Modeling synthesis

Ecosystemdrivers

Ecosystem characterization (data)

Scientific evaluation

&Stakeholderassessment

B) Guidance for future management actions

A) Guidance for future empirical efforts

Why was this project funded?

-Novel approach-Not business as usual-Uncomfortable-Confusing-May not be possible in

5 years-Potentially very exciting

$$$$

$$

$$

$$

$

$$

$

End-point #2: Fish production

Adaptive Integrated Framework (AIF): a new methodology for managing impacts of multiple stressors in coastal ecosystems

TEAM: Tomas Hook, Tammy Newcomb, Craig Stow, Scott Peacor, Steve Pothoven, Steve Brandt, Hank Vanderploeg, and Tom Nalepa

NOAA Great Lakes Environmental Research Laboratories: Craig Stow (lead), Stephen Brandt, Thomas Croley II, Julianne Dyble, Gary Fahnenstiel, Thomas Nalepa, Steven Pothoven, Henry VanderploegMichigan State University: Scott D. Peacor (lead), Michael D. Kaplowitz, Frank LupiUniversity of Michigan: Tomas Höök (lead), Dmitry Beletsky, Carlo De Marchi, Thomas Johengen, Donna KashianUniversity of Akron: Peter J. Lavrentyev Limno-Tech, Inc.: Joseph V. DepintoWestern Michigan University: Chansheng He Michigan Department of Natural Resources: Tammy J. NewcombMichigan Department of Environmental Quality: James H. Bredin

Overview

1) Background

2) Potential effects of stressors on fish production

3) Potential research approachesa) Modeling

b) Ecosystem survey and empirical research

4) Issues to consider

Yellow perch

Walleye

Percids in Saginaw Bay

Historically second largest commercial fishery in Great Lakes.

Walleye collapse ~1950 due to eutrophication, over-fishing, invasive species, etc.

Recent collapse of L. Huron alewives has led to improved reproductive success, but very poor growth and long-term survival.

Fielder and Thomas 2006 MI-DNR; Fielder et al. 2007 JGLR

Percid Production Requires a Balance Within the Ecosystem

Yellow perch

ZooplanktonBenthic invertebrates

Invasive Species

Moderate temperatures

Mesotrophic conditions

Balance between sufficient pelagic and benthic prey

Overview

1) Background

2) Potential effects of stressors on fish production

3) Potential research approachesa) Modeling

b) Ecosystem survey and empirical research

4) Issues to consider

WestR2 = 0.69p = 0.0002

April-May Maumee R. discharge(m3 / sec)

100 200 300 400 500

Lo

g10

(ag

e-2

yell

ow

per

ch)

(mil

lio

ns)

1

10

100

'87

'88

'89

'90

'91

'92

'93'94

'95

'96

'97

'98'99

'00

Ludsin et al, unpub. data

Loading effects on percid production

Recruitment linked to river discharge.

Mechanisms unclear.

Too much loading probably bad.

Ecosystem engineering and the disrupted food web of Saginaw Bay.

The substrate provided by the shells provides substrate for benthic plant and invertebrates, and increased light increases benthic plant production as well as potentially increasing predation intensity of visual (invertebrate and vertebrate) predators.

The red shaded species are non-indigenous species.

Invasive species effects on percid production

Black dot:Ponar samples for total benthos(1987 and 1988)

Circled black dot:Ponar samples for total benthos(1987-1996)

X:diver samples for zebra mussels(1991-1996)

Location of sites sampled for benthos in 1987-1996.

Year

1991 1992 1993 1994 1995 1996

Den

sit

y (

No

. m-2

)

0

10000

20000

30000

40000

Bio

mass (

g A

FD

W m-2

)

0

20

40

60

80

100

Density

Biomass

Mean density and biomass of zebra mussels at sites with hard substrate in inner Saginaw Bay in 1991-1996.

Year

1986 1988 1990 1992 1994 1996

Bio

mass (

g A

FD

W m-2

)

0.0

2.0

4.0

6.0

8.0

10.0

Mean biomass of non-dreissenids at deep, silty sites in inner Saginaw Bay (1991-1996)

TotalChironomids

Year

1986 1988 1990 1992 1994 1996

Bio

mass (

g A

FD

W m-2

)

0.0

1.0

2.0

3.0

4.0

Non-dreissenid biomass; Shallow, sandy sites

TotalGammarus

Climate change effects on percid production

• Temperature is the “master variable” for fish

• May also affect:

– Watershed loading

– Water levels

– Community composition

Yellow perch bioenergetics

24oC

11oC

15oC

• Respiratory cost(g g-1 d-1)

• 0.0053

• 0.003

• 0.002

• Respiratory cost (g g-1 d-1)

activity (x2)

• 0.011

• 0.006

*Model run for a 20 g adult yellow perch (5942 J/g); Chironomid (3138 J/g); Zoop. (2510 J/g)

Yellow perch bioenergetics

24oC

11oC

15oC

• Maintenance consumption(g/d)

