Adaptive Management

Lead, Marine Conservation Initiative at the Gordon and Betty More Foundation and former Chief of the Grand Canyon Monitoring and Research Center

“We don’t have a high tolerance for recognizing that we need to make long term investments
in managing our natural resources. Everyone wants to fix it and move on.”

Inspired by a 10^th grade biology teacher who instilled in him a passion for the process of learning, Dr. Barry Gold went on to become one of an emerging breed of environmental scientists who bridges the gap between science and policy. As Chief of the Grand Canyon Monitoring and Research Center (GCMRC), Barry was the science lead for the Glen Canyon Dam Adaptive Management Program an ongoing, 10+ year effort[1] to provide a process for the cooperative, science-based management of dam operations, in a way that integrates dam operations and downstream resource protection based on monitoring and research information.

Barry has extensive experience working at the interface of environmental science and policy. He has held senior positions at the National Academy of Sciences, U.S. House of Representatives and the Department of the Interior. He is currently Lead for the Marine Conservation Initiative at the Gordon and Betty Moore Foundation. He joined the Moore Foundation from the David and Lucile Packard Foundation where he led the development of a new science for oceans and coasts program. Barry has a DSc from Washington University, an MA from George Washington University, an MS from the University of Connecticut, and a BS from the University of Miami.

Many Leaf Litter readers are unclear of the definition of “adaptive management.” How do you define the term? Is it possible to do so in just a few sentences?

Definitions of adaptive management range from passive to active adaptive management. I’m on the end of the spectrum of those who believe that many of the activities that people call adaptive management aren’t. For me, adaptive management is where you propose a management action in the context of an explicit hypothesis that you’re trying to test from the outset. That is, you want to know if your management action is going to produce a predicted result. So, you implement the management action with an appropriate monitoring framework, to ensure that you’re measuring the variables that will tell you whether or not you are having the desired effect. Finally, you use the monitoring data to tell you if you need to make a change in the management action.

How did you come to be a believer and proponent of adaptive management (for use in) natural resources?

I had read the [Carl] Walters and [Buzz] Hollings material in the early days and theoretically, it made good sense to me. Then I used their adaptive environmental assessment and modeling framework for work that I did in Ghana in the mid-80s trying to understand how to manage an artisanal fishery. Just the modeling step, developing a conceptual model of the system, turned out to be very powerful – the notion that you could work with people who didn’t understand equations and create a model that incorporated their implicit assumptions and understanding of how the system worked and how it might respond to different interventions. So my first interest was simply in using the systems thinking approach that is part of adaptive management, as a means of creating a shared understanding between resource scientists and resource users and managers of the dynamics of the system they were all working with. (update on the fate of fisherman in Ghana)

The second piece came from my time in the Congress and then at the Department of the Interior, realizing that many people were seeking a silver bullet to address natural resource management issues and that there really is no silver bullet. Managing natural resources is a long term game, and if we are going to be in it for the long term, we need to do it in a way that is smart. To me that means a collaborative, science-based approach and that’s adaptive management. When I had the opportunity to go from being a theoretician to taking the job as the head of the biology program at the GCMRC, I jumped at the chance to try adaptive management on the ground. After I was at GCMRC for two years I became the Chief, and that experience convinced me that it’s probably the only way to go to successfully manage natural resources.

Tell me more about your work with fisherman in Ghana.

This was actually my dissertation research. I was trying to develop a model for managing a small scale artisanal fishery and link resource users and resource managers.  My major advisor was an early pioneer in what we now call sustainability research. He was interested in what he called “bioresources development,” how one could use biological resources in a way that meets peoples needs without substantially degrading the environment. My research had all the typical challenges: not a lot of data; people who relied on the resource; a very diverse population; and the challenge of getting stakeholders and managers to agree on the nature of the problem and the types of interventions to try. It was about a five year-effort. In the end, it really wasn’t successful. At that point in my life, I was really focused on the biological science, so I focused my efforts on building a population dynamics model for this fishery. But, I learned after the fact that it was the social dynamics that drove the level of effort applied to the fishery, and I didn’t really figure out how to incorporate that into the management experiment. Nonetheless, the fishery was managed pretty well for a while. I don’t really know the status of it today, I do know that some of the folks I worked with at the Fisheries Management Research Bureau in Ghana went on to get their own graduate education.

At what point when you were working on the Glen Canyon Dam Adaptive Management Program (GCDAMP) did you learn that adaptive management was the “way to go?”

