Introductory Guide AM
Introductory Guide AM
MANAGEMENT
for Project Leaders and Participants
January 1999
Brian Nyberg
Forest Practices Branch,
B.C. Forest Service
P.O. Box 9513 Stn Prov Govt
Victoria, B.C.
Canada V8W 9C2
Phone: (250) 387-3144
E-mail: brian.nyberg@gems6.gov.bc.ca
Introductory guide to adaptive management
Preface
i
Introductory guide to adaptive management
the Columbia River Basin, management of acid rain, and water management in the Florida
Everglades. Its application to forest management issues is now receiving increasing attention.
For example, ten "Adaptive Management Areas" are now operating in the U.S. Pacific
Northwest, the Clayoquot Sound Scientific Panel recommended an adaptive approach to
managing Clayoquot Sound, and Alberta-Pacific Forest Industries (AlPac) has adopted adaptive
management as principle for guiding operations in its large Forest Management Area in Alberta.
This introductory guide is intended to promote and assist with the application of adaptive
management to a range of forest management issues in BC.
Adaptive management is a systematic process for continually improving management policies and
practices by learning from the outcomes of operational programs. Its most effective form ”active”
adaptive management employs management programs that are designed to experimentally
compare selected policies or practices, by evaluating alternative hypotheses about the system
being managed.”
From Nyberg, J.B. 1998. Statistics and the practice of adaptive management. Pages 1-7 in
Statistical Methods for Adaptive Management Studies, V. Sit and B. Taylor, (editors). Land
Manage. Handbook 42, B.C. Ministry of Forests, Victoria, BC.
ii
Introductory guide to adaptive management
Introduction
The purpose of this document is to assist those who wish to apply adaptive management,
by providing a quick overview of the six main steps involved: 1) problem assessment, 2) design,
3) implementation, 4) monitoring, 5) evaluation, and 6) adjustment. The framework formed by
these six steps is intended to encourage a thoughtful, disciplined approach to management,
without constraining the creativity that is vital to dealing effectively with uncertainty and change.
The details of how the steps are applied and the level of rigour used depends on the problem and
on the imagination of participants. This guide is intended to provide direction, stimulate thought
and augment discussions with resource people; it is not a detailed "how-to" manual. Nor is it a
comprehensive discussion of adaptive management; those who want more background
information can refer to the key references cited at the end of this document.
The six main steps in adaptive management are shown in Figure 1. Step 1 (problem
assessment) is often done in one or more facilitated workshops. Participants define the scope of
the management problem, synthesize existing knowledge about the system, and explore the
potential outcomes of alternative management actions. Explicit forecasts are made about
outcomes, in order to assess which actions are most likely to meet management objectives.
During this exploration and forecasting process, key gaps in understanding of the system (i.e.,
those that limit the ability to predict outcomes) are identified. Step 2 (design) involves designing
a management plan and monitoring program that will provide reliable feedback about the
effectiveness of the chosen actions. Ideally, the plan should also be designed to yield information
that will fill the key gaps in understanding identified in Step 1. It is useful to evaluate one or
more proposed plans or designs, on the basis of costs, risks, informativeness and ability to meet
management objectives. In Step 3 (implementation), the plan is put into practice. In Step 4
(monitoring), indicators are monitored to determine how effective actions are in meeting
management objectives, and to test the hypothesised relationships that formed the basis for the
forecasts. Step 5 (evaluation) involves comparing the actual outcomes to forecasts and
interpreting the reasons underlying any differences. In Step 6 (adjustment), practices, objectives,
and the models used to make forecasts are adjusted to reflect new understanding. Understanding
gained in the each of these six steps may lead to reassessment of the problem, new questions, and
new options to try in a continual cycle of improvement.
