11-17-2016

The South Atlantic LCC’s Third Thursday Web Forum

Understanding marine ecosystem changes in the South Atlantic region using an integrated modeling approach

Call in for audio: 866-720-8724 2917595555#

• Introduction

• Monthly topic

• Questions and discussion about presentation

• South Atlantic LCC staff updates

• Questions and discussion about updates

2

Agenda

3

Tom Okey, University of Victoria

Understanding marine ecosystem changes in the South Atlantic region using an integrated modeling approach

11-17-16

AN INTEGRATIVE ECOPATH WITH ECOSIMMODEL FOR THE SOUTH ATLANTIC REGION FOR UNDERSTANDING COASTAL MARINE

ECOSYSTEM CHANGESTom Okey, Ocean Integrity Research

Roger Pugliese, South Atlantic Fishery Management Council

South Atlantic LCC Fall Seminar seriesOctober 17th, 2016

FOOD WEB MODEL

ECOPATH WITH ECOSIM

No fish is an island

7000+ users in 150+ countries (google analytics)

800+ peer-reviewed publication (ISI Web of Knowledge)

ECOPATH WITH ECOSIM

ECOPATH / ECOSIM / ECOSPACE

ECOPATH with ECOSIM (EwE) is a suite of combined models

developed for ecosystem and fishery analysis. These commonly

used programs can explore management or policy options and

address growing ranges of fisheries and ecological questions.

THE ECOPATH MODEL OF THE SOUTH

ATLANTIC BIGHT

SOUTH ATLANTIC BIGHT FOOD WEB

Figure by Kelly Kearney

BROAD PROJECT OBJECTIVES

A. Advance and integrate existing models

and data

B. Advance and refine the LCC

conservation blueprint

C. Support the SA Fishery Management

Council’s move to ecosystem-based

management and its Fishery Ecosystem

Plan

To build a flexible and integrative ecosystem modelling

suite for the South Atlantic Region

HOW DO WE GET THERE?

Establish an inter-disciplinary regional science team to ensure

the approaches taken include leading-edge thinking and

produce the best possible information for resource decision-

making

Work with managers and decision-makers to ensure models

meet their needs for conservation, management, and science

planning

Highlight the connectivities and dynamics between

coastal/ocean and estuarine-riverine ecosystems

Develop ecosystem models to understand ongoing changes in

the system and develop effective management strategies

OUR PROJECT TEAMS– Principal investigators –

Dr. Marcel J. Reichert – Marine Resources Research Institute, South Carolina Department of Natural Resources

Dr. Luiz Barbieri – Florida Fish and Wildlife Conservation Commission, FWRI

Dr. Thomas Okey – Ocean Integrity Research, Victoria, BC, Canada

Dr. Jerald S. Ault – Professor and Director, Fisheries Ecosystem Modeling and Assessment Research, Rosenstiel School of Marine and Atmospheric Science, University of Miami

Dr. Ruoying He – Department of Marine, Earth and Atmospheric Sciences, North Carolina State University

Dr. Peter Sheng – Professor and Director, Coastal and Oceanographic Engineering Program, University of Florida

Vembu Subramanian – RCOOS Program Manager and Data Management Coordinator, SECOORA

Dr. Patrick N. Halpin – Director, Geospatial Analysis Program Duke University

Roger Pugliese – Senior Fisher Biologist, South Atlantic Fishery Management Council

THE MODELS WILL:

Link to hydrodynamic oceanographic models and satellite data

Provide more realistic predictions about spatial policy options

Predict impacts of episodic events that are limited in space (oil spills, red tides, upwelling)

Meet the immediate needs of the SSC and the South Atlantic Council

PHASE 1- UPDATE

SOUTH ATLANTIC

ECOPATH MODEL

Element 1Develop South Atlantic Ecopath Model - Tom Okey

Element 2Data Conditioning- Marcel Reichert

Element 3Review of Estuarine Data and Models - Peter Sheng

Element 4A Coupled Marine Environmental Assessment and Prediction System for the Southeastern U.S. Coastal Ocean in Support of Effective Marine Ecosystem-Based Management - Ruoying He