Some people would argue that the GCDAMP has yet to demonstrate that it’s working. One big challenge in natural resources management is the lag time in response of the biological resources to management actions because the biological response has to cycle through the life cycle of the organisms we are interested in managing. I think the GCDAMP has demonstrated success in terms of the collaborative, science-based process that is being used and the response of the physical resources, to the management actions that have been tried to date.

By using adaptive management to implement a series of experimental flows, built around structured hypotheses, coupled with explicit monitoring protocols, GCMRC developed a solid understanding of how the hydrology and geomorphology of the Colorado River ecosystem in the Grand Canyon functions to create habitat. This understanding is 180 degrees from what was believed before the first management actions/ experiments were implemented in 1996.

The first experiment in 1996 was designed based on current knowledge of how scientists thought the system worked. Scientists assumed they knew where the sediment was stored and what would happen to that sediment if you ran an artificial flood. The goal of that experiment was to build beaches and create habitat. Even though it was initially celebrated as a success, scientists learned from the monitoring and research data obtained from that experiment and subsequent experiments that if one continued to run those kinds of management actions, you would cause harm to the physical resources. That was a major lesson. There had been ten or twelve years of research that led up to the ’96 flood, and then scientists and managers learned that it didn’t work at all the way they thought. That was the real learning point for me, because if we had said, “we have 10 years of science and we know what management action to implement, without embedding it in an adaptive management framework” we wouldn’t have done any of the monitoring and over time we would have degraded the resource without understanding why. So that is my definition of success.

On the biological resources side, I would say GCMRC has identified some parameters that need to be considered that weren’t being looked at before. They have a much better understanding of the ecology but still haven’t turned things around. For example, they know much more about the population trajectory of the humpback chub, and a better set of hypotheses for why the trajectory is going the way its going.  It’s the learning that occurs through using an adaptive management process and the dialogue created among scientists, managers, and resource users, that is the real hope, not necessarily that one has achieved the desired result in the short term. That’s one of the major challenges to adaptive management. The manager is faced with pressure from stakeholders and policy makers to go in and have a positive result right now. When one is not sure why degradation is occurring, which is the case for many of the natural resource challenges we now face, adaptive management is the way to go.

We asked our readers whether they thought there were specific types of projects for which adaptive management is most appropriate. Many believe it is appropriate for all types of projects in our work. Some believe it is most appropriate for particular types of projects such as invasive species management, stream restoration and wetland restoration. Others offered parameters, such as “projects where the problem is very site specific and not easily addressed using old regulatory approaches.”  Do you think there are specific types of projects, ecosystem, etc., where adaptive management is most appropriate and places where it’s definitely not the right approach?

I think the hypothesis testing framework and conceptual modeling parts of adaptive management are applicable in most places. The question is the level to which you need to do these activities and the resources you have available.

If everybody can agree on the restoration end point – let’s say you’re restoring a wetland and you’ve done this a hundred times in a very similar habitat and context and it’s been successful – then I’m not sure I’d go in and do a full blown adaptive management process. However, I would make sure that I know the minimum set of data that needs to be monitored on a regular basis to evaluate if the restoration is achieving the project’s goals over time.

One would make a more intensive investment in an adaptive management process for a more complex problem where there is not a clear sense of the right set of actions to take. Then, through implementing management actions as experiments, you narrow your solution set over time. A conceptual model will also help you focus on the key parameters you need to monitor. If you really embrace an adaptive management framework, you’re going to always want to have monitoring data on how well the system is doing – even after you leave. Say you’ve restored a prairie, and you’re not paying attention and there’s some natural or man-made event that occurs. Then all of a sudden the site is overrun with invasive plants. There’s got to be some sort of ongoing monitoring so one can try to understand the dynamics of the system.

74% of our readers said there were occasions when they wanted to use adaptive management but were unable to do so. Most cited “unwillingness to commit to a long-term process” or “general lack of understanding of adaptive management” as the primary obstacles. What do you see as the primary obstacle to widespread use of adaptive management?

I think the biggest obstacle to adaptive management is that the words “adaptive management” mean too many things to too many people, so they don’t really define a unique approach anymore. This was true even in the Glen Canyon example. The GCDAMP grew out of an environmental impact statement (EIS) process. In the course of doing the EIS, which normally leads to a preferred alternative, the various parties couldn’t agree on the preferred alternative because they had conflicting data and conflicting sets of assumptions about what would be the right restoration effort. So they included adaptive management as the implementation framework as a way of completing the EIS, not because there was a huge understanding of what adaptive management was. They said, “We can’t agree here. There’s this thing out there called ‘adaptive management’ where you implement what you think is the best of your alternatives and see what works. So let’s do that.”