Introductory guide to adaptive management
Assess problem
Adjust
Design
Evaluate
Implement
Monitor
In reality, some of the steps outlined will overlap; some will have to be revisited; some
may be better done in more detail than others. All steps should be planned in advance, though it
may be necessary to modify them later. All six steps are essential to adaptive management:
omission of one or more will hamper the ability to learn from management actions. In addition,
documenting the key elements of each step, and communicating results are crucial to building a
"legacy of knowledge", especially for projects that extend over a long time.
This guide summarizes the key points to consider in each step and provides a checklist for
quick reference. We have also included suggestions for creating the conditions that will facilitate
long-term learning by individuals and their agencies, companies or communities. Some potential
barriers to adaptive management, and suggestions for overcoming them, are noted in Appendix 2.
Learning how to do adaptive management is itself an adaptive process. Please let us know
what works and what doesn't. We would appreciate feedback on the value of the outlined
framework, how it could be improved, and how you applied it to different problems.
in some cases, adaptive management may also help detect cumulative, long-term, large-
scale, and emergent effects of actions
2
Introductory guide to adaptive management
1. ASSESS PROBLEM
Define spatial scale, temporal scale, and range of factors (i.e., values) to be considered.
Consider aspects of the system that affect indicators or that are likely affected by
management actions.
Avoid defining problem in terms of preconceived solutions, since this would limit the
development of imaginative alternatives.
1.2 Define measurable management objectives and list potential management actions.
Indicators are measurable attributes of system behaviour that allow you to weigh
management options and, eventually, assess outcomes.
Select indicators that are relevant to objectives and responsive to management actions.
3
Introductory guide to adaptive management
Take into account the cost and practicality of measuring each indicator.
Select some indicators that respond in the short term, some in the medium term, and some in
the long term. Select indicators that respond at different spatial scales (e.g., site, landscape,
region).
Develop a conceptual model of the system: outline linkages and describe the functional
relationships between actions and indicators (e.g., using box-and-arrow diagrams, graphs,
equations).
If warranted, modify an existing simulation model or build a new one to represent the
conceptual model. Simulation models are particularly valuable for projecting changes over
time and space and assessing the integrated consequences of a suite of actions.
Use the model (whether it is a simulation model or conceptual model) to explore the effects
of alternative actions.
1.5 Make explicit forecasts about response of indicators to alternative management actions.
Forecasts can be based on outputs from simulation models or, for simple problems, on the
graphs or diagrams used to describe the relationships between actions and indicators.
hrough exploring alternatives and forecasting responses, key gaps in understanding of the
system will emerge. Express these key uncertainties as alternative hypotheses of system
function. Hypotheses can be expressed as simple graphs, or where appropriate simulation
models exist, as functional relationships or sets of model parameters.
Consider the relationship between action(s) and indicators over a range of conditions (i.e.,
how will an indicator respond to different degrees of a treatment?).
conceptual models (e.g., box and arrow diagrams of potential impact pathways)
simulation models
1Understanding of complex and dynamic ecological systems will always be incomplete. However, not all gaps in
understanding necessarily need to be filled in order to decide between alternative management actions. For example,
where different assumptions lead to the same forecast, or to the same choice of management action, there is no need to
resolve the uncertainty about which assumption is "correct".
4
Introductory guide to adaptive management
Consider a number of management options, for example: a passive approach, where one
action is implemented; an active approach, where several alternatives are compared; or
testing a range of options at a pilot scale, before testing one or more at a larger scale.
If necessary, consider how and when to relax some of the design principles; note the
consequences this will have for how the results are interpreted, and for the value of the
resulting information.
5
Introductory guide to adaptive management
Evaluate the proposed plan or plans, on the basis of ability to meet long term objectives,
ecological and economic costs, risk of negative outcomes, and ability to fill key gaps in
understanding. Decide which proposed plan to implement.
Specify:
Set up system for managing data over the long term (e.g., storage, analysis, access).
Agree on who will interpret data and who will have access to it.
Identify who needs what information when in order to make timely changes.
