Salmon Science

Executive Summary

The science that serves as a basis for the salmon restoration cites human activities as the sole cause of the salmon decline. There are many other alternative reasons for the decline in the salmon population. For example, increase in number of predators, decrease in food supply and change in ocean conditions may have caused the decline in salmon population.

To date, the majority of effort to restore the salmon has been directed at restoration of river habitat. In 1950 a study showed that river habitat was not necessarily connected to salmon decline. Other data exists that show the decline in salmon population may not be related to river habitat. Present science is unable to determine the amount of salmon that will be restored by habitat improvement. Evidence exists that excellent stream habitat is available but is not being used by salmon. Thus, it is unclear how further improving the habitat will increase salmon population.

If we are going to save the salmon, we must get a model that accurately predicts the fluctuations in the salmon’s total life cycle. Present models stop at the river mouth. Any workable model for salmon restoration must go from egg to egg and food supply, the effect of predators, and the quality of the total habitat,(river, estuary and ocean).

Since the turn of the 20th century, legislation has been passed to protect many of the salmon’s predator species and their habitat. The result is that the total number of salmon predators has been on the rise since the early 1900s. It is one thing to protect a species and another thing to feed them. There is only a finite amount of food on this planet and very little goes to waste. Whereas the conservation movement had good intentions, it was naive to think that increasing one species would not seriously effect another.

To ignore the role of predators in the drastic reduction in salmon population over the last 100 years is a major oversight in the current science and can significantly change the conclusions. Consider that a partial list of salmon predators include at least 58 different species. The human fishing industry exists on salmon not caught by the other natural predators and evidence suggests that the present human catch is less than 1% of the total predator kill.

Sea lion, harbor seals and northern fur seals are responsible for the loss of 155 million salmon per year. In addition, these three species are also responsible for a substantial reduction in the salmon's food supply because they also prey on food sources common to the salmon. Further, since there is a strong connection between the amount of prey and the number of predators, if we increase the number of salmon by any means without artificial predator control, we will simply increase the number of salmon predators.

The current base line of 10 to 16 million salmon prior to 1800, often quoted by the popular press, is based on erroneous assumptions, and over simplified calculations. More realistic estimates from several sources place the size of the predevelopment runs between 1 and 2 million salmon, which is very close to the number of salmon currently in the river. Thus, by removing all the dams, farms, industry, towns and logging, completely restoring all the wet lands and the river to its 1805 condition, the probability of increasing the salmon runs is extremely low, since we may be currently at or near the 1805 runs.


According to the Oregon Coastal Salmon Restoration Initiative (OCSRI), the Salmon crisis is a product of a long sequence of assumptions and decisions made by humans.[1] One of the major reasons the OCSRI felt that past salmon recovery plans failed was because the conceptual foundations were largely based on untested assumptions.[2] It appears that history is about to repeat itself. The draft Environmental Impact Statement (EIS) for the lower Snake River Juvenile Salmon Migration states, "…the 1995 Biological Opinion provide the basis for the actions contemplated in the FR/EIS."[3] Note the word opinion; there is considerable difference between fact and opinion and one of those differences is, that opinion is based on untested assumptions.

All the current plans to save the salmon are based on a body of science that while large, is disjointed and contains numerous opinions and untested assumptions, lacks objectivity, and focuses mostly human activity in the river habitat. While considerable research shows that various human activities do degrade the river habitat for salmon use, degradation is not an absolute variable. There are degrees of degradation and no research is yet available that connects degradation to the amount of salmon actually lost due to the specific degradation.

For example, the current science does not or cannot estimate the number of salmon that will return if the four dams on the Snake are removed. The costs in dollars and degradation of human habitat are considerable. If we do not know how many salmon will return, that action is not unlike buying a pig in a poke. There is also evidence that not all habitat areas suitable for spawning are used to their capacity. It is unclear how further improving the habitat will increase salmon population.

Although much of the science behind salmon restoration has been reported to be peer reviewed, this short review has found several basic errors. Peer review implies that other scientists agree 100% with all of the findings in the report. The errors in these studies cast suspicion on the reviewers, the review process or both. To be meaningful, the peer review process must address all reviewers’ comments to the satisfaction of the reviewers. It is apparent that the peer review process as well as the science is flawed.

Salmon restoration is an important issue with far reaching consequences, and deserves to be decided based on fact; not a poll of scientific opinion. Restoration is complex and revolves around the ocean, estuary and river habitats; food supply, and predator populations. To date, the effort to restore the salmon has been directed at restoration of river habitat. Over the past 35 years, human activities along the river habitat have been subjected to increasing control by government agencies. These actions have not altered the decline in salmon population.[4] The common explanation for this failure is that actions taken have not been sufficiently restrictive, and that more regulation is needed. However, an equally valid conclusion to the lack of results is that regulating river habitat has little effect on salmon population.

Flawed Science

Saving salmon should not be left to guessing or biological opinion as to the cause of the decline. There is a bid difference between the number one answer and fact. Good science starts with sound assumptions and moves to a logical conclusion based on observed facts and examination of all possible hypotheses. Many individual studies on salmon restoration are based on sound scientific principles.

However, the large body of science that exists is almost entirely made up of various studies that focus on habitat degradation. These studies have shown that dams, river flow patterns, organic and sediment input, riparian habitats, migrational impediments, dissolved gas levels, pesticides, and industrial and municipal waste all degrade salmon habitat. However, degradation is not an absolute variable. Degradation has a wide range of values. Consider the difference in habitat degradation from adding a few pounds of silt to a stream compared to adding 500,000 yards of slit to the same stream.

Presently we do not have the capability to distinguish between the precise effects on the salmon due to these two extremes, let alone compare silt degradation to dissolved gas degradation. Currently the studies are qualitative, what we need is quantitative studies. Until we can put numbers on the amount of degradation, we do not have the science to judge the outcome of any habitat mitigation action.

For example, if we take out a dam, what we do not know how many of those fish that we saved will be killed by predators, starve in the ocean, or be killed in the rapids and shallows created by the natural river when they return to spawn. In other words, what is the net outcome of our action relative to the overall salmon population? Evidence suggests that the net action will be negligible. This was essentially confirmed by Will Stelle from NMFS who was unable to tell how many salmon would be returned by removing the dams. The NMFS and Corp. scientists working on the project concede that it may be none.

Where as a few individual studies, are flawed, the major flaw occurs in the over all body of the work. The “science” is not yet at a stage where cause and effect connections can be made.[5] Flawed science means, incomplete research, lack of objectivity and substitution of opinion for fact.

Incomplete Research

To date most of the science behind salmon restoration is directed at the effects of human activities in the river habitat. There is evidence that the river habitat is not that important to salmon survival.[6] This is a little like trying to determine if a plane will fly by studying its landing gear. Fortunately there are a growing number of scientists who realize that the issue also concerns oceans, weather, food supply and predator populations. If we are going to understand the salmon puzzle we need to identify all the pieces and connect them correctly.

Lack of Objectivity

Lack of objectivity shows in the type of studies performed and the lack of looking at alternative, valid explanations. There are a preponderance of studies that support the notion that human activities caused the decline in salmon and few that detract from that idea. Where are the studies to show the positive effects of dams on salmon population? To simply assume there are none is too naïve and not good science.

For example, rapids and shallows are dangerous areas for salmon. Predators find the fish easy prey at these constrictions in the rivers. Rapids are notorious for bruising and damaging fish, shallows sunburn their backs. We know exactly how many smolts are kill be a turbine. We have no idea how many are destroyed by the natural river itself. Dams replace these bruising, fish damaging natural barriers with fish ladders designed with resting pools to make the passage less demanding on the salmon’s food reserves. The fish ladders also protect the salmon from predator attack. It is entirely possible that dams net effect on salmon survival is positive. Good science lets the facts lead to the truth rather than the hypotheses under consideration lead to predetermined conclusions.

The infamous Caspian tern study was designed to show the damage that the Corp. of Engineers was causing by building spoil piles at the mouth of the river. It’s intent was to stop the dredging. The lack of objectivity has reach such proportions that it never occurred to the scientist in charge of the study that their data would implicate predators as a major problem. From recent discussions with OSU, they are now scrambling to lay the blame at the feet of hatchery salmon. Their refusal to admit that the terns are a problem is a strong indication that objectivity is lacking.

Consider that a single cow in a stream causes degradation to the stream, but a herd of elk do not. The science is pushing to have all cattle fenced from natural drainage paths. Is the sciences also suggesting we fence all the streams in the natural forests to keep deer and elk away from them as well.

There are two basic approaches to scientific study, one where the investigators try to prove themselves right and another where they try to prove themselves wrong. It is easier to maintain objectivity using the approach of trying to prove the theory under question wrong. In science, wrong is more powerful than right. For example, if the theory is proved right 100 times and proved wrong once, the theory needs to be revised to accommodate the new data.