Adaptive management has become everything to everyone and in my mind it has lost some of its power. It is not simply going out there, doing something and seeing what happens. It requires creating a conceptual model that clarifies ones assumptions about how a system works, and explicitly testing these assumptions through management actions designed as experiments with an accompanying monitoring and research program.

Modeling is another impediment to adaptive management. Adaptive management does not require complex models when you think about the whole spectrum of mathematical modeling, but it does require building a model that makes explicit one’s assumptions about how the system works and how it’s going to respond to the management action. The model is the thing that starts building the scientific framework within which you’re going to then test some hypotheses. A lack of expertise in systems thinking and model building can be an impediment. Also, you want to build the model with managers and stakeholders as well as scientists, and managing that dynamic is a challenge. Those communities don’t necessarily collaborate well. They speak different languages. However, we have learned that this can be an important part of building a strong collaborative relationship that is so important to success of these types of projects.  This is a key piece because at the heart of it, adaptive management is a collaborative, science-based process, and unless you build the collaboration into it, I don’t think you can move forward.

The third impediment to adaptive management is that it’s not a quick fix. Most people would like to do a stream restoration, for example, and say, “we’ve fixed it” and walk away. There are many cases where one may know enough about how to restore a system to a given end point that you can do that, but there are many more places where you’re going to want to implement your restoration or conservation activity, and then maintain a monitoring program. We don’t have a high tolerance for recognizing that we need to make long term investments in managing our natural resources. Everyone wants to fix it and move on.

Another challenge to adaptive management is the extent that people accept it as a scientific process since there are no controls. It’s hard to find two systems in a very similar state, go in and manipulate one system and leave the other alone, and then watch how they respond. So the issue of control is difficult. There’s also no replication. Once you’ve implemented one management action on a system, you can’t go someplace else and replicate it. So it’s challenging from a traditional scientific perspective. As a result of these challenges, one often designs the management action to be aggressive enough that you can see a result, and given the uncertainty around the likely response, it’s hard to get the OK to do that. Even in the case of the GCDAMP, scientists never quite got the flows they wanted. The Bureau of Reclamation, which operates the dam, the Fish and Wildlife Service, which has responsibility for endangered species, and the users of the power generated from Glen Canyon Dam all had reasons why the scientists couldn’t get exactly the experiments they wanted.

Where do you think we should start in changing people’s perspectives on adaptive management?

First, we need to include people. As ecologists and environmentalists, we need to recognize that people are part of the systems we are being asked to manage or restore, and recognize that the end point is probably not to restore the system to a “pristine” state. I was part of a group of authors who wrote a Policy forum piece that was published in Science called “Ecology for a Crowded Planet,”  in which we argue that ecologists and scientists need to think more about what our ecological end points are when people are part of the system.

Second, with respect to the policy, management and stakeholder communities, we need to help them understand the value of these systems. My hope is that the ecosystems services framework International Society for Ecological Economics or the Millennium Ecosystem Assessment  that is starting to take hold will result in meaningful valuation work that can help us all understand how the value of ecological systems change as they move along a spectrum from less degraded or altered to more degraded;

Third, we need to think about natural infrastructure the way we think about physical infrastructure.  For physical infrastructure – whether it’s roads or buildings, or utility systems – we recognize the need to make continued investment in the operation and maintenance of those systems if we want them to continue to function and provide societal benefits. We have to adopt that framework for natural systems; we have to be willing to make similar long term investments in the operation and maintenance of natural systems and that includes adaptive management. Once we get those perspectives aligned people will recognize that natural resource management is an ongoing commitment and abandon the quick fix notion, “that we can restore a system and walk away and move on.”

Finally, we need to train more people in the collaborative, science-based approaches, including conceptual modeling and monitoring program design, at the core of adaptive management.  I think this can be done through networks of practitioners and continuing education.  And, we need to learn to scale this approach to meet the needs of a given problem.

Adaptive management is thought to be one of the most inclusive approaches to Ecological Restoration, Conservation Planning and Regenerative Design. Do you agree?