Element 5South Atlantic Fisheries Ecosystem Modeling & Prediction- Jerald S. Ault

PHASE 2 –CONNECT TO OTHER MODELS

Respiration

Predation

Yield

Net migration

UnassimilatedOther mortality

Food

consumptionProduction

Polovina, J.J. 1984. Coral Reefs, 3:1-11; Pauly et al. 2000. ICES J. Mar. Sci., 57: 697-706; Christensen and Walters. 2004. Ecol. Model., 172(2-4): 109-139

Bi Biomass

Pi/Bi Production

Qi/Bi Specific consumption

DCij Fraction of prey i in diet of predator j

BiAi Biomass accumulation

EEi Production used in the system

1-EEi Unexplained mortality

ii

i

iii

n

1Pred_jijj

j

i

i

EE1BB

PBAYEDCB

B

QB

B

P

1.

2.

iii

i

i

i

UNRBB

PB

B

Q

ECOPATH

Diet compositione.g., for a tuna

Use volume or weight!

Partly digested

fish 31.6%

Others 19.3%Portunids 15.8%

Euphausiids 3.5%

Squids 12.3%

Anchovies 8.8%

Sardines 7%Auxids 1.7%

HISTORY OF THE SAB MODEL

2001 - Strawman 48-box model constructed

2004 - Preliminary 98-box model developed

2013 - Model refined to address forage fish

questions

Construction

& refinement

Sponsored by

SAFMC

42-box model

98-box model

Application to

forage question

Sponsored by

Pew Charitable

Trusts

Forage groups

articulated

99-box model

COLLABORATIVE CONSTRUCTION

Adult Herring

Arrowtooth Flounder

CetaceansPinnipeds

SablefishLingcod

Pacific Cod

Seabirds

Omnivorous Zooplankton Herbivorous

Zooplankton

Shallow Small

Epifauna

Fishery

Primary contributors Behzad Mahmoudi (FMRI) Bob Feller (USC) David Whitaker (SCDNR) Doug Vaughan (NMFS) Marty Levissen (SCDNR) Jack McGovern (NMFS) Larry DeLancey (SCDNR) Bill Sharp (FMRI) Whit Gibbons (UGA) Joan Browder (NMFS) John Carlson (NMFS) Larry Cahoon (UNC) Galen Johnson (UNC)

Megan Gamble (ASMFC) Brad Spear (ASMFC) Toni Kearns (ASMFC) Peter Verity (SKIO) Wilson Laney (USFWS)

Secondary contributors

Elizabeth Wenner (SCDNR)

Robert George (GIBS)

Carolyn Currin (NOAA)

Chuck Hunter (USFWS)

Craig Watson (USFWS)

Damon Gannon (Mote Lab)

Desmond Kahn (DEDNR)

Enric Cortez (NMFS)

George Sedberry (SCDNR)

Greg McFall (GRNMS)

Hans Paerl (UNC)

Jennifer Wheaton (FMRI)

Jenny Purcell (WWU)

Jim Nance (NMFS)

John Merriner (NOAA)

Doug Forsell (USFWS)

Jon Hare (NOAA)

Jose Castro (Mote Lab)

Ken Lindeman (ED)

Mark Epstein (USFWS)

Martin Posey (UNC)

Paul Carlson (FMRI)

Steve Ross (UNC)

Buddy Powell (WT)

Alan Bolten (UFL)

Karen Bjorndal (UFL)

Bob Noffsinger (USFWS)

Sean McKenna (NCDENR)

Pat Tester (NOAA)

Lance Garrison (NMFS)

Species / Groups in SAB 99-box model

Dogfish sharks Demersal coastal omnivores Birds -- shelf piscivores Benthic meiofauna

