Adaptive Resource Management in Complex Systems

Vol. 26 No. 3

Mr. Kornfeld is a Giordano Fellow at the Faculty of Law, Hebrew University, Jerusalem, Israel.

What are the systemic implications of saving endangered species, reallocating fresh water, or limiting emissions of carbon dioxide into the atmosphere? Regulatory agencies and environmental and conservation communities can generally provide qualitative assessments. However, deeper critical analysis is required to more accurately determine the extent to which efforts to affect these ecosystems are working and to diagnose why some actions succeed while others do not. In recent years, there has been a convergence of thinking among regulatory agencies and environmental and conservation organizations as to how to better assess, plan, and implement these actions. One method that has emerged is adaptive resource management (ARM), also known as adaptive management. ARM is a structured, iterative process of optimal decision making in the face of uncertainty. It is aimed at reducing ambiguity over time via systematic monitoring, focused on one or more resource objectives, that accumulates data required to improve future management. Ecosystem adaptive management is based on a learning process designed to improve long-run management outcomes. However, the ARM process must meet the challenge of achieving the proper equilibrium between enlarging one’s knowledge base and making governance decisions aimed at achieving an optimal short-term outcome based on contemporary knowledge.

ARM uses a number of tools in addressing a resource system. These tools, scientific and social processes, are the vital components of the ARM process. They include: (1) management linked to appropriate temporal and spatial scales; (2) management focused on statistical power and controls; (3) computer models to build synthesis and an embodied ecological consensus; (4) ecological consensus-building to evaluate strategic alternatives; and (5) communication of alternatives to politicians for negotiation of a selection.

If ARM is done correctly and transparently, particularly by government agencies, the democratic process is the winner, since the methodology is open to all stakeholders. Indeed, when a given objective or goal must be accomplished, and the ARM process is open to all of the impacted stakeholders—each of whom is subsequently represented at the table—the collection of ideas and approaches is filtered through public discourse, with the remaining ideas incorporated into the solution or remedy. Accordingly, in order to be successful, the ARM process needs to maintain political open-mindedness, create an open and cooperative process, and rationally employ the scientific method in its decision-making analysis.

This article defines the ARM process and provides a historical framework for its development, organization, and process for cooperative integration of natural resources conservation. The article will then address the methodologies central to developing new ARM institutions and strategies and to harmonizing scientific hypotheses and resultant field studies. Finally, this article will focus on the ARM process in practice, using as examples the Bureau of Reclamation’s monitoring and research information for the Glen Canyon National Recreation Area and Grand Canyon National Park, and the Everglades ARM.

What Is Adaptive Resource Management (ARM)?

When resource management agencies, such as the Bureau of Land Management, the Fish and Wildlife Service, or the Bureau of Reclamation, are required to make regulatory or policy decisions, they have employed a multiplicity of approaches that encompassed diverse levels and types of scientific data, political, economic, and social concerns, and historical precedence. However, these time-honored decision-making approaches frequently failed to realize ecosystem management objectives for complicated problems in large, complex systems such as the Colorado River Basin or the Everglades. In contrast, the July 2, 2011, ruptured Exxon-Mobil pipeline underlying the Yellowstone River in Montana demonstrated that conventional decision-making approaches can be comparatively successful at tackling relatively uncomplicated problems in discrete ecological systems.

Complex systems, however, require an approach to resource management that is grounded in and admits of uncertainty, such as the modeling of an aquifer or analyzing climate change. First used in the 1970s, ARM is a methodology for dealing with complex systems like ecosystems. In essence, ARM involves an iterative process whose practitioners recognize that in order to understand, manage, and regulate a complex environmental or resource system, one must fully understand it (i.e., gain a holistic knowledge about a system and comprehend its responses to various stressors or inputs). By holistic, I mean the aggregate of the variables or properties that make up a particular system (e.g., a watershed, which includes the physical, biological, chemical, economic, social, sustainable biodiversity, and other components). Indeed, the system cannot be determined or explained by its component parts alone. Rather, the sum of the properties and their interrelationships determines how each part behaves. Measuring and understanding a system’s responses to changes in variables or properties over time decreases the uncertainty related to future management decisions. It is akin to building a case, where every new piece of evidence allows one to build a clearer and more realistic picture of what occurred.