Define the intensity and degree of response in an indicator that will trigger a change in
management actions or objectives.
djustments should reflect the trade-off between the costs of acting if preliminary results
later prove to be incorrect, and the costs of not acting if they later prove to be correct.
quantitative decision analysis (see Keeney, 1982; McAllister and Peterman, 1992)
project plan that documents uncertainties, design, and other parts of Steps 1 and 2
6
Introductory guide to adaptive management
In some circumstances, it may be necessary to deviate from the original plan: decide when
and what type of deviations are acceptable. Ensure that these circumstances are clear and
accepted by all partners. Otherwise, the "tyranny of small decisions" may eventually
invalidate the plan or lead to loss of confidence and support by partners.
4. MONITOR
Monitoring is often neglected in conventional models (i.e., hypotheses) of how the system
approaches to management, yet it is critical to functions, and take appropriate corrective
improvement. Monitoring allows you to assess action. Monitoring can also determine if actions
how actions actually affect indicators. This were implemented as planned, and may detect
information then allows you to evaluate the "surprising" events.
effectiveness of alternative actions, adjust
consider developing new, innovative and inexpensive monitoring techniques (e.g., videos or
photographs of sites)
7
Introductory guide to adaptive management
5. EVALUATE
Evaluate the reasons underlying any differences between actual and forecasted outcomes.
Were the objectives met? If not, why not?
5.2 Document results and communicate them to others facing similar management issues.
6. ADJUST
Information must be used in order to have adjustments are rarely as simple as those
value. Information gained through the proposed initially. In addition, management
preceding five steps should be used to verify or experiments may yield some useful information
update the models used to make the initial that was not anticipated. Well-defined feedback
forecasts, and adjust management actions as loops are intended to ensure that information is
necessary. Objectives should be reviewed and used promptly and appropriately; they are not
adjusted to ensure that they remain consistent meant to be rigid rules that frustrate adaptation.
with overall goals and values. Often, new information will suggest new
In order to facilitate change, participants should management solutions, or new questions to
consider at the outset (i.e., in Step 2) how answer...leading to another cycle of assessment,
actions might be adjusted. However, results design, implementation, monitoring and
are rarely as clear as anticipated, and thus evaluation.
6.1 Identify where uncertainties have been reduced, and where they remain unresolved.
6.2 Adjust the model used to forecast outcomes (Step 1) so that it reflects the hypothesis
supported by results.
8
Introductory guide to adaptive management
6.3 Adjust subsequent management decisions and policies, and reevaluate objectives, as
necessary.
Future actions should be based on which hypothesis of system function was supported by
the results.
6.4 Make new predictions, design new management experiments, test new options.
reasoning behind the choice of management plan, monitoring program and expected
outcomes;
2. Ensure that such information is accessible over the long term, and in the event of turnover
in participants.
3. Define who is responsible for co-ordinating and for carrying out each task.
4. Set time lines for carrying out each task or part of project.
In communicating the results, specify which uncertainties have been reduced, and how this
affects understanding of the system and future management actions.
9
Introductory guide to adaptive management
mistakes are recognized as the price of innovation and are treated as opportunities to
learn;
people are explicitly rewarded for innovation and learning (i.e., in performance
evaluations).
unanticipated and uncontrollable events that compromise design (e.g., wildfires that
destroy replicates);
interruptions in funding;
4. Anticipate potential barriers and develop strategies for overcoming or minimizing them.
See Appendix 2.
A Final Comment
In using this introductory guide, keep in mind that adaptive management is like painting: knowing
the steps is important, but it isn't enough to create great art. Potential project leaders and
participants are encouraged to use other resources, including those offered by the BC Forest
Service Adaptive Forest Management initiative. We must learn how to do adaptive management
by doing it.
10
Introductory guide to adaptive management
APPENDIX 1
AEAM is a process where participants with a diversity of skills and expertise are brought together in a
workshop or series of workshops to assess a management problem and explore management options.
