Basic principles of adaptive management

1
Basic principles of adaptive management
?
?
?
James T. PetersonUSGS, Georgia Cooperative Fish
Wildlife Res. UnitWarnell School of Forestry
and Natural ResourcesUniversity of Georgia
2
All Natural Resource Conservation and Management
Involves Decision-making

• Endangered species listing
• (Yes, No)

• Endangered species recovery planning
• (where, what, when, how)

• Habitat conservation planning
• (where, what)

• State wildlife management planning
• (where, what, when, how)

• Harvested population management
• (season length, bag limits)

3
Example ES Recovery Planning
4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
Basic Elements of Natural Resource
Decision-making
Current species status
Actual future status
Management action
Notions of system dynamics
Anticipated future status
Revise/update ideas
But how are decisions really made??
8
Information
Conservation priority
Terrain and Geomorphology
Rookie
9
Problems with Black Box Approaches

• Generally not explicit or transparent

• Many unidentified and unstated assumptions

Low priority!
High priority!
Conservation potential
Conservation potential
Fragmentation
Fragmentation
10
Problems with Black Box Approaches

• Not transferable or repeatable

• No formal learning component

Rookie
11
Failure to use quantifiable objectives or feedback
12
Quantitative Decision Analysis
External Physical Influences (weather, habitat
conditions)
Management Actions (habitat improvement, predator
control, captive breeding)
External Biological Influences (competitors,
predators, exotics)
Populations/ Communities
Stakeholder Benefits (Public satisfaction)
The Basis of Adaptive Management
13
Decision Analysis Process

• Identify the decision situation and objectives

• Identify and separate fundamental and means
  objectives (essential!)

Means objectives network
Minimize extinction risk
Maximize spatial distribution
Minimize mortality
Exotic control
Increase habitat
Establish corridors
14
An alternative perspective

• Focuses efforts things that matter most to the
  decision-maker

Ecologist perspective
15
The importance of identifying and structuring
objectives
Habitat availability
Probability persistence
Time
16
The importance of identifying and structuring
objectives
Agreement on management objectives? (where do we
go)
Yes
No
Routine management science
Negotiation, compromise
Yes
Agreement on science? (how do we get there)
No
Adaptive resource management
Conflict
Biology
Politics
from Lee (1993)
17
Decision Analysis Process

• Identify the decision situation and objectives

• Identify and separate fundamental and means
  objectives (essential!)

Utility species A species B
18
Decision Analysis Process

• Identify the decision situation and objectives

• Identify and separate fundamental and means
  objectives (essential!)

• Identify decision alternatives
• e.g., land management, captive breeding,
  regulations

19
Decision Analysis Process

• Identify key uncertainties
• - those things that can affect the utility

?
?
?
Arrgh jargon
20
Jargon Busting Uncertainty
Wet year
Dry year
21
Jargon Busting Uncertainty
Wet year
Dry year
22
Jargon Busting Uncertainty
23
Decision Analysis Process

• Construct the model (need to estimate the
  outcome!)
• - Simple (simple is good!)
• - Complex

Where do we get the information?
Empirical data
Published reports (meta-analysis)
and when information is completely lacking
24
Expert Judgement
Population size Low Moderate High
Natural mortality
Habitat availability
?
?
?
Low
Low
. . .
. . .
. . .
Medium
Low
Low
High
. . .
. . .
25
Making the Decision
Probability weighted utility
Cost
Goal minimize cost of conservation
action while increasing persistence
colonize
0.60
70k
(reserve)
Yes
(reserve habitat)
Purchase reserve
0.40
270k
No
Improve habitat
200k
(habitat)
What would I decide if the probability of
colonization were 0.4?
26
Making the Decision
colonize
0.40
70k
Yes
Purchase reserve
0.60
270k
No
Improve habitat
200k
Certainty-weighted cost (thousands)
Purchase reserve 0.470 0.6270
190
Improve habitat 200
27
Making the Decision
Optimization (number crunching)
Optimal
Persistence probability
Population size
Conservation policy
28
Identify key uncertainties Sensitivity Analysis
Model components
Colonization mechanisms (system uncertainty)
Carrying capacity
Juvenile mortality
Exotic density
Precip. during breeding
Natural adult mortality
0.25
0.35
0.45
0.55
0.65
Persistence probability
29
Reducing Uncertainty Learning
New Information
Prior Estimate
Prior Estimate
Posterior Estimate
Posterior Estimate
30
Learning how a system works

• Conduct additional studies/experiments
• - Time consuming (decisions cant wait)
• - Expensive
• - Risky

Population size
time
31
Learning how a system works

• Conduct additional studies/experiments
• - Time consuming (decisions cant wait)
• - Expensive
• - Potentially wasteful

• Learn while managing (Adaptive Management)
• - Decisions are made
• - Requires sequential dynamic
  decision-making time and/or space

Sequential decision-making through time
Population
Population
Population
Population
Decision
Decision
Decision
Decision
C. Moore
32
Learning how a system works

• Conduct additional studies/experiments
• - Time consuming (decisions cant wait)
• - Expensive
• - Potentially wasteful

• Learn while managing (Adaptive Management)
• - Decisions are made
• - Requires sequential dynamic
  decision-making time and/or space

Sequential decision-making through space
Site A
Site B
Site C
Site D
Site E
Site F
Site G
C. Moore
33
An illustration Robust redhorse
Why are they so rare?
34
Management action
System dynamics
Predicted redhorse
70
Flow variability
0.5
Decrease power generation
10
0.5
Predation
10
Flow variability
0.5
Control flatheads
50
0.5
Predation
Redhorse abundance
35
Management action
System dynamics
Predicted redhorse
70
Flow variability
0.6
Decrease power generation
10
0.4
Predation
10
Flow variability
0.6
Control flatheads
50
0.4
Predation
Redhorse abundance
36
Management action
System dynamics
Predicted redhorse
70
Flow variability
0.3
Decrease power generation
10
0.7
Predation
10
Flow variability
0.3
Control flatheads
50
0.7
Predation
Redhorse abundance
37
Learning how a system works

• Conduct additional studies/experiments
• - Time consuming (decisions cant wait)
• - Expensive
• - Potentially wasteful

• Learn while managing (Adaptive Management)
• - Decisions are made
• - Requires sequential dynamic
  decision-making time and/or space
• - Requires monitoring
• Current state of the system
  (where are we?)
• Actual outcome of the decision
  (where did we end up?)

38
Adaptive Management
Basic elements
Info (monitoring data)t
Info (monitoring data) t1
Actual future population
Management action
Current population
Model A (hypothesis)
Predicted population
Bayes Rule
Model B (hypothesis)
Predicted population
39
Additional sources of uncertainty (yes, more
jargon)
Action A
Action B
True
40
Additional sources of uncertainty (yes, more
jargon)
Action A
Observed
Action B
True
41
Additional sources of uncertainty (yes, more
jargon)
Partial controllability
The mismatch between the management intention
and the actual action - weather patterns
(similar to environmental stochasticity)
Actual density
Intended moist soil plant density (action)
42
Basic Forms of Adaptive Management
Passive
Active
Relatively simple Choose decisions as if current
uncertainty will not change Information gained
through monitoring is incorporated, but not in a
planned way
Computationally more complex Potential
information returned by each decision is given
value when evaluating each decision Probing is
valued when - uncertainty is very
high -management loss is expected to be high
if uncertainty is unresolved
43
Misconception Mania
Adaptive management is Management A formal
quantitative decision making process for choosing
the best strategy and learning Flexible and
can be used with a wide range of model types and
complexity
Adaptive management is NOT Research Trying a
management strategy until it fails, then adopting
another Implementing a strategy (experiment)
until an effect is detected Too complex, too
expensive, too hard
44
Cautionary thoughts
When the only tool you have is a hammer,
you tend to see every problem
as a nail. - A. Maslow
Adaptive management is not the answer to all
management problems
It is only useful when Biological
uncertainty is the main issue, i.e., management
objectives are clearly articulated and agreed
upon Decision-making is sequential The
system and decision outcomes can and will be
monitored