Adult mackerel Benthic oceanic piscivores Rock shrimps Deep-burrowing infauna

Juvenile mackerel Benthic oceanic invertivores Penaeid shrimps Carnivorous zooplankton

Bluefish Benthic coastal piscivores Megafaunal predators Aquatic and other insects

Weakfish Benthic coastal invertivoresEchinoderms and

gastropodsOther zooplankton

Red drum Benthic coastal planktivoresEstuarine infaunal

crustaceansIchthyoplankton

Atlantic menhaden Reef associated piscivores Birds -- herbivores Microbial heterotrophs

Mullets Reef associated omnivores Birds -- wading piscivores Phytoplankton

Other Drums & Croakers Triggerfish Birds -- shelf invertivores Microphytobenthos

Striped bass Shallow water grouper/tilefish Birds -- raptors Benthic macroalgae

Highly migratory pelagics Goliath grouper Encrusting fauna Pelagic macroalgae

Dolphinfish Nassau grouper Squids Seagrasses

Pelagic oceanic

piscivoresDeep-water grouper/tilefish Stomatopods Marsh vegetation

Pelagic coastal piscivores Shallow-water snapper Octopods Estuarine benthic detritus

Nearshore piscivores Mid-shelf snapper Blue crabs Offshore benthic detritus

Pelagic oceanic

planktivoresJacks Horseshoe crabs Water-column detritus

Sardines Red porgy Golden crabs Dead carcasses

Scads Grunts and porgys Calico scallops

VALUED FISH SPECIES

FORAGE GROUPS IN THE 99 BOX MODELGroup Species included B

(tˑkm-2)P/B

(year-1)Q/B

(year-1)Anchovies Bay (Anchoa mitchilli), striped (A. hepsetus), silver

(Engraulis eurystole)

3.75 1.45 17.50

Atlantic menhaden Brevoortia tyrannus (not B. patronus) 7.05 1.70 7.84

Atlantic silverside Menidia menidia 1.18 2.00 14.90

Halfbeaks Ballyhoo (Hemiramphus brasiliensis), balao (H.

balao), common or Atlantic silverstripe

(Hyporhamphus unifasciatus)

1.22 2.60 11.70

Mullets Striped (Mugil cephalus), other (Mugil spp.) 0.11 0.70 11.03

Sardines Spanish (Sardinella aurita), scaled (Harengula

jaguana)

1.93 1.11 11.82

Scads Round (Decapterus punctatus), rough (Trachurus

lathami), bigeye (Selar crumenophthalmus)

2.28 0.92 10.00

Shad Alosa spp. 3.97 0.50 3.80

Thread herring Atlantic thread herring (Ophistonema oglinum) 0.28 1.60 13.26

Pelagic oceanic

planktivores

Chub mackerel (Scomber japonicus), lanternfish

(Diaphus spp.), antenna codlet (Bregmaceros

atlanticus), striated argentine (Argentina striata),

flyingfish (Exocoetidae)

3.95 0.87 11.71

Squids Shortfin (Illex illecebrosus), longfin (Loligo pealei) 0.45 2.67 36.50

Shrimps Rock shrimps and penaeid shrimps 2.53 5.38 19.20

NEW 99 BOX SAB MODEL (FORAGE)

ELEMENT 2: SC DEPT. OF NATURAL

RESOURCES LONG-TERM SURVEY DATA

CONTRIBUTIONS

Coastal Trawl Survey

SEAMAP-SA

Annual, seasonal

Nearshore (depths <10m)

75 ft. Falcon trawls

Common taxa:

Croakers, mackerels, anchovies

Penaeid shrimp

Coastal sharks

Rays

SE Reef Fish Survey

Begun by MARMAP, now with SEAMAP-SA and SEFIS

Annual (May-Sept.)