Learning is not simply a byproduct of adaptive resource management. Rather, it is integral to the resource management process because the ARM process is simply an extension of the scientific method. The scientific method is a means of asking and answering questions by observing phenomena and undertaking experimentation to reproduce the observation. It has seven steps, which include: (1) inquiry or asking a question (Why does an apple fall from a tree?); (2) performing background research (examine the phenomenon of why and how the apple falls); (3) forming a hypothesis (an explanation of why the apple falls); (4) conducting an experiment to test the hypothesis; (5) evaluating the experiment’s outcome or data and examining whether the data support the hypothesis (i.e., drawing a conclusion); (6) assessing whether the conclusion matches the result seen, which may necessitate more experimentation; and (7) communicating the result to the wider scientific community and testing it to see if it can be replicated.

Under the ARM approach, learning more about the system being assessed is valued to the degree that increased understanding helps improve the system‘s long-term performance. The ARM process recognizes that resource management decisions must be made even if one does not have all the information one would like to have, or one is not quite certain what the entire suite of impacts might be. ARM also recognizes that lessons learned and better understanding of the system gleaned from the process will be feedback into the ARM process for appropriate adjustments. Thus, the ARM process is both a management tool for achieving resource objectives and a method of inquiry for the discovery of additional information about the resource or system being managed.

Indeed, the process “is an attractive one, promising a way to make decisions in the face of current uncertainty while also reducing uncertainty over time.” Holly Doremus, Adaptive Management as an Information Problem, 89 N.C. L. Rev. 1455, 156–67 (2011). The key concept of adaptive management is that the execution of an objective should assimilate, rather than follow, learning about the system to be managed. Id. at 1457. Accordingly, an ARM framework overtly builds in potential for learning and corrections. Ideally, the process produces a pliant organizational configuration that acclimatizes to change. One field that will most likely employ and test the efficacy of the ARM process is global climate change. However, given ARM’s long-term iterative process, gathering workable results from longitudinal study may take too long, frustrating policy makers and other scientists.

Although the ARM process has strong adherents, it also has fervent doubters. It is in the field of climate change that its greatest doubters are found. Id. One area for skepticism is adaptive management’s implementation. There are questions, at this juncture, about where the learning ends and the management begins. Moreover, ARM’s requirement for collaboration and adaptive approaches, as well as its iterative and integrative process, creates ambiguity about how to balance learning and implementation. So, in enormously complex systems, such as climate change, “in the absence of a clear statement of federal priorities for sustaining or restoring ecosystems and the minimum level of ecosystem health needed to do so, ecosystem management has come to mean different things to different people.” U.S. Gen. Accounting Office, Ecosystem Management: Additional Actions Needed to Adequately Test a Promising Approach 38 (1994).

Additionally, in the United States., where science is routinely challenged and ridiculed, the politics and decision-making dynamics of various stakeholders, including those in government agencies, may well impact the process and impair its efficacy. Furthermore, the ARM process may cause new accountability conundrums, including affording cover to resource management agencies that allows them to postpone or delay “imposing politically controversial limits on economic activity.” Id. For instance, the ARM process might encounter agency bias when economic activity potentially conflicts with environmental or health safeguards. One example of this conflict is the regulation of Marcellus Shale, a large natural gas reservoir located primarily in New York and Pennsylvania. Accusations have been made that Pennsylvania’s governor and the state’s Department of Environmental Protection have favored drillers over drinking water quality. See, e.g., Eliza Griswold, Situation Normal All Fracked Up, N.Y. Times, Nov. 20, 2011, at MM44. A second example is the environmental justice plight of American Indians and the poor. See generally, Carlton Mark Waterhouse, The Past, Present, and Future of Environmental Justice, and Philomena Kebec, Environmental Justice and the American Indian Bands, Presentations at the Florida A&M College of Law’s Green Justice for All: International and Comparative Dimensions of Environmental Justice (Nov. 3–4, 2011).

Adaptive Resource Management Case Studies

The Comprehensive Everglades Restoration Plan (CERP). The Everglades, comprising at least half of southern Florida’s watershed, is the largest subtropical wilderness in the United States. It commences with the Kissimmee River and the Fisheating Creek, near Orlando, both of which flow southeastward and discharge into Lake Okeechobee. Designated as a World Heritage Site, International Biosphere Reserve, and Wetland of International Importance, the Everglades ecosystem is home to numerous rare and endangered fauna and flora. The Everglades is a brackish water ecosystem that includes sawgrass marshes, tropical hardwood hammocks, cypress swamps, estuarine mangrove forests, pine rockland, and the marine environment of Florida Bay. Over the past century, human activity, including sugar agriculture and other industries, has considerably altered the Everglades ecosystem in south Florida. Today, more than 1.5 million acres today are riddled with exotic and invasive plants and over 1 million acres are contaminated with mercury. Moreover, 68 plant and animal species are threatened or endangered, and populations of wading birds have declined by some 95 percent. National Research Council (NRC), Adaptive Monitoring and Assessment for the Comprehensive Everglades Restoration Plan 1 (National Academies Press, 2003).

