Science as a Way of Knowing, Forestry as a Way of Shaping Landscapes

Notes based on a lecture at the University of British Columbia, 12/97

Dan Binkley

Colorado State University

The most common view of Science in forestry is the application of the scientific method to solving specific problems. An example from the SCHIRP project on Northern Vancouver Island:

"the objective of this decade-long research effort was to understand the underlying causes of poor conifer growth…and to recommend the most effective methods…"

This is only a small corner of the domain of Science -- the rest of the territory has a much broader range, and more power.

In this seminar, I’ll discuss why Science is so powerful relative to other "ways of knowing," and ponder a few thoughts on why forest scientists and managers do not use the full power of science more often.

The world around us has events that we witness, and these events have causes. If we understand the causes, we have a chance to predict which events will happen when, and perhaps even to dictate what happens. The central philosophical problem is that different people come up with different explanations for the same events. Where ideas conflict, they can’t all be true. Science is unique in the method it uses for figuring out which ideas are wrong.

An example comes from the journal of a European visitor to Eastern Canada, 200 years ago. He observed that the Cree seemed to have a strange view of wildlife management:

"They kill all they can, having an incontrovertible maxim among them, which is ‘the more they kill, the more they have to kill’ and this opinion…is…pertinaciously held." (Umfreville 1790)

Most of us would disagree with this "knowledge", but how do we know if this model works better or worse than the views we hold? It sounds strangely reminiscent of the classical forestry idea that if we cut down all the old growth forests, we’ll speed up the rate at which new wood grows.

Many philosophical views have full, complete explanations for any event that occurs. The uniqueness of science is that a scientific idea must "forbid" certain events. If a "forbidden" event is observed (in nature or in an experiment), then the scientific idea is "falsified" and we move on. This falsification is the key that separates science from other ways of knowing, and the power of this approach is illustrated by Galileo:

"But in the physical sciences when conclusions are sure and necessary and have nothing to do with human preference…. a thousand Demosthenes and a thousand Arisotles would be left in the lurch by any average man who happened to hit on the truth for himself."

This may be a bit naive – few major truths have been hit on by "any average man." If we think that our knowledge should progress, we have to expect that some of our ideas are wrong. How can we tell which of our ideas are wrong? It may be useful to view Science as resting on 4 cornerstones: Ideas; Information; falsification/experimentation; and willingness to change views.

If any of these cornerstones is missing, or too small, then the edifice tilts and may fall down. Using this 4-cornerstone view of science, how has science been used most powerfully in dealing with forest issues? I’ll discuss 2 science case studies in this seminar before moving on to forest management:

Why does forest growth decline with age?

Will too much N from acid rain harm Sweden’s forests?

Why does forest growth decline over time?

The classic answer was that total production plateaued (or even declined a bit), while the accumulation of respiring biomass left little carbohydrate to allocate to wood growth. Here is the classic view from Kimmin’s Forest Ecology text:

This model has the first two cornerstones of science: ideas and information. The "net production" part matches the growth of forests all around the world. But what about the other 2 components of science? Until the late 1970s, no one bothered to experiment/falsify the ideas. Mike Ryan’s PhD thesis (with Dick Waring at Oregon State University) found respiration was not substantially greater in old forests. He concluded the pattern must result from some other ecophysiological process – time to abandon the old idea and develop (and test a new one). But old ideas live unless we’re diligent in discarding them…

I thought that nutrient supply might generally decrease with time, lowering leaf area, total production, and therefore net (wood) production. I tested this on Ryan’s chronosequence:

I found that nutrient supply did not go as predicted – net N mineralization increased with age. Fertilization of single trees led to strong basal area responses in all 3 ages. So my idea had to be rejected – but this study wasn’t strong enough to throw out my idea completely (it was not replicated, "stand age" was produced by nature, not randomized, etc.). How could we have a stronger test within a reasonable time?

In Hawaii, plantations of Eucalyptus peak in wood production at about 2 years of age, and then decline in increment. Although the peak comes early, it happens when trees are about 10-15 m tall, which is similar to temperate forests. Therefore, Eucalyptus plantations might serve as a "model" of how forests around the world develop. Given a warm climate year-round, it would also be a strong test of the respiration idea (respiration increases with temperature). Our experiments (with Mike Ryan, Jim Fownes, Randy Senock; unpublished data below) in Hawaii are continuing, but we’ve found that:

Increment peaked between 18 months and 24 months, following the trend in leaf area and light interception;

Leaf area and light interception remained high – but we found declines in wood production, belowground production, and total production (NPP and GPP).

Fertilization (4 times/yr!) increased leaf area and productivity, but could not prevent the decline in production.

It’s time to dismiss both respiration and nutrient limitation as universal explanations of declining growth in old forests. Nutrients are still important, of course (!), but if nutrient limitation commonly exists in all ages of forests, then it is not a candidate for explaining an age-specific pattern.