Offshore (15-300 m)

Traps, longlines

Common taxa:

Groupers and sea basses

Porgies

Snappers

Triggerfishes

Element 3: Reviews of Estuarine Data and Models

Element 3: Reviews of Estuarine Data and Models

Peter Sheng is looking at a few estuaries in Florida

Relate fishery data with salinity and other parameters obtained from numerical model simulations using dynamic models.

St. John's River Apalachicola Bay Rookery Bay

Element 4A Coupled Marine Environmental Assessment and Prediction System for the Southeastern U.S. Coastal Ocean in Support of Effective Marine Ecosystem-Based Management - Ruoying He

ECOSIM

Temporal-dynamic module of EwE, initialized from Ecopath

Includes biomass and size structure dynamics

Requires only a few extra parameters

Used, among others, to assessQuantify combined effect of species dynamics, fishing

impacts, and environmental impacts on a food web over time

Replicate past scenarios (time series fitting)

Explore future scenarios

Explore fishing policy alternatives

Test model robustness

ECOSIM

Walters et al 1997 RFBF, Ahrens et al 2012 Fish and Fisheries

Foraging arena:

CALIBRATING THE MODEL

TIME PREDICTIONS FROM AN ECOSYSTEM

MODEL OF THE GEORGIA STRAIT, 1950-2000

Okey, T. A. 2004. Can oil spills shift marine ecosystems to alternate stable states?: Preliminary simulations with an Ecopath model of Prince William Sound,

Alaska. Pages 84-103 In: Shifted community states in four marine ecosystems: some potential mechanisms. PhD. University of British Columbia, Vancouver.

Prince William Sound, Alaska

1

2

3

0Bio

mas

s / o

rigi

nal b

iom

ass

0 5 10

System-wide changes in biomass

Time (years)(V = 3)

REMOVAL OF A SEA CUCUMBERS

EXPLAINS GALÁPAGOS ANEMONE

BARRENS

Anemones

Damselfishes

Hexaplex gastropod

Sessile filter and suspension

Pepino (S. fuscus)

Fishing zone

No fishingzone

Island

Sea cucumber recovery on a

Galápagos rocky reef

SPATIAL SIMULATION

ECOSPACE: SEASONAL OR FULL-TIME CLOSURES

Advancing Ecospace

Jeroen Steenbeek, Ecopath International Initiative

ECOSPACE

Spatial temporal component of EwE, executes

Ecosim for every ‘water’ cell in a grid

Requires extra inputs, related to movement,

habitat, fishing, environment

Groups and fleets try to move to

nearby optimal conditions

ECOSPACE

ECOSPACE

ECOSPACE

Used, among others, to assess

Distribution of marine species and fishing effort

Spatial impact of fishing

Management options, e.g. impact of MPAs

Impact of environmental change (EwE version

6.3+)

Running model has been linked to Marxan & Atlantis

Includes an IBM approach

ECOSPACE

2011 Ecospace had three major limitations

1. Unable to represent sub-cell features

2. Unable to explicitly incorporate environmental

effects on species: “why are the species where

they are?”

3. Limited facilities to exchange data with the

outside world, thus unable to include

environmental variability

RECENT DEVELOPMENTS IN ECOSPACE

1. REPRESENTING SUB-CELL FEATURES

PROBLEM

Cells may contain small but important features that

cannot be represented well, such as coral reefs

Old Ecospace accepted only one habitat type per

cell, and yes/no habitat usage

Using ever smaller cells is no option: computationally

demanding and wasteful, with risk of over-

representing features and dynamics

Example: small but ecologically important coral reef

SOLUTION

Ecospace (EwE version 6.3+) allows habitats to

overlap, and habitats can occupy a fraction

[0,1] of a cell

Species can utilize a fraction [0, 1] of each

habitat type

Old Ecospace models are directly translated

to this structure, and work as expected

2. WHY ARE SPECIES WHERE THEY ARE?

HABITAT CAPACITY MODEL

Christensen et al. 2014. Ecosystems, E17, 1397-1412.

