Modeling Atlantic menhaden in support of

nutrient and multispecies management in Chesapeake Bay

Mark J. BrushRobert J. Latour

Elizabeth A. Canuel

Virginia Institute of Marine ScienceCollege of William and Mary

Chesapeake 2000 calls for …

• Assessment of the potential impacts of filter feeders on bay water quality.

• Continued nutrient loading reductions

• A move towards multispecies, ecosystem-based management* These issues are linked by

the transfer of nutrient-fueled primary production up the

food chain via filter feeders such as Atlantic menhaden …

Atlantic menhaden

PiscivoresOther forage spp.

PlanktonNutrients

Top-down

effects

Bottom-upeffects

Menhaden-based food web:

striped bass bluefish weakfish

Atlantic menhadenbay anchovy

Approaches: • Ecopath with Ecosim• Bioenergetics

PHYTOPLANKTON ZOOPLANKTON

FISHING

• Ecopath with Ecosim (EwE)

Types of Multispecies Models

• Multispecies Bioenergetics Models (MSBE)

• Multispecies Production Models (MSP)

• Multispecies Virtual Population Analysis

(MSVPA)

Latour, Brush & Bonzek (2003)Toward ecosystem-based fisheries

management: strategies for multispecies modeling and

associated data requirements.

Fisheries 28(9):10-22

Ecopath

• Parameterization of an Ecopath model is based on satisfying two ‘master’ equations:

1) Production = catch + predation + net migration + biomass accumulation + other mortality2) Consumption = production + respiration + unassimilated food

• Master equations translate to:

Biomass of i

ji

n

jjjiiiiii DCBQBBAEYEEBPB

1

)/()/(

Total mortality

Ecotrophic efficiency

Yield

Net migration

Biomass accum

Biomass of predator j

Food consumed per unit biomass of predator j

Fraction of i in diet of j

MSP MSVPA EwE MSBE

Age / size structure no yes yes yes

Biomass predictions yes yes yes yes

Data requirements low high high high

Mass / energy balance no no yes yes

Network analysis no no yes no

Number of species low intermediate high intermediate

Model output intermediate quantitative qualitative intermediate

Physiological information no no limited yes

Predictions with variable F yes yes yes yes

Spatial resolution no no possible possible

Taxonomic resolution species or groups

species species or groups

species

Temporal resolution annual annual annual daily

Why bioenergetics?

Source:Latour et al.

(2003)

Types of Multispecies Models

Bioenergetics Approach

… piecing together environmental data and biological relationships

C

TEMPERATURE

R

TEMPERATURE

C

FOOD

GROWTHOR

LOSS

Example: Atlantic menhaden

INPUTSconsumption (C)

OUTPUTSrespiration (R)

feces (F)excretion (U)

specific dynamic action (S)

Growth = C – R – F – U – S

MULTISPECIES BIOENERGETICS MODELS (MSBE)

Consumption IndividualWeight (W)

W

C

Respiration

W

R

T

Ractivity

multiplier

EgestionC * (1-assimilation

efficiency)Excretion

C * fraction excreted

Specific Dynamic ActionC * SDA coefficient

Predation

B

Ppredator

abundance

day

T

T

C

F

C

day

T

day

F

B

PopulationSize (N)Recruitment

Background mortality

Fishing mortality

C = Cmax * f(T) * p

Cmax = aWb

R = Rmax * f(T) * ACT

Rmax = aWb

Menhaden: Rippetoe (1993) Gottlieb (1998) Luo et al. (2001)

Bay anchovy: Luo & Brandt (1993)

Piscivores: Hartman & Brandt (1995)

Estuarine Food Web

SmallPhytoplankton

LargePhytoplankton

Micro-Zooplankton

Meso-Zooplankton

MicrobialLoop

SeaNettles

CombJellies

FishLarvae

ForageFish

PredatoryFish

Cloern, J.E. 2001. “Our evolving conceptual model

of the coastal eutrophication problem”

Mar Ecol Prog Ser 210:223-253

Nutrients

Climate

?

