"Across the Concrete Mortality" in Mainstem Dams

“. . . the fish passage and propagation facilities of the Corps of Engineers are essentially designed, constructed, and operated on the basis of recommendations and criteria received from the fish and wildlife agencies. Therefore, when a facility works well we can all share a sense of pride. Conversely, when facilities do not perform to meet our expectations, we all must share that responsibility as well.” E. Mains (1976).61

Application of a scientific perspective propelled the federal government to supply nearly all the dams in the Pacific Northwest with fish ladders. The fish ladders were widely recognized as a success. The U.S. Department of Interior observed that salmon climbed the fish ladders with “far less effort than their forebears that fought upstream through the swirling rapids that are now buried beneath fifty feet of water”; the Oregon Fish Commission considered them “entirely successful”.62

From the perspective of removing obstacles to salmon migration, even the drowning of Celilo Falls by construction of The Dalles Dam was helpful to salmon. As one fishery agency official candidly told reporters at the time, “it would be easier for the fish to go over a ladder in the dam than to fight their way over Celilo Falls”.63 Today such statements are cited as examples of “at least a lack of sensitivity toward Indians among government officials”.64 Truth and sensitivity are often at odds.

Science was not perfect, because providing fish ladders alone is not enough. Forty years ago the question of the downstream migration was rising to prominence, as public figures like reporter (and later Senator) Richard L. Neuberger asked: “But what, in the roaring darkness of the turbines and penstock shafts, what happens to the tiny fingerlings on their turbulent voyage to salt water?”65

Environmentalists often proclaim the dams are “rigged with turbines that churn like giant Cuisinarts making salmon and steelhead puree”66, like the infamous "Bass-o-Matic" demonstrated by Dan Ackroyd on Saturday Night Live. The Idaho environmentalists, the most extreme of their breed, went so far as to run a television ad in the spring of 1995 that showed a blender, followed by a red screen leaving the results to the audience's imagination.

This is a hoax too. Turbines rarely kill salmon by cutting them up. They have large flat blades that spin about the same speed as bicycle tires—80 rpm. The vast majority of salmon get through untouched.

In the early days, before the engineers learned to run the turbines so as to minimize injuries to salmon, they were run in a way that involved "cavitation", in which bubbles form because of the extreme turbulence and shear forces in the water. Fishery biologists putting marked groups of juvenile salmon through the turbines found mortality anywhere from 4% to 19%. But it is hard to recover marked salmon in a huge river, and the precision of early estimates of losses is highly suspect.

More advanced studies in the 1980s produced a wide range of results at different projects, ranging from 2.3% losses at the second (and newer) powerhouse at the Bonneville Dam to as high as 16.9% at Lower Granite Dam. Behind these figures are enormous experimental problems and difficulties in the interpretation of data.

For example, fishery managers object to running the second powerhouse at Bonneville, even though, because of through advanced design, its direct turbine mortality of 2.3% is low. That is because predator concentrations seem to be particularly high below the second powerhouse, cancelling out the benefits of good turbine design. Conversely, the 16.9% figure at Lower Granite is regarded as suspect by the scientists who conducted the study, because a critical statistical assumption—that control and test groups mixed—did not appear to be true.67 The devil is always in the details.

By the 1990s, technology had advanced to the point where researchers could attach tiny balloons to salmon that inflate after a delayed chemical reaction.68 Salmon passed through the turbine pop to the surface quickly after turbine passage. Sometimes a dead and battered salmon comes up. But not very often—about three times out of a hundred. Researchers have observed salmon passing through turbines to compute 120 hour survival estimates, so that immediate and delayed mortality can be computed. The results: 95% survival through the turbines at Lower Granite Dam.69 Even those results may overstate turbine mortality if “culling” is occurring; it is possible that the weakest members of the population may be weeded out at the first dam, tending to lead to higher estimates of turbine mortality there.70 Similar results have been obtained at Rocky Reach Dam in the upper Columbia River.