Responding to the massive deterioration of the Everglades ecosystem, Congress enacted the Water Resources Development Act in 1992 (WRDA). WRDA authorized an across-the-board review of the Central and South Florida Project. Its main goal was to examine the restoration potential for the Everglades ecosystem. The review resulted in the Comprehensive Everglades Restoration Plan, probably the largest ecosystem restoration endeavor in the world. The restoration will entail enhancing and modifying the system of canals, levees, and water-control structures carved into the Everglades over the past century, in order to reestablish south Florida’s ecosystem. The restoration plan’s “fundamental premise . . . is that restoring the historical hydrologic regime to the remaining wetlands will reverse declines in many native species and biological communities.” Id. at 2.

Two agencies have been designated to implement the plan: the South Florida Water Management District and the U.S. Army Corps of Engineers. To date these agencies have yet to establish a timeline. Army Corps of Engineers, Central Everglades Planning Project (Dec. 2011). However, to place the CERP’s ARM in proper context, it is important to stress that human activity over the past 100 years has wrought incalculable damage upon the Everglades ecosystem. Indeed, the extent of the destruction of this once vibrant ecosystem is unimaginable, and it will take decades to restore it.

The CERP includes a Monitoring and Assessment Plan (MAP), which is grounded in contemporary scientific theory as well as ARM. Unfortunately, the CERP’s ARM is seriously flawed because of its underdeveloped feedback mechanism. Comprehensive Everglades Restoration Plan, at 51. Recognizing this problem when it studied the ARM, the NRC recommended three changes needed in the CERP’s ARM: (1) adaptive management must not simply mean “flexibility in decision making under uncertainty.” When considering a variety of possible strategies, actions should be taken that are informative, reversible, and less uncertain or at least robust to uncertainties; (2) institutional mechanisms should be created and sustained to ensure that scientific information is available and accessible to the decision-making process; and (3) opportunities for flexibility in design should be identified and operational features of the restoration plan components should be assessed to help prioritize monitoring and assessment activities.

The NRC panel also concluded that the “restoration goals, objectives, and targets for the Everglades are inadequately defined and are not reconciled with the large-scale forces of change in south Florida.” Id. The NRC urged that “[t]argets should be set as soon as possible that define the extent of compatibility between the built and natural systems and that address possible conflicts between ecological restoration and other policies, statutes, and social demands.” Id. Additionally, “[r]esearch and monitoring must continue to better conceptualize and describe current conditions in the Greater Everglades Ecosystem. Continued support and coordination of hydrologic and ecological monitoring and coordination among them are important components of monitoring and assessment. Integrated modeling is the best method for extrapolating findings over large areas and long periods.” Id.

Given the panel’s criticism of the ill-defined goals and targets, its recommendations are tepid. For example, declaring that “[t]argets should be set as soon as possible” leaves too much room for dalliance. Instead, concrete, prompt deadlines should be set, if the ARM process is to come to fruition. Similarly, the assertion that “continued support and coordination of hydrologic and ecological monitoring . . .” is just too “squishy.” The foregoing terminology provides no force or impetus to get things done. Each of the stakeholders, in a concerted effort, must push for continual progress though hard and fast dates. Of course, deadlines must be realistic—establishing hard and fast dates or targets does not mean that they will be achieved. See, e.g., the Clean Water Act.

Adaptive Management of Glen Canyon Dam and the Colorado River Ecosystem

The Glen Canyon Dam is situated on the Colorado River in northern Arizona’s Coconino County. The dam was constructed pursuant to the Colorado River Storage Project Act of 1956. Operated and managed by the Bureau of Reclamation (BOR), it was completed in 1963 and began operation in 1966. Glen Canyon spans the Colorado River south of the Arizona–Utah border and forms Lake Powell, which stretches upstream for 186 miles. Downstream, the Colorado passes through a 15-mile stretch of Glen Canyon and the Glen Canyon National Recreation Area into Marble Canyon at Lee’s Ferry, where it enters Grand Canyon National Park. The river then flows almost 300 miles through the Grand Canyon, before it arrives at Lake Mead, which is impounded behind Hoover Dam, 35 miles south of Las Vegas, Nevada. Nat’l Academy of Sciences, Downstream: Adaptive Management of Glen Canyon Dam and the Colorado River Ecosystem (1999)[hereinafter Downstream].