?

Fishing?

Systems-level process models

Why bioenergetics?

MULTISPECIES BIOENERGETICS MODELS (MSBE)

Individual models for:

Menhaden Juveniles Age-0 Adults Age-1 Age-2 Age-3 Age-4 Bay anchovy Juveniles Age-0 Adults Age-1 Striped Bass Juveniles Age-0 Residents Age-1 Age-2 Age-3 Age-4 Migrants Age-5 Weakfish Juveniles Age-0 Adults Age-1 Age-2 Bluefish Juveniles Age-0 Adults Age-1 Age-2

Population models for:Menhaden Juveniles

Menhaden Adults

Bay anchovy

Striped Bass Juveniles

Striped Bass Residents

Striped Bass Migrants

Weakfish JuvenilesWeakfish Adults

Bluefish JuvenilesBluefish Adults

Groups in EwE model:

Starting point: Individual Bioenergetics Modelsfor each Species & Age-Class

2:44 PM Wed, Aug 13, 2003

0.00 182.50 365.00 547.50 730.00

Days

1:

1:

1:

2:

2:

2:

0.00

20.00

40.00

1: Wind[Men0] 2: vBavg[Men0]

1

1

1

1

2

2

2

2

growth: p1 (Untitled)

Age-0 Menhaden

modelaveragetrajectory

2:44 PM Wed, Aug 13, 2003

0.00 182.50 365.00 547.50 730.00

Days

1:

1:

1:

2:

2:

2:

3:

3:

3:

4:

4:

4:

0.00

200.00

400.00

1: Wind[Men2] 2: vBavg[Men2] 3: vBmax[Men2] 4: vBmin[Men2]

1

1

1

1

2

2

2

2

3

3

3

3

44

44

growth: p3 (Untitled)

Age-2 Menhaden

min & max

trajectories

MULTISPECIES BIOENERGETICS MODELS (MSBE)

Age-specific individual models

Average rates for each species/group

Population-level biomass models

Predator consumption

Fishing Mortality(limited to season)

Non-predation, natural mortality

Diet compositionor

foraging model

Prey loss

Modified byco-occurrence

• In Ecosim, this density-dependent effect is modeled as:

Ft =QREt/[1+(QR-1)Bt/Bo] where QR=qmax/qo

is the specified catchability increase ratioBt = stock biomass at time t Bo = Ecopath base biomass

Model calibration – Ecosim• Menhaden QR = 3.50

Some Key Limitations

• Need Bay-specific stock assessments for:- abundance- biomass- fishing mortality (F)

• Need to refine menhaden feeding ecology

Current CBP study

Objective 1:Menhaden Stock Assessment

• R. Latour in collaboration with S. Martell and NMFS Beaufort Lab

• MDDNR and VIMS seine survey data (juveniles)• Fisheries-independent trawl data (adults)• Bay-specific landings • Simple biomass-based (e.g. surplus production) models • More complex age-structured models (e.g. VPA, catch-at-age, etc.) • As much spatial and seasonal resolution as possible

Timing: Year 1, with updates in Years 2-3

MDupper

MDmid

MDlower

VAlower

VAupper

PatuxentRiver

PotomacRiver

RappahannockRiver

YorkRiver

JamesRiver

MDupper

MDmid

MDlower

VAlower

VAupper

PatuxentRiver

PotomacRiver

RappahannockRiver

YorkRiver

JamesRiver

MDupper

MDmid

MDlower

VAlower

VAupper

PatuxentRiver

PotomacRiver

RappahannockRiver

YorkRiver

JamesRiver

Objective 1:Menhaden Stock

Assessment

Goals:

• Age Resolution – by year class

• Temporal Resolution:AnnualSeasonalMonthly (?)