Usually, participants work with modellers to develop a computer simulation model that they then use to
explore various "what if..?" scenarios and evaluate potential outcomes of different management actions.
In addition, significant benefits are derived from the process of building the model. The workshops are
intended to encourage debate about system response to management actions and to stimulate a creative
search for new solutions, rather than build consensus around a single solution. In some cases, the scope
of the problem and the options explored will be bounded by prior decisions about land-use (for example,
those defined in Land and Resource Management Plans, or other regional or local land-use plans).
The AEAM workshops are valuable for:
provides a consistent basis for participants to discuss and evaluate management options.
AEAM workshops typically involve the steps outlined below, although they can be tailored to suit the
management problem and available budget. In some cases these steps will be done in a single
workshop, over one or several days; in other cases, they may be done in a series of workshops, with
some participants working on particular aspects of the problem in between. Work is done in both
plenary sessions and sub-groups.
While model-building is valuable for focusing discussion, workshops where time or budget constraints
preclude computer modelling can still be useful. The development of conceptual models is itself a very
worthwhile exercise. To maximize the value of the workshop, objectives and expectations must be
clearly understood by all participants, including the modellers and facilitators.
AEAM is an iterative process - later steps may lead to reevaluation and reiteration of earlier steps. For
example, it may be useful to model the problem at several different spatial scales. Similarly, exploring
the effects of management options may suggest alterations to model parameters or relationships, or new
options to try.
Steps that were outlined in section 1 (Assessing Problem) are noted.
11
Introductory guide to adaptive management
Prior to the first workshop, typically in a meeting between the facilitator(s) and the project
leader(s):
1. Identify key participants.
The first workshop may involve 20-30 people, including:
modellers/facilitators
"knowledge experts" from a range of disciplines (e.g., hydrologists, wildlife biologists, fisheries
biologists, social scientists, etc.)
policy-makers
forest workers
other stakeholders
Participants can be drawn from government agencies, industry, community groups, environmental
organizations, First Nations or other groups with expertise to contribute or who will be affected by
management decisions. The mix and quality of participants is critical to workshop success. Participants
should not only have relevant expertise, but should also be creative, innovative thinkers.
2. Define the initial scope of the problem and key problem features.
(e.g., spatial and temporal scale, range of factors to consider, key indicators)
For some problems it also may be useful to summarize existing, accepted knowledge about the
system. This can "jump start" the process and minimize time spent on issues about which their is
already general agreement.
It is critical that participants not feel constrained by this initial problem scoping;
leaders/facilitators must be prepared to alter or abandon any boundaries or framework
developed prior to the workshop.
In some cases, the scope of the problem will be influenced by land-use decisions made by other
bodies (e.g., LRMPs). Decisions made by AEAM participants should not conflict with or
supersede those made by other decision-making bodies.
The participants in the workshop should have the opportunity to revise the scope defined by the
facilitator(s) and project leader(s) prior to the workshop, to ensure that issues of concern are
addressed. This is the first step in building a common understanding of the problem.
12
Introductory guide to adaptive management
Actions and indicators are then grouped into logical sub-groups (e.g., based on theme or scale).
Participants are assigned to these sub-groups based the knowledge and skills they can contribute. In
each sub-group:
7. Draw impact hypothesis diagrams (for given group of actions and indicators).
These "box and arrow" diagrams outline the linkages between management actions and
indicators and represent a synthesis of existing knowledge on each subsystem.
Impact hypothesis diagrams are not intended to show detailed links between all components of
the system - only those that affect management outcomes/decisions.
Evaluate links in impact hypothesis diagram, considering: quality of existing information, level
of influence on outcome, feasibility of filling information gap. One suggested system for
assessing alternative hypotheses is:
1. unlikely, not worth testing
2. already tested
3. possible, too difficult to test
4. uncertain, research needed
5. likely, worth testing
Work in sub-groups alternates with discussion in plenary sessions. This ensures that the sub-groups
understand and have a chance to comment on each other's assumptions, and ensures that the diagrams
form a coherent whole. "Outputs" of one diagram will be "inputs" to another.