The dam’s construction engendered a huge wave of opposition. Indeed, as early as 1963, when the dam was built, David Brower, the first executive director of the Sierra Club and founder of the Wilderness Society, argued that “Hoover, Parker, and Davis dams already exist and control the river adequately; they could probably continue to do so until Lake Mead is silted in completely, perhaps two hundred years from now.” Eliot Porter, The Place No One Knew: Glen Canyon on the Colorado (1963).

To “protect, mitigate adverse impacts to, and improve the values for which Grand Canyon National Park and Glen Canyon National Recreation Area were established, including, but not limited to natural and cultural resources and visitor use,” Congress enacted the Grand Canyon Protection Act of 1992, Pub. L. No. 102-575, 106 Stat. 4669 (Oct. 30, 1992), which called for an adaptive management (AM) program focused “on the Colorado River Ecosystem.” Strategic Plan, Glen Canyon Adaptive Management Program (Final Draft, Aug. 17, 2001) [hereinafter Strategic Plan]. The Act mandated a Glen Canyon Dam environmental impact statement (EIS) and required the secretary of the interior to undertake a long-term monitoring of dam operation impacts on “resources of the Colorado River downstream of Glen Canyon Dam.” § 1804(d).

After extensive public participation and comment, interdisciplinary discussions, and professional analysis of the inputs and technical data, the final EIS was concluded in March 1995. Nine alternatives were chosen for detailed study by an interagency Glen Canyon Dam EIS team. These alternatives were selected “to allow the Secretary of the Interior to balance competing interests and to meet the statutory responsibilities for protecting downstream resources and producing hydropower, and to protect affected Native American interests.” Strategic Plan at 8. The EIS’s preferred alternative, “Modified Low Fluctuating Flows,” proposed quantified minimum and maximum flow rates and provided for controlled floods to enhance, protect, and restore downstream resources. Downstream at 2.

The EIS “identified a set of expected benefits associated with the preferred alternative, but it also recognized scientific uncertainties regarding the extent and ways in which those benefits could be achieved. The preferred alternative was and is an experiment.” Id. As part of the EIS, the BOR proposed a program of “adaptive management” to put the experiment into operation, and to improve results over time, based on a long-term monitoring and research plan. The AM program and the Glen Canyon Protection Act also faced a number of legal constraints, including “the compacts, federal laws, court decisions and decrees, contracts, and regulatory guidelines collectively known as the ‘Law of the River,’ which apportions the water and regulates the use and management of the Colorado River among the seven basin states and Mexico,” U.S. Dept. of Interior, The Law of the River, the Endangered Species Act, and federal trust responsibilities to Indian tribes. These limitations have been incorporated into the AM over the past 10 years.

The EIS, the ARM, and AM program were incorporated into a Record of Decision (ROD), which then Secretary of the Interior, Bruce Babbitt, signed on October 8, 1996. The ROD established an Adaptive Management Program (AMP), which is composed of five elements: (1) the Adaptive Management Work Group (AMWG); (2) the Technical Work Group (TWG); (3) the Grand Canyon Monitoring and Research Center; (4) independent review panels; and (5) the appointment by the secretary of the interior of a designee (today, that person is Anne Castle, the Department of the Interior’s Assistant Secretary for Water and Science). Each of these entities has its own jurisdiction and members are appointed by either federal or state agencies. For example, the AMWG is a federal advisory committee. Its members are appointed by the secretary of interior. This group is represented by each of the cooperating agencies, members from each of the Colorado River basin states, environmental groups, recreation interests, and contractors for federal power from Glen Canyon Dam.

Similarly, the TWG is a subcommittee of the AMWG. It is chaired by a TWG committee member, who is selected by the other committee members. The TWG’s responsibilities include developing criteria and standards for monitoring and research programs; providing periodic review and updates; developing resource management questions for the design of monitoring and research by the Grand Canyon Monitoring and Research Center; and providing information, as required, for the preparation of annual resource reports and other reports, necessitated by the AMWG. See generally, U.S. Dep’t. of the Interior, Bureau of Reclamation, Glen Canyon Dam Adaptive Management Program (2011).

