetic mixing and loss of genetically separate stocks, and the dilution of the genes that had been formed to adapt a wild stock of salmon to its habitat (Hilborn 1992).

In addition to these effects many scientists criticized management decisions, such as water-releases at dams, made in the interest of hatchery smolts over the interests of wild smolts. Many scientists also returned to emphasizing the claim, first issued by salmon advocates twenty years prior, that hatcheries simply provided an ongoing justification for continued over-harvest of salmon, including endangered wild fish (Meffe 1991). Preoccupation with artificial propagation had led to continued neglect in both research on and expenditures for habitat restoration and control of over-harvest, destructive impacts of logging, grazing, agricultural pollution, mining and manufacturing pollution. Over 90% of the total funds spent on restoration of fish runs during the 1980s had been spent on fish hatchery and dam by-pass technology, compared to the less than 10% spent on habitat restoration and improvement (Petersen 1995).