In some cases there may be an existing model that captures the relevant pathways and
conceptual model expressed in step 7 above. If so, this model can be used to explore alternative
actions. If a suitable model is not available, develop a new one based on the impact hypothesis
diagrams.
While a crude model can be developed in the workshop, typically it will have to be refined
before it can be used for exploring scenarios and forecasting outcomes.
10. Test and validate the model; do sensitivity analysis on model parameters.
Test model to determine how sensitive outcomes are to variations in model parameters.
Sensitivity analysis helps narrow in on those areas of uncertainty that most affect management
outcomes.
This "gaming" with the model is usually done in the group as a whole.
13
Introductory guide to adaptive management
12. Make explicit predictions about response of indicators (see section 1).
A reminder: AEAM is an iterative process- not sequential as summarized above. The model will
be continuously refined and tested, new options developed and explored, and new predictions
made.
documented assumptions about how the system functions, how variables respond (i.e., the
model);
a list of key uncertainties in model parameters and relationships (i.e., those that have most
influence on outcomes);
a list of plausible alternative hypotheses that should be tested, in order to improve understanding
of system behaviour and its representation in the model, and thus improve management
decisions;
an idea of what "probing actions" can be taken to test these alternative hypotheses.
14
Introductory guide to adaptive management
APPENDIX 2
Bayesian statistics)
(potential constraints on design include: long
draw on other sources of knowledge to help interpret
response times, large spatial scales, high results (e.g., retrospective studies, descriptive studies,
variability between sites, high variability in process research, local knowledge)
measurements)
designs that are less than ideal may still provide useful
information - consider consequences of relaxing design
rules and accepting lower levels of precision, confidence in
results
lack of powerful design does not preclude adaptive
management - it just makes interpretation of results more
difficult
2 use quantitative methods of risk analysis to assess and
3. reluctance to "experiment" with high
16
Introductory guide to adaptive management
all,
17
Introductory guide to adaptive management
all,
18
Introductory guide to adaptive management
APPENDIX 3
Definitions
Indicators: - measures of system behaviour in terms of meaningful and perceptible attributes
- what do we really need to know in order to make a decision?
- what do we need to measure to determine if management objectives have been met?
- allow us to assess alternatives, weigh options
Hypotheses: - predictions about how one or more indicators will respond to management actions
- alternative explanations for the mechanisms that underlie observed responses
- the most useful hypotheses state the degree of response, or describe the response over
a range of conditions; hypotheses that state only whether an indicator will or won’t
respond to a treatment are less useful for deciding between alternative courses of action,
or making informed trade-offs
Taylor, B, L. Kremsater, and R. Ellis. 1997. Adaptive Management of Forests in British Columbia.
British Columbia Ministry of Forests, Forest Practices Branch, Victoria. 93 p. (Note that in developing
this introductory guide, the elements of adaptive management (outlined in chapter 5) have been updated and
revised.)
Walters, C.J. 1986. Adaptive Management of Renewable Resources. McGraw-Hill, New York.
374pp.
20
Introductory guide to adaptive management
QUICK REFERENCE:
1. Assess Problem
Design management plan that will provide reliable feedback and fill gaps in understanding.
3. Implement
4. Monitor
5. Evaluate/ Adjust
Document results and communicate them to others facing similar management issues.
6. Adjust
Identify where uncertainties have been reduces and where they remain unresolved.
Adjust the model used to forecast outcomes, so that it reflects the hypothesis supported by
results.
Make new predictions, design new management experiments, test new options...repeat cycle.
21
Introductory guide to adaptive management
Define who is responsible for co-ordinating and for carrying out each task.
Set time lines for carrying out each task or part of project.
Create Success
Ensure that all participants have a clear, consistent understanding of adaptive management.
Anticipate potential barriers and develop strategies for overcoming or minimizing them.
22