I think it can be but I think it’s not always practiced that way. What has happened in many instances is that the technical experts have taken over. To really be successful, adaptive management needs to be a collaborative, science-based process. It’s critical at the beginning to engage all parties interested in the restoration or conservation action that’s being contemplated. It can be intimidating for resource users – let’s say farmers, when you’re thinking about some sort of riparian restoration – to sit in a room with a bunch of scientists and articulate their understanding of how a system works. From my perspective, unless you’re willing to develop a common understanding about the system – what’s wrong with it, how we think it works, how we think it will respond to various management interventions, and what will be measured and monitored – you’re not going to have the support over the long run for the project. This can be done in a good workshop, using simple modeling that takes maybe a day and a half of people’s time at the front end. This can really create the support that’s needed.

In a previous issue of Leaf Litter, in which we addressed the topic of stakeholder engagement, we spoke with Pamela Mang of Regenesis. She talked about how important it is to understand the history of a place, and how you can’t do that without engaging the people of that place, hearing their stories, and gathering “data” you couldn’t obtain otherwise.

As restoration professionals, engineers or scientists, our data is, hopefully, observer independent and fairly testable, so we somehow think our data is superior to others’. What we really want to do is objectively look at everyone’s understanding of how a system works and the assumptions that are implicit in that understanding. It’s important to create dialogue with parties of interest, people who have used or been around the resource for a long period of time and gather both their understanding of a system and the values they bring to the table.

Any suggestions on how to better educate ourselves and our colleagues, clients, and stakeholders so that people can begin to see that the added time, cost, and perhaps acceptance of uncertainty are worth it in the long-term?

Start simple. Your readers can probably look at a project that they’ve been engaged in over a long period of time and try to make the case that “Hey, we’ve been doing this for three, five or ten years, and here’s how it has changed over time, and we’ve done that in not as organized or rigorous a fashion as one might have.” Then, compare that to a process where you start with an adaptive management approach: creating goals, articulating values; doing a workshop to create a simple model — even at the level of a conceptual diagram[2] – clarify assumptions, define a set of hypotheses to be tested and design a monitoring approach.

The next step is to propose a set of management activities that test some of those assumptions, around which you’ve thought through what you are going to monitor. If it’s some sort of prairie restoration where you’re going to reintroduce fire, what is it you’re going to monitor after each burn? Do you have a good baseline before and after? One of the things that turns out to be important is being able to say to managers and stakeholders, “Here’s the data. This is what we thought would happen, and this is what happened.” That feedback piece is important. You’ve actually followed through and can produce something.

Are there any places within or outside of the U.S. where you are seeing this perspective changing on all sides? Is anyone ahead of us on this?

There are examples of people working together on various parts of it, and they tend to be on a more local scale. Steve Yaffe at the University of Michigan has a good web site. He calls it “collaborate ecosystem management” and he has identified a number of projects where people are working together on ecosystem management. They don’t necessarily embrace adaptive management. As you go up the scale there are also a number of federal projects. They tend to be mostly river-related projects. There’s one happening on the Trinity River. There is work starting on the Missouri River. There is also an effort where the Fish and Wildlife Service manages one of the flyways and the number of birds that can be harvested out of the flyway using an adaptive management framework that has been fairly successful. It’s not a restoration project, but they do at least have some models, some hypotheses about how much harvest they can allow, and on a regular basis they reexamine that in the context of data and set new targets. There is a lot of activity in Canada, and some in Australia.

Is there anything else you’d like to communicate to our readers about adaptive management?

People who have looked carefully at the largest project with which I’ve been engaged often ask is, “Why do you think adaptive management is so great? That project is still underway and you still haven’t recovered the endangered humpback chub?”  In my mind, it took a long time for the Colorado River ecosystem to get to the state it’s in, and it’s going to take a similar time for it to recover, and then it won’t be to the state of a pre-dam ecosystem. This is true in many other projects. So one has to be careful about the expectations you set up for using adaptive management. It’s not a silver bullet. It’s going to require a lot of hard work and active intervention. The real key is to share with people both the learning that comes out of the process and the real change that’s taking place in the resource. In the case of Glen Canyon, there were so many things that weren’t understood in 1996 that are understood today. That’s a real advance. The solution set for the kinds of management actions and what the GCDAMP can accomplish has been narrowed. In some cases, you’ll come down to understanding what needs to be done, and there won’t be the will to do it. That won’t be the failure of adaptive management. That just becomes a decision.