Another idea tested by Holly Barnard came from Barb Bond and Mike Ryan -- tall trees may not be able to conduct as much water in mid-day as short trees. If the resistance to water flow in xylem is the same across age, then a tree that is twice as tall can conduct only half as much water per unit time (given the same soil moisture and vapor pressure deficit). This simple, first-principles idea predicts that tall trees will have a mid-day shutdown of stomata when short trees could continue photosynthesizing. The various tests of this idea (in lodgepole and ponderosa pines, and in southern beech) were supportive -- but Holly's data clearly refuted the idea as an explanation for growth decline rates in Eucalyptus in Hawaii. So, is the theory of hydraulic limitation on tree growth dead? Not quite -- it's still reasonable to see if it fits in some cases (even if it's not universal), and it's still reasonable to test for whether it might apply once Eucalypts are more than 30 or 40 m tall.

A Great Swedish Hypothesis

The deposition of nitrogen (N) from the atmosphere can have an acidifying effect on ecosystems. This is a rather complex issue, but conceivable consequences include lower soil pH, and greater leaching losses of calcium, potassium, and magnesium. Some people think this may lead to declining health of forests. In Sweden, some modeling by environmental scientists led to predictions of a 20% growth loss within a few decades, and severe mortality on 2-4% of the forested landscape. How can we decide the amount of confidence this model warrants? If we think it’s bogus, how do we convince others to agree with us? Peter H` gberg and I were funded to prepare an assessment. Should we look for logical flaws? Read the dozens of papers that have overwhelming details of greater or lesser strength? Argue with the experts on the basis of relative professional reputations?

The key to good science is remembering the 4 cornerstones, especially the experimentation/ falsification stone. One should judge an idea by what it forbids; if it forbids many things, then it’s a good scientific hypothesis. If good experiments fail to produce forbidden events, then the idea warrants more confidence. In this case, the modelers made a very clear, strong prediction that forbade many observations: forest growth declines strongly if the ratio of base cations (Ca+Mg+K) to aluminum drops below about 1:1. We expected that forest fertilization (with N) should reduce this ratio, yet N fertilization typically increases forest growth in Sweden. So we plotted the response in the cation ratios for two fertilization studies (at StrD san and Skogaby) in relation to forest growth – and found that the results jumped strongly into territory "forbidden" by the model. Therefore, we know the model is invalid, without needing to challenge all the incredible details of the ideas and data that went into it. Nifty!

Applying Science to Shaping Forest Landscapes

Much of the science applied in forestry focuses on technical problem solving, and that works very well. However, if we recognize the occasions when old ideas are wrong, we’ll have more power in shaping landscapes. Sometimes, this opportunity arises when we search for the right answer to a question that is simply wrong. An example from a couple decades ago in BC: forest sites were classified by the availability of K+Ca+Mg, presumably because the scientists thought these elements would limit forest productivity. So they set out to determine the best method for soil analysis. They collected new information by sampling 42 soils, analyzing them by various methods, and relating the indices of cation availability with site productivity. They found that all indices of cation availability correlated quite poorly with site productivity, so they concluded: "We recommend the NaOAc pH 4.8 extraction method." Why didn’t they go on to toss aside the idea that base cation supply limits productivity? (It could be that they did not have strong confidence in their experiment, but I think they just weren’t looking for errors in their view of the world.)

Twenty years ago, I was the teaching assistant in a forest hydrology course at UBC. One of the exercises asked 4th-yr students to decide where to build a road for a proposed clearcut. The correct calculations showed the forested slope was already unstable; cutting trees would guarantee a landslide. Where did the students chose to put the road?

None recommended a mid-slope road, as this would be expensive and would trigger a landslide.

37 students recommended a valley-bottom road, as it would be cheap, and could be rebuilt after the landslides.

3 recommended building the road along the ridgetop, which would be very expensive but safe from the roadslide.

Just one student recommended that the plan to clearcut the slope should be reconsidered.

I’m sure the thinking of UBC foresters has broadened since then, and that they would have a broader view of applying knowledge to forest management.

Asking the right questions

The best use of science in forestry needs to be developed from the base of asking the right questions. Good science addressing the wrong question isn’t likely to be rewarded. For example, a research organization might ask one of these alternative questions:

"How can we achieve successful regeneration of Douglas-fir after clearcutting in the Coastal Western Hemlock Biogeoclimatic Zone?"

"What are the biological and silvicultural impacts of alternative forest harvesting methods?"

The better question depends on who owns the land, and what basis they need to make decisions about the future shape of their landscapes. The forestry establishment in BC was left out of planning process for the Clayoquot Ecosystem Management plan on Vancouver Island – foresters were thought be largely irrelevant to strategic planning in forest management, only useful for implementing decisions made by other people. Were foresters relegated to mere technicians because the forestry establishment tended to seek answers to the first question, without asking the second?

Main points:

Routine science provides technical insights and solutions.

The full richness of science comes in to play when we’re ready to figure out where we’re wrong about reality. Falsification can be easy and decisive, if we do it creatively.

Forestry harnesses science to accomplish social goals -- irrelevant science (and management mindsets) can be fun, but that’s the path to extinction!