There is also good reason to believe that turbine design can be improved substantially.71 Since many of the mainstem dams on the Columbia and Snake Rivers will need substantial renovation over the coming decades, there will be many opportunities to reduce turbine mortality even further.

The studies on turbine mortality have had little impact on the turbine component of the Great Salmon Hoax. Newspaper articles continue to report that “[s]almon are being spun around and chewed to bits in the giant turbines of the Columbia-Snake River dam system”.72 Harvest managers and environmentalists continue to assert as a scientific fact that turbines kill 10-15% or even more of the salmon that pass through them.73 The Northwest Power Planning Council’s Independent Science Group essentially ducked its charge to apply science to the question of turbine mortality, instead declaring that "[a] generally accepted figure now is 15%" per dam.74 In salmon recovery, what is "generally accepted" and what is true are often very different things.

Even if everyone agreed that direct turbine mortality was 5% or less per dam, the cumulative result, over eight dams, is significant. Statistically, if every salmon had to go through the turbines at all eight dams, it would work out to 34% mortality over the course of the migration. That sounds pretty bad. Luckily, that's not the right calculation.

First, the mortality is not entirely additive. Some of the 34% of salmon killed by dams would have been killed by something else, particularly predators. There is a surplus population which is doomed; the cause of death varies but the certainty of loss is invariable. This is a well-recognized phenomenon common to all animal populations called "compensatory mortality" or "competing causes of death". In human populations, for example, curing a disease that kills some fixed percentage of population will not increase the population by that same rate, because other causes of death "compete" to fill the void.

Thus we cannot get a one-for-one increase in salmon populations by reducing any single source of mortality, and we get even less for solving a juvenile mortality problem, because the other causes of death have a long time to compete to fill the void. The fishery agencies, biased to ignore adult mortality, refuse to acknowledge this phenomenon. Indeed, as far as I know, no one has ever even tried to make a quantitative estimate of the extent to which dam-related mortality is compensatory.

It is hard to estimate how much incremental harm that 34% mortality would do to salmon populations. If half the salmon were likely to die anyway, would the dams be killing an additional 10%? 20%? 30%? While both the Northwest Power Act, the Endangered Species Act and other federal law requires agencies to assess the true effects of the dam operations, they have never done so, claiming ignorance.

Instead, all involved simply assume that the turbines could, collectively, kill a third or more of the fish. And to fix the problem, the U.S. Army Corps of Engineers has put in fish bypass systems so that the vast majority of the salmon never go through the turbines.

Figure 14: Schematic Drawing of Dam Passage Routes

The most expensive devices are huge, submerged traveling screens that guide the fish away from the turbines (identified as "bypass" in the drawing). Current extended screen technology is highly effective at helping fish avoid the turbines. The screens deflect 80% or more of the yearling chinook, including endangered Snake River spring/summer chinook, at Little Goose and McNary Dams. Results on smaller subyearlings, including endangered Snake River fall chinook, range upwards of 50% at many projects. Thus at projects with screens (not including The Dalles),75 most salmon never go through the turbines, so the turbine losses are far lower.

The screens are not perfect, in that, in theory, fish can be damaged by bouncing against them. However, studies at Lower Granite dams have found that “juvenile fish suffer minimal injury and delay with extended screens”.76 There is another problem associated with all bypass systems: large numbers of juvenile salmon collected in the system are released from pipes below the dam. Critics have theorized that this makes it easier for predators to catch the juvenile salmon, an effect that has been demonstrated at the bypass system of the second Bonneville powerhouse.77 Evidence suggests much less of an effect at upriver dams, perhaps because predator concentrations are lower upriver. Of course, no one knows how many predators used to concentrate below rapids and waterfalls before the dams were built.

Yet another way salmon can avoid the turbines is when the turbines are turned off and water is spilled over the top of the dams. Survival of fish passing over a spillway can be as high as 98% or more. Spill is the favored passage method of environmentalists, but destroys the economic benefit of dams since spilled water cannot be used to generate electricity. About 20% of the annual planned resources for salmon recovery are presently squandered on spill programs, the subject of Chapter 12. The principal vice of spill, beyond the enormous economic waste, is that at several projects, a decision to spill water siphons fish away from the bypass systems where they could collected for "transportation".

Smolt transportation means collecting juvenile salmon with a bypass system and putting them into barges. They are towed through the locks of the dams, and avoid "across-the-concrete" mortality as well as natural predators. The overwhelming success of this "transportation" program—discussed in Chapter 5—makes it a key target of environmentalists. The transportation program stands as the single most significant obstacle to their ultimate goal: removal of the dams.

As a result of transportation, bypass and spill the vast majority of migrating salmon and steelhead never go near a turbine. Federal fisheries scientists acknowledge that “transportation mitigates for direct losses through the hydropower system under a broad range of flows”.78 Overall survival rates through the Columbia and Snake Rivers, with maximum transportation, can be upwards of 80%. Even without transportation, the most recent analyses by scientists at the National Marine Fisheries Service suggest that in-river survival is at least “66% of historic rates”.79 The true number is probably higher.

61 E. Mains, “Corps of Engineers Responsibilities and Actions to Maintain Columbia Basin Anadromous Fish Runs”, in E. Schwiebert (ed.), Columbia River Salmon and Steelhead 40-41, Spec. Pub. No. 10 (Am. Fish. Soc. 1977).

62 Quoted in K. Petersen, River of Life, Channel of Death 109.

63 Memo to File, S. Hutchinson (U.S. Fish & Wildlife Service), Jan. 16, 1951 (recounting conversation with reporter Herb Lundy); excerpted in J. Cone & S. Ridlington, The Northwest Salmon Crisis: A Documentary History 207 (OSU Press 1996).

64 J. Cone & S. Ridlington, The Northwest Salmon Crisis: A Documentary History 213.

65 Quoted in A. Netboy, Salmon of the Pacific Northwest: Fish vs. Dams xi (Binfords & Mort Portland 1958)

66 B. Matsen, Ray Troll’s Shocking Fish Tales 91 (Ten Speed Press 1993).

67 The research results are summarized in ISG, Return to the River 273-75.

68 See generally, D. Mathur & P. Heisey, “Debunking the Myths about Fish Mortality at Hydro Plants”, Hydro Review, April 1992, at 54-60.

69 J. Skalski et al., “Turbine Passage Survival of Juvenile Spring Chinook Salmon (Onchorhynchus tshawytscha) at Lower Granite Dam, Snake River, Washington”, U.S. Army Corps of Engineers Contract No. DACW68-95-C-0031 (October 1995), Table 3-6.

70 See C. Steward, “Assessment of the Flow-Survival Relationship Obtained by Sims & Ossiander (1981) for Snake River Spring/Summer Chinook Salmon Smolts”, Final Report, BPA Contract No. DE-AM79-93BP99654, at v (April 1994) (“evidence suggest that a disproportionate percentage of chinook smolts died at the first dam encountered on the Snake River, presumably due to the culling of unfit fish”).

71 See generally J. Ferguson, “Improving Fish Survival through Turbines”, Hydro Review, April 1993, at 54-61.

72 “Our debt is overdue”, The Astorian, May 13, 1997.

73 B. Harden, A River Lost 71 (10-15% mortality).

74 ISG, Return to the River 273.

75 Plans are in place to install extended screens at The Dalles. Unfortunately, the States and tribes have begun resisting structural improvements to bypass systems dams they think should simply be removed.

76 “Fish Research: Filling the Gaps”, Salmon Passage Notes, at 2 (USACE NPD Feb. 1996).

77 Ledgerwood et al. (1991).

78 Id. at 9.

79 J. Willliams, G. Matthews & J. Myers, “The Columbia River Hydropower System: Does It Limit Recovery of Spring/Summer Chinook Salmon?”, at 4.

Previous PageTable Of ContentsNext Page

This Web page was created using a Trial Version of HTML Transit 3.0.