• Spatial Resolution:BaywideSpatial elements

• Output: Abundance, biomass, & F

Objective 2:Feeding & Diet Study

• Stable isotopes in field-caught fish and prey sources:

- size-fractionated plankton, SAV & Spartina detritus

- long-term, integrated diet composition

• Expt 1: Diet selectivity• Expt 2: Functional response (consumption vs. food conc.)• Expt 3: Density dependence (consumption vs. fish density)• Expt 4: Excretion rate

Timing: Years 1-2

Menhaden Feeding Experiments

1. Starvation period (48 hours, 1 m filtered water)

2. Sample 3 fish stomachs to confirm guts are empty

3. Early A.M. - Fill tanks (n=20) with raw seawater from pier - Collect samples for size-fractionated Chl-a & CHN, TSS (pre- and post-combustion), particle size distribution, size-fractionated isotopes, and lipids

4. Expt start: place fish in tanks (1 per tank)

5. Expt end: collect samples from each tank for … - Size-fractionated Chl-a & CHN, TSS (pre- and post-combustion), particle size distribution

? Decide on expt duration & mid-point sampling

- Stomach contents- microscopic ID

- stable isotopes- lipid biomarkers

Pilot Experiments

0

2

4

6

8

10

12

14

16

18

20

t0 t10 t20 t30 t40 t50 t60

TIME, minutes

CHL-

a ,

mg m

-3

1 23 45 67 89 10Control

Tank:

2002 CBP Suspension Feeder Workshop

• Research and monitoring• Basic consumption estimates• Management-relevant intermediate complexity models• Application of the 3-D water quality model

… called for a multi-faceted approach to estimating the

effects of suspension feeders on water quality …

Thisproject

Objective 3:Basic Consumption Estimates

• Individual daily rations (literature – seasonal variations)• Stock assessment (annual, seasonal, and regional abundance)• Planktonic food supply (Brush et al., C. Buchanan)

Timing: Year 1

0.0

0.4

0.8

1.2

1.6

J F M A M J J A S O N D

g C

m-3

PC

phyto-plankton detritus

microzooplankton mesozooplankton

a.

0.0

1.0

2.0

3.0

J F M A M J J A S O N D

g C

m-3

b.

ug/L2

30

MDupper

MDmid

MDlower

VAlower

VAupper

PatuxentRiver

PotomacRiver

RappahannockRiver

YorkRiver

JamesRiver

MDupper

MDmid

MDlower

VAlower

VAupper

PatuxentRiver

PotomacRiver

RappahannockRiver

YorkRiver

JamesRiver

MDupper

MDmid

MDlower

VAlower

VAupper

PatuxentRiver

PotomacRiver

RappahannockRiver

YorkRiver

JamesRiver

Objective 3:Basic Consumption

Estimates

Goals:

• Age Resolution – by year class

• Temporal Resolution:AnnualSeasonalMonthly (?)

• Spatial Resolution:BaywideSpatial elements

Objective 4:Multi-Layered MSBE & EwE Modeling

Timing: Years 1-2

• Refine with stock assessment and experimental results• Force plankton concentrations• Daily, seasonal, & annual consumption, N cycling, and N export

Objective 4:Multi-Layered MSBE & EwE Modeling

Timing: Years 1-2

• Add dynamic feedbacks with plankton populations

• Effect of menhaden cons. & excr. on plankton biomass & productivity• Fraction of menhaden yield supported by Bay production• Run over range of menhaden biomass & nutrient loading

Objective 4:Multi-Layered MSBE & EwE Modeling

Timing: Years 1-2

• Coupling to piscivores; full MSBE development

• Role of menhaden as a forage base• Effect of fisheries management scenarios

Objective 4:Multi-Layered MSBE & EwE Modeling

Timing: Years 2-3

• Full coupling

• Simultaneous impact of multiple stressors (a la Cloern 2001)• Indirect effects• Comparison of nutrient and fisheries mgmt effects on menhaden yield & control of water quality

Objective 5:Comparison of Menhaden, Zooplankton,

Oyster, and Clam Filtration

Timing: Year 3

• Add zooplankton, oyster, and clam information to MSBE and EwE models

• Run for different regions of the Bay