Attending the conference hosted by the organization in Portland, Oregon last week — I was struck (again) by the huge dependence of scientists on mathematical and computer modeling in trying to ‘understand’ salmon. An entire day was spent — in a dimly lit room — watching PowerPoint after PowerPoint presentation of various equations, models, and graphic representations suggesting we “know” about salmon.

The image below is not one of them… it’s my sanity-finder:

One equation was so damn long that I swear it took a couple of PowerPoint slides to show it all… And the equation apparently had all of the various “factors” affecting salmon survival– the freshwater environment, estuary, ocean, and so on and so on.

Yet… to “know” about salmon, and to model salmon populations, we need to make “assumptions”.

It is this type of language that peaks my curiosity. See, assumption, has a few definitions. There’s the one that I use in my marriage — as in don’t assume, it makes an “ass” out of “u” and “me“. Generally, assumptions in a relationship are not a very good thing.

Dictionary definitions suggest:

The act of taking possession or asserting a claim; The act of taking for granted; Something taken for granted or accepted as true without proof; a supposition; presumption — arrogance.

Now there’s no question that there is no shortage of arrogance in science… it was not in short supply at this particular conference, and other scientifically-slanted events I’ve had the pleasure of attending… However, when it comes to modeling, or ecological modeling, there are things that must be taken for granted, or accepted as true without proof.

Without proof… that sounds so… so… un-scientific.

Now various definitions of ecological modeling suggest that these are meant to “simplify complex foodwebs“. Ok, now we’re getting somewhere.

Wikipedia continues with some enlightening thoughts:

Ecosystem models are a development of theoretical ecology that aim to characterise the major dynamics of ecosystems, both to synthesize the understanding of such systems and to allow predictions of their behavior (in general terms, or in response to particular changes).

[hold on to this thought of ” response to particular changes” for a few more paragraphs..]

Because of the complexity of ecosystems (in terms of numbers of species/ecological interactions), ecosystem models typically simplify the systems they are studying to a limited number of pragmatic components. [my emphasis]

Uh, huh. Simplify to a limited number…

The Wikipedia definition continues by suggesting there are various factors driving the simplification process inherent in ecological models:

ignorance:

while understood in broad outline, the details of a particular foodweb may not be known; this applies both to identifying relevant species, and to the functional responses linking them (which are often extremely difficult to quantify)

computation of complexity:

practical constraints on simulating large numbers of ecological elements; this is particularly true when ecosystem models are embedded within other spatially-resolved models (such as physical models of terrain or ocean bodies…)

[oh yea… like salmon, maybe?]

and limited understanding:

depending upon the nature of the study, complexity can confound the analysis of an ecosystem model; the more interacting components a model has, the less straightforward it is to extract and separate causes and consequences; this is compounded when uncertainty about components obscures the accuracy of a simulation.

[hmmm… maybe like wild salmon that spend lives in gravel, freshwater, estuaries, the North Pacific, estuaries again, fresh water again… and so on, and so on]

“Uncertainty… obscures… accuracy”… important points to ponder.

_ _ _ _ _ _

On day two of the conference, the agenda suggested that we would get into the: Human responses to hatcheries: understanding the social, cultural, legal and economic dimensions.  (of hatchery and wild salmon).

One of the panel presentations discussed an: “Economic analysis of a Columbia River fish hatchery program”. One of the most stunning comments to come out of this presentation was:

…it is impossible to economically model social and cultural impacts…

Thus, despite all of the various economic metrics, modeling, graphing and equations as part of this study and thousands of others… “it is impossible to model social and cultural impacts“.

During the question period, I asked the obvious question: “ok, how do we measure and deal with the social and cultural impacts then?”

There was no answer of substance from scientists present — to this thorny issue…

Is there an answer?

_ _ _ _ _ _ _ _

On the afternoon of the second day, the famed “break-out” sessions came. I attended a sure-to-be-interesting session involving northern British Columbia and Alaskan representatives — two regions that have very different approaches to salmon hatcheries.

I won’t regurgitate the entire conversation here; however, some curious points:

Alaska pumps out billions of baby salmon from around southern portions of the state in salmon ranching programs run by non-profit (cost recovery) operations. These ranching (hatchery-like) operations continue despite the fact that many Alaskan government reps suggested most rivers naturally get enough spawners to support salmon populations.

It’s strictly an economic enterprise — not an ecological (even though there is an ecological impact).

British Columbia pumps out about 600 million baby salmon — some for ecological reasons, but most for economic.

Much of the discussion in this session, and much of the conference, surrounded “carrying capacity” of the North Pacific. What sort of ecological impact is occurring as a result of sending out over 5 billion baby salmon from hatchery/ranching operations around the North Pacific?

I kept asking the obvious question to me: “what is the carrying capacity of the North Pacific?… It sure as heck has seen a lot more baby salmon in the past then it does now…”

The answer quickly spouted by many scientists around the table was “we’ll never really know… it’s impossible to model… it’s too big…”

But then that would be followed up by comments such as: “it’s sure not what it used to be”… “it’s way less now”… “climate change is affecting it”.

Oh, OK, well if climate change is impacting it… how much is it impacting it — the carrying capacity? What are the effects of climate change? How are these effects changing the carrying capacity?

“Well… we don’t know… we’ll never really know… it’s impossible to model… it’s too big…

however, we are seeing some of the impacts in ocean acidification…”

I pointed out, that from what I have read, all the various models used to predict ocean acidification rates were way off, way wrong. We are already seeing rates that were not anticipated until at least 2050 (according to the models).When it comes to wild salmon, ocean acidification can be devastating, as the acidification will dissolve the shells of little critters like copepods that are essential food sources for baby salmon as they head out to sea.

So, OK, if ocean acidification can be devastating to salmon; is happening faster than expected — how can we get a better understanding of what the impacts might be?

you probably know the answer… “we’ll never really know… it’s impossible to model… it’s too big…”

_ _ _ _ _ _ _ _

Remember that piece from above though, about “response to particular changes”:

Ecosystem models are a development of theoretical ecology that aim to characterise the major dynamics of ecosystems, both to synthesize the understanding of such systems and to allow predictions of their behavior (in general terms, or in response to particular changes)

If we don’t understand:

1. Carrying capacity of the North Pacific
2. Effects and rate of climate change, and thus also
3. Effects and rate of ocean acidification

And we can’t model the social and cultural impacts of losing wild salmon, of building thousands of hatcheries, and of building hundreds of salmon farms along wild salmon migration routes.

WHAT THE HELL ARE WE DOING?

Sanity-finder #2: