Eli Fenichel, Jean Tsao, Michael Jones, and Graham Hickling

1
Eli Fenichel, Jean Tsao, Michael Jones, and
Graham Hickling
Prepared for the Sea Lamprey Integration
Committee, 4/24/2007
Presentation available at www.msu.edu/fenichel
2
Investors, Benefits, Costs, Risks
Benefits value of assets received value of
those given up.
Cost forgone opportunities
Risk (likelihood x consequences) of an event

• Investors experience risk because
• They are uncertain about the world
• Once an investment is made, assets used often not
  be recovered

Shrewd investors ask How do my decisions and
actions affect the likelihood and consequences of
events I experience?
3
Making Decisions in a Certain World Net Present
Value (NPV)
Proceed with program if EBenefits Costs ? 0
EBenefits from the lake using Lake Ontario sea
lamprey in the SMRT program Benefits from the
lake without using Lake Ontario sea lamprey in
the SMRT program ? 0
4
Decision Boundary Certain World Traditional
Expected Net Present Value ? 1 creating a 45º
angle
Lamprey from Lake Ontario too
Decision boundary, Slope ? 1
Expected Net Benefits of translocating lamprey
Lamprey from upper Great Lakes only
Expected Net Benefits of NOT translocating
lamprey
5
Investment Approach for an Uncertain and
Irreversible World
When we think of risk, we think of
precaution Adjust the decision rule a
precautionary adjustment (PA)
Value of flexibility or option value
Proceed with the program if
)
E


Benefits
PA
Costs
(
0
?
where PA gt 0
6
Options Thinking
Option a right but NOT an obligation to do
something e.g. the right, but not the obligation
to continue the transfer of lamprey from Lake
Ontario to the St. Marys River as part of SMRT
Evaluating options asks When to exercise the
option? OR How much is the option worth?
How much we need to know to justify the
action Gives us a measure of how much precaution
is needed
with few options
7
Sterile Male Release Decision Tree
Outcomes
Action
Uncertainty
More lamprey suppression Higher probability of
additional disease problems ?Fishery benefits
??
Obtain lamprey from Lake Ontario
Pathogen moved with lamprey
More lamprey suppression Lower probability of
additional disease problems ? Fishery
benefits ??
do
Pathogen not moved with lamprey
dont
Less lamprey suppression Lower probability of
additional disease problems ? Fishery
benefits ??
Lamprey not moved
8
Option to Continue Sterile Male Transfers
Decision Tree
Pay-off 1
Success (P1)
Pay-off 7
Exercise option
Success (P2)
Pay-off 2
Failure (1-P1)
Exercise option
Failure (1-P2)
Experience (changes P)
Hold option
Pay-off 6
Hold option
Pay-off 5
9
Decision Boundary Traditional Expected Net
Present Value ? 1 creating a 45º angle
Lamprey from Lake Ontario too
Decision boundary, Slope ? 1
Expected Net Benefits of translocating lamprey
Lamprey from upper Great Lakes only
Expected Net Benefits of NOT translocating
lamprey
10
Decision Boundary Using Real Options ?gt1 but ?lt?
Lamprey from Lake Ontario too
Decision boundary, Slope ?
Expected Net Benefits of translocating lamprey
Lamprey from upper Great Lakes only
Expected Net Benefits of NOT translocating
lamprey
11
Conceptual Model for SMRT Decisions
Commercial fishery benefits
Haeseker et al.

Lamprey induced mortality
Recreational fishery benefits
Attacks on fish
Fish population effects
Sea lamprey parasites
Lamprey related disease mortality
Non-use/ ecosystem benefits
Other sources of pathogen transfer

Lamprey spawning potential
Sea lamprey ammocoetes
?
SMRT decision
Pathogen transferred with lamprey
Screening Defensive expenditures
?
Extra lamprey Expenditures
12
The most asked question with regards to fish
health and fish transfers
How many fish should we screen?
But, what we really want to know is what is the
probability of transferring infected fish, given
screening effort And is it acceptable?
13
Likelihood of Translocation Depends on BOTH the
number of fish screening and the number of fish
transferred
Pr(t) 1 Pr(d) Pr(b)
Joint probability of pathogen transfer Pr(t)
The probability of a pathogen being present in
the transfer batch, Pr(b)
The probability of detecting a pathogen by
screening, Pr(d)
Outcome
No transfer
No transfer
Detect Pr(d)
0
Present Pr(b)
Chance of pathogen transfer
1-Pr(d)Pr(b)
Fail to detect 1-Pr(d)
Not Present 1-Pr(b)
Chance of no pathogen transfer
1-Pr(d)1-Pr(b)
14
What a Sample Really Means
The number of fish needed for testing to achieve
95 confidence that the prevalence of the
pathogen is less than a given level
Assuming a perfect test
Required Sample Size
95 Confidence
15
Sample Size and Transferring and Infected Fish
1000 fish transferred
Probability of transferring an Infected Fish
16
What has sea lamprey screening told us?
Consider 2004 - 600 sea lamprey were capture from
Lake Ontario for SMRT 60 were screened Provided
95 confidence Heterosporosis sp. would be
detected is the true prevalence were 5. If only
200 where moved the probability of moving at
least 1 infected fish ? 5. OR Provided 45
confidence Heterosporosis sp. would be detected
is the true prevalence were 1. If only 200 where
moved the probability of moving at least 1
infected fish ? 47. (if 600 lamprey where moved
? 55).
17
Summary of Screening
Addressed the probability of moving one infected
fish. We CAN NOT address the consequences of
moving one infected fish from a probabilistic
screening model alone. Therefore, we can not talk
about the RISK of moving infected lamprey without
understanding the impact on the system.
Screening 60 fish is clearly arbitrary, and the
proper effort depends on a wider set of tradeoffs.
An integrated and systems approach to evaluating
risk and tradeoffs is needed.
18
Future Directions

• Combining existing models to take a systematic
  look at the problem and the tradeoffs.
• Incorporating in the uncertainties such as
  pathogens, alternative pathogen screening
  strategies, and societal values.
• Conducting real options analysis to determine a
  decision boundary.
• Make a recommendations to either proceed with
  SMRT transfers or delay.
• Serve as a base/model for other systematic
  approaches to addressing tradeoffs in sea lamprey
  control and in general. Including prioritizing
  research.

19
Thank you
Especially Great Lakes Fishery
Commission Department of Fisheries and Wildlife
_at_ Michigan State University Quantitative
Fisheries Center _at_ Michigan State University The
Saginaw Bay Walleye Club