There are many credible alternative reasons for the decline in the salmon population. For example, an increase in number of predators, a decrease in food supply or changes in ocean conditions may have caused the decline in salmon numbers. To ignore the effects of these alternatives is flawed science at its worst.

Soft Science

Soft science contains words like “possibly, we think, in our opinion, could be, may be,” and so on. Soft science is not science at all; it is more like propaganda and should be avoided. Science needs to deal in facts and only facts. It is better to say nothing, than try to make conclusions on possibilities. Imagine storing nuclear waste in downtown Portland based on the scientific opinion that it “could be” all right. The saving of salmon is no less important that proper storage of nuclear waste.

Complete Salmon Model

The inconsistencies discussed above are due primarily to the lack of any realistic model of the salmon life cycle. Science is the process of predicting what will happen before it happens. For example, predicting the direction a ball will go when you drop it. In order to do that you need a model. In the case of the ball, the model is simple because the system that controls the falling of the ball is simple. In the case of salmon survival, the system that controls the survival of the salmon is complex; therefore, the model also must be complex or the answer it produces will be inadequate.

Without an accurate model, we are reduced to randomly twiddling the knobs of nature with no real assurances of what will happen. If we are going to save the salmon, we must get a model that accurately predicts the fluctuations in the salmon population. Presently we are asking an agency of the government to make profound decisions without the proper tools. We urgently need a good model that considers all the necessary and complex factors involved in salmon survival.........continues...

Breaching dams won't save salmon  -- April 10 2000, Oregonian, Letter to the editor by Don Dodds.   
A March 26th commentary on breaching dams to increase salmon in the Snake River failed to discuss some powerful reasons why breaching those dams won't achieve that result.  First, the commentary did not consider ocean conditions, especially food supply that is essential to survival of Snake River salmon.
 A recent study found that returning salmon declined 90% over the last 30 years in the pristine, undammed and heavily environmentally reconstructed Keogh River in British Columbia. The study concluded the decline was due to a 30-year decline in food productivity in the North Pacific waters where these salmon mature. This is the same ocean area where the Snake River salmon mature.  In the light of this evidence, the construction of the dams on the Snake 35 year ago appears to be unrelated coincidence. By comparison, the prolific salmon stock in the Hanford Reach of the Columbia River feed farther north in colder waters where food supply is adequate. Other evidence shows that ocean conditions, rather than river conditions, are primary salmon survival factors.  For example, fish populations that rely on the same food sources as salmon but do not enter the rivers as salmon show a similar population decline as the salmon since 1883.  Why breach dams to try to increase salmon if the salmon starve to death because of inadequate ocean food supply?  Second, the commentary failed to take predators, especially protected predators, into account. The National Marine Fisheries Services- own data reveals that the populations of many predators soared while salmon populations declined.   For example, seals, sea lions and sea otters (which were slaughtered during the 1800s for oil and furs) increased from about 3000 in the early 1900s to about 650,000 today. Some bird predator populations have increased from a few hundred pairs to over 30,000.  Salmon predators, other than man, kill 99 of every 100 salmon. Those who advocate dam breaching to save protected salmon seem unwilling to make tough choices, such as reducing the population of one protected species to save another. It is hard to defeat the age old relationship between prey and predator.  Without predator control, any increase in salmon production will simply provide an increase in predators, and no net gain of salmon in the river.  Finally, after calling for unjustified dam breaching, the piece urges other drastic controls on human activity along rivers, vigorous habitat restoration, which will likely have little or no impact on salmon populations and may in fact decrease their biodiveristy. A paired river study in British Columbia compares two rivers, one river in disarray due to human activities to a second river where the habitat was been heavily restored according to the latest biological opinion. The biologically repaired river habitat is being out produced by the unaltered and supposed heavily damaged river.  Management of salmon over the past century has concentrated on the hatcheries and harvest while considering the ocean habitat to be an inexhaustible pasture. The success of this approach has been limited. Over the last 10 or 12 years, the idea that the ocean is constant and unbounded has been slowly changing.   In 1995, Congress directed that ocean conditions be considered in salmon survival.  Since that time, considerable evidence has been compiled that indicates the problem is not in the rivers and the current 4H (Habitat, Harvest, Hatcheries and Hydropower) approach will not solve the problem. Is there a need to rush into drastic and irreversible actions that will not save salmon? The tide of biological opinion is changing.  We can have it both ways, save the salmon and the dams.