• 0.934

• 0.517

• 0.301

• Maintenanceconsumption (g/d)

activity (x2)

• 1.89

• 1.05

*Model run for a 20 g adult yellow perch (5942 J/g); Chironomid (3138 J/g); Zoop. (2510 J/g)

Overview

1) Background

2) Potential effects of stressors on fish production

3) Potential research approachesa) Modeling

b) Ecosystem survey and empirical research

4) Issues to consider

Fish Models

Coupled bio-physical 3D

models

Empirically based

(Bayesian)

Simple statistical (e.g.

regression)

Artificial Neural

network

EcosystemStressors

Eco

syst

em

Mo

del

s

Ecosystem Characterization

Fish community dynamics

Water quality &

Human health

Eco

syst

em

en

dp

oin

ts

• Land & Resource use

• Climate Change• Invasive Species

• Watershed model• Hydrodynamic model• Biophysical data and

processes

Recommendations for ecosystem characterization•Experimental•Monitoring•Synthesis

Socio-economic integration to guide management•Economic models•Public preference •Workshops

MI-DNR Saginaw Bay Surveys

• Spring sampling of river spawning walleye (1981-present)

• Fall bay-wide surveys of fish populations– Gillnet (1989-present)– Bottom trawls (1971- or 1986-present)– Abundance, size, condition, age, diets

• Smattering of additional data from 1926-present

Empirical models

1) Regression models -(e.g., Fielder et al. 2007 JGLR)

2) Neural networks -(trained on meta-data and then applied)

3) Bayesian probability networks -(hocus-pocus vodoo)

Coupled 3-D process –based models

• Saginaw Bay Ecosystem Model (SAGEM)– Deterministic, process-based model– Phytoplankton model first developed in 1970’s– Updated to include:

• Multi-class phytoplankton• Zebra mussel bioenergetics• PCB fate, transport and lower food web bioaccumulation• Benthic algae productivity

– Latest version: Bierman et al. 2005 JGLR 2005

• Will be coupled to fish individual-based model

Next Individual

Add new individuals

Forageƒ(fish size, temperature, prey densities,

water clarity)

Respire (bioenergetics)ƒ(fish size, temperature, food consumed)

Predation Mortalityƒ(fish size, current location)

Starvation Mortalityƒ(fish size, % storage tissue)

Move?ƒ(fish size, current location)

Next Day

Loop through individuals

Swimming distance

Reactive distance and encounterrate

Optimal foraging

Foraging:

Stochastic capture success Consumption

Overview

1) Background

2) Potential effects of stressors on fish production

3) Potential research approachesa) Modeling

b) Ecosystem survey and empirical research

4) Issues to consider

Fish-related ecosystem chracterization

Coupled bio-physical 3D

models

Empirically based

(Bayesian)

Simple statistical (e.g.

regression)

Artificial Neural

network

EcosystemStressors

Ec

os

yste

m

Mo

de

ls

Ecosystem Characterization

Fish community dynamics

Water quality &

Human health

Ec

os

yste

m

en

dp

oin

ts

• Land & Resource use

• Climate Change• Invasive Species

• Watershed model• Hydrodynamic model• Biophysical data and

processes

Recommendations for ecosystem characterization•Experimental•Monitoring•Synthesis

Socio-economic integration to guide management•Economic models•Public preference •Workshops

Fish samplingCollect (monthly spring-fall)

1) Young percids 2) Potential interacting biota

preycompetitorspredators

Fish samplingQuantify:fish sizeagedietsgrowth (incl. short-term)mortality

Relate to:available preyphysical conditionspredation pressure

Dreissenid Abundance and Biomass:Needed for estimates of nutrient excretion, filtering capacity, and materials cycling (carbon, energy).

Key question: has the population remained stable since 1994-1996?

Macroinvertebrate Abundance and Biomass:Needed for estimates of food available to fish.

Emphasis on the dominant taxa: (Chironomidae at deep sites and Gammarus sp. at shallow sites.)

Proposed Effort for Benthic Community

New Links—emphasizing the microbial food web (conceptual) and invaders (real)

Zooplankton sampling

Vertical tows of 64µm net

Collect small zooplankters

Oblique tows of 153µm net

Collect large zooplankters

Experimental Research

Examples:

• Effects of water clarity on fish foraging

• Effects of dreissenid shells on fish foraging efficiency

• Effects of dreissenids on P-availability

Overview

1) Background

2) Potential effects of stressors on fish production

3) Potential research approachesa) Modeling

b) Ecosystem survey and empirical research

4) Issues to consider

Nearshore sampling

• Nearshore areas potentially important nursery grounds

• Nearshore sampling not described in proposal

• How will we consider these habitats?

Invasive species

• Proposal refers to invasive species, but really focuses on dreissenids

• Is the invasive species stressor really just a dreissenid stressor?

• Or, will we explicitly consider effects of other stressors?– Bythotrephes– Round gobies– Phragmites