A major concern about the plan is its scientific frailty. For example, the National Academy of Sciences (NAS) reported that the “preferred alternative was and is an experiment” and questioned the long-term efficacy of the AM program, noting that “the concept is still evolving.” Downstreamat 2. Indeed, Secretary of the Interior, Ken Salazar, in a letter dated August 7, 2009, addressed to the members and alternates of the Glen Canyon Dam Adaptive Management Work Group, noted that “President Obama and I have repeatedly stressed the importance of ensuring that sound science is used to inform the policy decisions facing our Government and our Nation. This principle is fundamental to the Department’s continued efforts to implement and balance the complex provisions of Federal law that apply to the Glen Canyon Dam Adaptive Management Program.” (Emphasis added).

Given the evolving nature of the proposals, the NAS report evaluated two strategic plans, one from 1997, a revised version in 1998, as well as a succeeding development. Subsequently, on July 7, 2008, the Bureau of Reclamation issued a Final Environmental Assessment for Experimental Releases from Glen Canyon Dam, Arizona, 2008 through 2012. U.S. Dep’t of the Interior, Bureau of Reclamation, Upper Colorado Region (2008). In that document the BOR proposed a series of experimental releases of water from Glen Canyon Dam to help native fish, particularly the endangered humpback chub, and conserve fine sediment in the Colorado River corridor in Grand Canyon National Park. These elements were not considered in the EIS or the ROD. They were, however, addressed by the NAS and were subsequently incorporated into the Finding of No Significant Impact (FONSI). Thus, the iterative process appears to be working.

On July 5, 2011, the Bureau of Reclamation issued two draft environmental assessments. The first is a Draft Environmental Assessment (EA) Non-Native Fish Control, U.S. Dep’t of the Interior, Bureau of Reclamation, Upper Colorado Region (2011). The Draft EA addresses actions that are necessary due to heavy competition and predation by nonnative rainbow and brown trout, which are reducing the survival of young native humpback chub and are threatening the survival of this threatened species. The second report is a Draft Environmental Assessment Development and Implementation of a Protocol for High-flow Experimental Releases from Glen Canyon Dam, Arizona, 2011–2020, U.S. Dep’t of the Interior, Bureau of Reclamation, Upper Colorado Region (2011). This report proposes “to develop and implement a protocol for high-flow experimental releases from Glen Canyon Dam to better determine whether and how sand conservation can be improved in the Colorado River corridor downstream of Glen Canyon Dam including areas within Grand Canyon National Park.” Id. Both drafts are protocols that are part of the ongoing Glen Canyon Dam ARM. The findings in each Draft EA are included in the learning and iterative process.

One might think that the lengthy process described above is interminably long. However, given the damage to the environment, occurring over 100 years in the case of the Everglades or 59 years for the Glen Canyon, the time spent resurrecting these ecosystems, whose damage we still do not fully understand, is not long at all. As a parallel, consider the following analogy: A bottle of syrup drops off a kitchen counter. It takes less than a second to fall a distance of four feet. However, it will take at least a half hour to clean up, not only picking up the glass shards and cleaning the sticky syrup but also mopping up the floor. And, let us not forget the effort of buying another bottle of that tasty maple syrup.

The importance of ongoing learning, as part of the ARM process of altering a plan or program as one gains new data, cannot be underestimated. Contrast the ARM process with that of an EIS under the National Environmental Protection Act (NEPA) where “[t]he combination of narrow judicial oversight and lack of accountability for error [in the EIS process] can produce terrible consequences.” David Hodas, The Role of Law in Defining Sustainable Development: NEPA Reconsidered, 3 Widener L. Rev. 1, 45–48 (1998). Because an EIS has no substantive mandate or substantive value, courts appear to be more concerned with whether an agency meets its procedural touchstones in evaluating the environmental impact of a particular project rather than examining whether the project is implemented in a manner consistent with the EIS assumptions. See Hodas, The Role of Law, supra at 46, citing Ogunquit Village Corp. v. Davis, 553 F.2d 243, 244–45 (1st Cir. 1977) where the court rejected a “challenge to a beach erosion project that utilized inland sand (an option not mentioned in the EIS) instead of the type of sand evaluated in the EIS.” Under NEPA, there are no post-EIS iterative adjustments based on project monitoring and feedback.

In conclusion, one cannot understate that the central idea of the ARM process is a valuable tool for managing big, complex systems in the face of scientific uncertainty. ARM is a learning and feedback process that takes time, diligence, and patience. But the process does not operate in a vacuum. Like so many human endeavors it is buffeted by the turbulence of political and social pressures. 


  • About NR&E

  • Additional Resources

  • Contact Us

Stay Connected


Book: International Environmental Law: The Practitioner's Guide to the Laws of the Planet