Holly Doremus on Adaptive Management

by Noam Ross, 20 April 2012


Holly Doremus, J.D., Ph.D., visited Davis on 4/20/2012 as part of a seminar series on interfaces between ecology, economics and policy. Dr. Doremus is at the UC Berkeley School of Law, but has Ph.D. in Plant Physiology from Cornell. She is former former UCD Faculty.

These are my notes from Dr. Doremus’s lecture. Any errors or misrepresentations are my own.

Adaptive Management – Knowing When to Say When

Adaptive management is thought of as a solution to two problems:

  • The science problem - What do we know and what can we learn about managed systems and the consequences of our actions?

  • The policy problem - Given what we know and we don’t know, what should we do?

It has some potential downsides, though:

  • Confusion over definitions and goals - People ascribe different meanings to the term, and it becomes more symbolic than a way to confront the difficult parts of their job. (Sounds a lot like resilience to me)
  • Resource demands and extended controversy - It requires more time and effort by the agency. More constant monitoring, modification of decisions. This can have political and morale/psychological costs.
  • Short-term/long-term trade-offs - Learning may have to be put ahead of what might be the optimal short-term decision
  • Accountability versus flexibility - Agencies do not necessarily respond to the public interests, but respond to political and other pressures. Accountability mechanisms are needed to get agencies to use adaptive management in ways that are in the public interest.

The last of these is a principal agent problem. The objectives of an agency will differ from those of the principal (legislatures, the public), and there need to be ways to force alignment. For instance, lawsuits to enforce the Endangered Species Act are such mechanisms.

Much of this arises when agencies are given multiple goals that have trade-offs among them. For instance, the public lands agencies are told to manage lands for multiple uses so as to best meet the public interest, and all accountability is through the political process.

Adaptive management is becoming the de facto standard for natural resource management, but when it’s legislated or prescribed, there is little conversation as to what it actually means, whether it is the best strategy, and whether the value of learning outweighs the costs.

Courts rarely interfere with scientific decisions and agencies. This is, in part, because lawyers and judges are culturally averse to scientific topics. It is also because of the belief that legislatures have created these agencies so as to outsource technical decisions.

Is adaptive management the right choice?

  1. Is action inevitable or essential, and why?

    Is it feasible to delay action? Adapive management should not be used to justify ill-considered action. For instance, agencies can grant a take permit under the ESA, which normally requires proof that there would not be harm. Adaptive management can be used as an excuse to provide such a permit without proof.

    This may not apply if some action is inevitable but several options may be available. For instance, salvage logging may be better or worse that the risk of large fires for old-growth forests. Adaptive management might be a good tool here.

    Of course “inevitability” may be political.

  2. What are the information gaps, and what difference will filling them make to management?

    Articulate a model of the system, including uncertainties, goals, and metrics. Distinguish genuine unknowns from information that just isn’t shared. Often there is information that is in the hands of landowners, etc., which might be demanded, but adaptive management is used as an excuse to avoid doing so.

  3. How likely is significant learning, on what time frame, and at what cost?

    Need to be honest about what experimentation is feasible and likely to be politically palatable. What information will it provide? What information can be gained from other approaches (modeling, observational science)?

  4. Can management be adjusted as new information becomes available?

    Will impacts be irreversible? How “sticky” are policy decisions, legally or politically? Can decisions be made incrementally?

    Adaptive management has part of its premise that policy choices are iterative. However, some decisions are irreversible. Because of pressure to make policy decisions as final, many decisions are effectively irreversible, and no action is a better choice.

The value of explicit analysis

Rarely are the questions above answered in real management situations. They ought to be, because it will help to:

  • Clarify goals
  • Help combat hubris and the tendency toward over-optimistic assumptions
  • Increase accountability
  • Inform budget decisions and frame expectations
  • Highlight removable barriers to learning

NEPA and other legislation requiring explicit analysis of impacts have been useful; adaptive management should not be allowed to erode such approaches.

Tailor the strategy to the problem

Where goals are multiple and potentially conflicting, adaptive management is likely to be abused (e.g., the “co-equal goals” of the Delta Stewardship Council). Priorities and approaches to making trade-offs must be defined first.

Ensure accountability and enforceability

  • Define triggers and responses - before taking action
  • Allow citizen enforcement - e.g., lawsuits. This can can actually give agencies cover for politically difficult decisions.

Reducing barriers to learning

  • Find substitues or mitigators for risky experiments
  • Take advantage of “natural experiments”
  • Remove budget disincentives - in the USFS and other agencies, the management and research budgets are distinct, which hampers learning
  • Facilitate information sharing and diffusion

Promote systematic learning

  • Make models explicit
  • Build in systematic monitoring
  • Recognize that data is not knowledge - Monitoring alone just generates data that sits on a shelf. Monitoring plans must include a plan to analyze, disseminate, and act on findings.
  • Create positive career incentives.

More from Dr. Doremus on the Berkeley Blog