The Salmon Decline

The Decline of the Northwest Salmon?

This network of dams provides about 2/3's of the regional power requirements of the Pacific Northwest. Recently, however, hydropower has become disfavored due to its possible impact on salmon and other migrating fishes.

The apparent decline of the salmon in the Columbia river system is one of the most complex environmental problems we have.

However, this problem has been over-simplified by the media and others to one single root cause, namely the construction of hydroelectric dams on the Columbia river system.

Most people believes Dams are the root cause of the Salmon Decline. They cite the following data as evidence:

This graph shows an apparent decline of the pounds of fish that have been caught (harvested) in the Columbia river. You should think critically about this data to see if it really is relevant to the more general problem of hydropower induced decline of the salmon.

In fact, the most recent salmon count data now show unprecedently large numbers of salmon returning.

We wish to access of this simplification is valid by examining some of the broader issues and looking at the actual data which is available.

The Salmon controversy is an excellent example of arguments made largely on belief, instead of knowledge that is supported by the actual data.

As the dams are obvious targets of opportunity this belief system can be supported. But can it really be supported by the data? .

It is also an excellent example/reflection of another problem that effectively prohibits meaningful discourse; reactive vs proactive positions

Let's look at some data. Most of this data comes from fish counts done at Bonneville Dam starting in 1938.

Blue line represents growth of non-harvested fish in the Columbia River system. The implication is that competition for food/nutrients is increasing rapidly.

Its getting warmer earlier in the season with time in the Columbia river. Is this cause local (e.g. dams/ag) or global (e.g. Global Warming)

Graph shows harvesting by Tribal communities with protected fishing rights. Some degree of "over harvesting" is apparent.

Graph shows total salmon catch off the coast of North America (including Alaska).

Decline of In-River catches which is frequently used as the "best" indicator for overall salmon decline in the Columbia River System:

Implication is that increases in ocean catch are directly responsible for decreases of in-river catches also shown.

The Salmon Population

Factors Causing Salmon Decline:

Salmon Decline Matrix

Problem Dams Logging Changing
Agriculture Pop. Growth Over Fishing Other
Salmon Migration Barriers Big Timenononononono
Barriers to Reaching Spawing Grounds Yes noYesnonoYes!no
Reduced Number of Young Salmon reaching seayes yesnonononobird
Pesticide Exposureno YesnoYesnonono
Industrial Pollutants Exposureno nononoYesnono
Loss of Streamside Vegetation and Functionsno YesnoYesYesnono
Increased Sediment Loadno YesYesYesYesnono
Natural Habitat Destructionyes yesnoyesyesnono
Reduced Fresh Water Flow in Rivers and Streams yes nonoyesyesnono
Abnormal water temperaturesyes noyesyesnonono
Reduced upwelling and nutrient/food supplyno noyesnononocompetition with hatchery fish
Estuary Degradation no YesnonoYesnono
Loss of cool deep pools in streamsno Yesnonononono

Qualitative Scorecard:

Abstract of a Recent Technical Paper

Evidence of a Nutrient Deficit in the Freshwater Systems of the Pacific Northwest. A comparison of historic and current levels of salmon abundance was used to determine the current deficit of marine-derived nutrients for Pacific Northwest streams. This nutrient deficit may be one indication of ecosystem failure that has contributed to the downward spiral of salmonid abundance, further diminishign the possibility of salmon population recovery to self-sustaining levels.

This is the raw salmon count data at Boneville Dam that we will inspect more closely below. Evidence of quasi-cylical behavior is apparent. Generating count Data (last year and this year are high!)

Some relevant WWW resources:

White Men Luv Fish

Pacific Northwest Index

Canadian Harvest Increases

Some Salmon Links

The above documents and data strongly suggest that a significant reason for Columbia river salmon decline is due to increased harvests in Alaska and Canada.

This is likely a reflection of where the Salmon are and this is strongly coupled to decadel climate changes in the Alaska-BC-PNW region.

This implies, that at any particular point in this region, over timescales of 50 years, salmon counts/catches will be cyclical in nature.

World Salmon Harvest

What about the viability of Dam Breaching?

Pros and Cons of Dam Breaching:

Arguments For:

  • restore Salmon runs
  • restore river to its more natural state
  • create short term jobs (sledgehammer crew).

Arguments Against:

  • Loss of Power
  • Loss of flood control
  • Loss of Irrigation
  • Loss of River Navigation
  • Increased Turbidity
  • Toxic Waste movement down the river

How we can gather objective data on each of the items in the list.?

First we noticed that the listed items under Aguments against are quite obvious and really don't require justification. All of these things would happen, it just matters to which degree they would happen and this is very difficult to model.

So let's concentrate on the other items in the Arguments For column.

The principle hypothesis here is whether or not dams have caused a "catastrophic" loss of population.

We can use salmon count data as a means of establishing a baseline. We can then convolve that data with the following simple model:

X ----------|-------------|--------|------|--------|----|---|--|->  X'

            1             2        3      4        5    6   7  8

Suppose we have a system where X salmon enter and have to pass 8 dams to reach X'.

The issues are:

These hypothesis can be tested to some extent by the actual salmon count data which exists at each dam. That data is readily available .

But we aren't doing an indepth analysis here. Let's keep it simple.

Under the model where all dams are equal then we should see something like this in the data.

X1 < X2 < X3 < X4 < X5 < ...

While there are complicating factors, this basic pattern should be defined by the data on upstream salmon counts. It is not.