This post was chosen as an Editor's Selection for ResearchBlogging.org

Many ecological systems have tipping points - thresholds where small changes in impacts can have very large effects on on ecosystem functioning, often in a bad way.  Lakes, for example, might show little impact from nutrient pollution until a threshold level is reached, and then massive algal blooms form that choke off many other species growth.

In the absence of knowledge of exactly how far one can push a system before reaching a tipping point, many invoke the precautionary principle, which states that in the face of uncertainty, one should take the most conservative approach.

There is broad disagreement about whether the precautionary principle is useful and exactly how to apply it.    In Ecological EconomicsNicholas Brozović and Wolfram Schlenker explore the economics of this challenge.   They build an economic model of a system where too much pollution causes a threshold change economic benefits, and then explore how the optimal behavior - that is, the optimal level of pollution - change in this system as uncertainty changes.  The unique contribution here is that they model two types of uncertainty, and they find some interesting things:

Systemic uncertainty emerges from the fact that ecological system are complex and multifaceted.  Even if we knew how much pollution, on average, a lake could absorb before it collapsed, this amount could change due to many other factors at play in the system (like the weather).  Sometimes this is called “irreducible uncertainty” because more knowledge of the system does not reduce it.

Image source: Flickr

It turns out that such uncertainty has a complex relationship with the optimal management strategy.  If there is some systemic uncertainty, it pays to manage with a “buffer” to avoid hitting the tipping point.   However, lots of systemic uncertainty reduces the incentive to be precautionary, because random, uncontrollable events are more likely to cause a collapse than one’s own actions.

Knowledge uncertainty comes from limited knowledge.  This is “reducible uncertainty” because learning can improve our knowledge.  This, too, has a complex relationship with strategy.  If one is far from the tipping point, more uncertainty drives precautionary behavior, as one becomes less sure about one’s margin of safety.  However, if one is close to the tipping point, lots of uncertainty can obscure the high-risk situation one is in.

So…it’s complicated.  Brozović and Schlenker make a very interesting observation about this complexity:  

disagreements in some environmental disputesmay be a result of different beliefs about threshold uncertainty, even when there is broad consensus about underlying processes and system dynamics. For example, some regulators argue for an immediate reduction in the production of a pollutant until uncertainty about the underlying natural process is reduced while others suggest that no costly reductions should be undertaken until the same uncertainty is reduced. Our analysis suggests that there may be much more common ground between these two views than might otherwise be thought. 

I would think this has straightforward applications in stakeholder relations or negotiating strategy: parties’ assessment of uncertainties in the system can strongly influence their actions.

I see a couple of limitations in this work.  First, while the authors explore the uncertainty of when ecological thresholds are reached, they do not examine the uncertainty of how big a change occurs at the threshold.  It’s rare that we know how bad the changes will be before a sytem collapses.

Second, they model a reversible shift.  They acknowledge that some shifts are irriversible and leave exploration of such a system to other analyses.  However, there is a continuum of dynamics between reversible and irriversible systems, notably those with hysteresis, where considerably greater change is required to return a system to its original state than the initial change.  

ResearchBlogging.org Brozović, N., & Schlenker, W. (2011). Optimal management of an ecosystem with an unknown threshold Ecological Economics DOI: 10.1016/j.ecolecon.2010.10.001