Bridging from single species tactical advice to ecosystem-based strategic advice Lesson 1: We should expect our models to fail, sometimes spectacularly.

Bridging from single species tactical advice to ecosystem-

based strategic advice

Lesson 1: We should expect our models to fail, sometimes spectacularly so. But when we do we should take the opportunity to learn from the failure

Traditional approach to fisheries management

• Management approach focuses on F via catch & effort regulation

– Stock assessments recommend B and F reference points

– SSC recommend ABC (<=OFL) that includes scientific uncertainty –tactical advice

– Council recommend ACT (<=ABC) that includes management uncertainty (implementation uncertainty)

• Stakeholders limited to

– Commercial fishery interests

– Managers (protecting societal interest)

)(1 MFRGBB tt

Approach to incorporating scientific uncertainty

• Select %ile of probability distribution

• Lack of data in many systems means that P* is becoming ordinal as cFlim

But this assumes that all stakeholders will want a maximal F

4

Stakeholders can be responsible

Council recommendation• 10M Lbs quota• 3 fish bag limit• >24 in min size

Stakeholder choice• 8M Lbs quota• 2 fish bag limit• >28 in min size

How and why did recreational anglers, CCA, and NGOs become

more conservative than managers?

5

Stakeholder-centered approach

Stakeholders propose objectives, options and performance

measures

Revise options and performance measuresModel development

and modification Stakeholders

Review

model results

Recommendations

6

Model Schematic

“Natural” Deaths

Reproduction

Growth & Maturity

Fishing

Migration

7

Model Structure• Abundance

• Mortality

• Catch

N = Abundance F = Inst. Fishing mort. RateM = Natural mort. Z = Total mort.p = migration ratey = year s = seasona = age x = sexo = area f = fishery

oxsayZoxay

osaoxtaty eNpN ,,,

,,,,,,,

f

foxasyxaoxasy FMZ ,,,,,,,,,,

oxasyoxasyfoxasy NFC ,,,,,,,,,,,,,

8

Parameter uncertainty

• Simulations drew from parameter distributions that reflected either– Scientific uncertainty– System uncertainty

• Recreational F – used 3 scenarios: either increasing, constant and decreasing scenarios with white noise variability (lognormal CV 10%)

• Ran multiple simulations to yield distributions of outcomes

9

Recommended optionsSSB F Season closure

8M, 2

Fis

h, 2

8in

8M, 2

Fis

h8M

, 32i

nSt

atus

Quo

8M, 2

Fis

h, 2

8in

8M, 2

Fis

h8M

, 32i

nSt

atus

Quo

8M, 2

Fis

h, 2

8in

8M, 2

Fis

h8M

, 32i

nSt

atus

Quo

Lesson 2

• Don’t under-estimate the ability of stakeholders to understand complicated management questions and processes.

• Stakeholder involvement is critical at all stages of management– Objectives– Management alternatives– Evaluation

EAM

Strategic advice: Management policies that support long term fishery and ecosystem sustainability

What kind of EAM do you want?

Lesson 3: Understanding of EAM by managers and stakeholders is at a infant’s level - education is critical

Case Study: Chesapeake Bay Fishery Ecosystem Plan

• Ecosystem boundaries– ~350 fishes described in Chesapeake Bay– Variation in life history causes boundary problems

FEP: Boundary issues

• Biological– 2 Large marine ecosystems

• Human– 11 states– 3 multi-jurisdictional commissions– 2 regional management councils– 2 nations

FEP: Simple strategic advice

• Patterns based on total removals• Empirical evidence

suggests ~250,000 mt may be sustainable

• 75,000 mt of non-menhaden harvest has been sustained

Ecosystem dynamics

• Used to examine ecosystem context of single species management decisions– What would happen if the region achieved a goal of a

10-fold increase in oyster biomass?• Concerns

• Fisheries model in a “microscopic” ecosystem• My species wasn’t included• Are we modeling the tail or the dog?

• Modeling effort was not integrated into management

• Maryland Sea Grant asked to facilitate an ecosystem based process for the Chesapeake Bay

Single-species ecosystem-based FMPs

Single-species ecosystem-based reference points

• Striped bass

Case study: Atlantic menhaden• Estuarine-dependent pelagic planktivore that provides

ecosystem-services – by filtering phytoplankton– important prey resource for charismatic piscivores

• Stock migrates along coast from Florida to Maine• Juveniles use estuarine nursery areas

• Stock assessment (2006) indicates neither stock is not overfished, nor experiencing overfishing.

• Yet concerns expressed over “localized” depletion in Chesapeake Bay

Localized depletion?

• A spatially-restricted decline in abundance that impairs ecological function– Predator production– Nutrient export

• Otolith chemistry evidence suggests spatial structure

150.00

200.00

250.00

300.00

350.00

400.00

Lower Mid Upper Upper 7 Upper 9

Li C

once

ntr

atio

n (ppb)

150.00

200.00

250.00

300.00

350.00

400.00

Lower Mid Upper Upper 7 Upper 9

Li C

once

ntr

atio

n (ppb)

Nutrient export services

MS-models

• Predator-prey model– Climate-driven recruitments are strongly negatively

correlated – Stochastic, correlated SR function

1,))((

,

10

21 t

NfMat

at NeNN t

10

21,

)(,1

at

Matt PeNP a

Pa

MS – models: effects of recruitment correlation on depletion

Biophysical EM

Luo et al.

Biophysical EM Sample output

Lesson 3

• Defining the question is essential– What is localized depletion?

• Models with different outputs can’t be compared– What are the common metrics

• Reference points for non fishery endpoints are needed

Case study: Regime shifts in the northwest Atlantic

• Dramatic shift from a groundfish-dominated community to an elasmobranch-dominated community

• Multiple explanatory hypotheses

0

20

40

60

80

100

120

140

160

1963

1966

1969

1972

1975

1978

1981

1984

1987

1990

1993

1996

Year

Bio

mas

s in

dex

Gadids Flatfish Skates Other Dogfish

0

1

2

3

4

5

6

7

8

1960 1964 1968 1972 1976 1980 1984 1988 1992Year

Lan

din

gs (

106 k

g)

0

20

40

60

80

100

120F

ish

ing

effo

rt (

103 d

)Groundfish & flounders Pelagic Elasmobranchs

Winter skate length frequencies

Frisk et al. 2008

Closed Ecosystem?

• Could broadscale movement of adult skate from Scotian Shelf be responsible

Ecosystem change on the Scotian Shelf

Figures from Frank, 2003

Lesson 4:

• Even the most successful EM may not forecast the future state of the ecosystem– CBO budget forecasts have all been wrong, but

that doesn’t mean that they haven’t been useful.

• Managing expectation is critical

Recommendations• Given a goal, what approaches are feasible or useful?

– There needs to be a clear statement of what question is the model being used to ask

– Not all ecosystem questions require an EM• What comes first – goal or stakeholders?

– The stakeholders you have in the room will affect the stated goal or vision

– Engage stakeholders early and often• Strategic advice may not be of the form of control rules, e.g.,

ecosystem services– How then do EM feed into the management process,

particularly ?