Original Ecospace could not explain species distributions. Habitat usage was an aggregated assumption implying environmental preferences

Ecospace (EwE version 6.3+) defines spatial foraging arena size from species’ response to environmental conditions

Ecospace has become an integrated food-web / species envelope model

HABITAT CAPACITY MODEL

Dynamic habitat model predicts how productive individual cells

are for each species, based on multiplicative effect of

environmental responses

Christensen et al. 2014. Ecosystems, E17, 1397-1412.

SETTING UP THE

HABITAT FORAGING CAPACITY MODEL

1. SELECT GROUP CAPACITY MODEL

2. DEFINE ENVIRONMENTAL DRIVERS

3. POPULATE ENV. DRIVER MAPS

4. DEFINE ENV. RESPONSE CURVES

HAB. CAP. CASE STUDY

Full Mediterranean EwE model

90+ functional groups, assigned to 4 MSFD zones

Time frame 1950 – 2010

Entire basin at 0.167 dd grid

Piroddi et al (in progress)

Foraging habitat capacity model case study

Primary production

Salinity (surface and bottom)

Temperature (surface and bottom)

Depth

MSFD area restrictions

HAB. CAP. CASE STUDY

1. Define environmental drivers

HAB. CAP. CASE STUDY

Piroddi et al (in progress)

2. Define environmental responses

Here we are using a plug-in to import environmental

responses from AquaMaps species envelopes

HAB. CAP. CASE STUDY

4. Ecospace computes capacity (cetaceans - depth)

HAB. CAP. CASE STUDY

4. More capacity (Western sardine - depth, MSFD W)

HAB. CAP. CASE STUDY

4. Run

ECOSPACE SPIN-OFF

Functional responses are also available to

Ecosim

Previous Ecosim only supported two

normal distributed response functions

New Ecosim shares functional response

curves with Ecospace

Environmental drivers in Ecosim are

functions; not maps

Affects foraging arena size over time

System is backwards compatible with

older models

Not published yet, application in

progress in PhD thesis

3. EXCHANGING DATA

WITH THE OUTSIDE WORLD

Ecospace internal data model was hard to access• Almost impossible to vary input maps over time

• Almost impossible to exchange data with other

models

• Changing environmental conditions could not be

included in spatial temporal analysis

ECOSPACE SHORTCOMINGS

SPATIAL TEMPORAL DATA FRAMEWORK

Steenbeek et al. 2013. Ecological Modelling 263, 139-151.

SPATIAL TEMPORAL DATA FRAMEWORK

SPATIAL TEMPORAL DATA FRAMEWORK

CASE STUDY

Steenbeek et al (2013) Ecological Modeling

GIS DATA FOR MANY ECOSPACE LAYERSConnected to existing Ecospace driver

layers

Primary production

Environmental drivers

Habitats

Fishing cost

MPA layouts

Contaminants

Migration

Computed foraging capacity

Coming soon

Advection

1. CALCULATE ADVECTION

CURRENT STATUS

Connected to most Ecospace layers

Reads and writes 20+ GIS data formats, both raster and vector

Designed upon EwE plug-in system. Easy to extend with new capabilities

Not (yet) publicly released: more R&D needed.

Can only be applied by directly involving EwEteam

Needs better support for projections

Needs integration of new data types

Needs user testing to streamline workflow

2. CHOOSE ADVECTED GROUPS

ADVECTION

Computed advection, per month, is stored

with the EwE model

Advection patterns can be overwritten with

external hydrological model output via the

spatial temporal data framework

Will be made available in a next EwE version,

but development version can be applied by

involving the EwE development team

Publication is in progress

The evolution of the South Atlantic Conservation Blueprint

Next Third Thursday Web Forum

1-19-2017

10:00 am

Rua Mordecai

Science Coordinator, South Atlantic LCC

4

5

How to get involved with your cooperative

• Join the South Atlantic LCC web community

• Connect with a staff or other cooperative member

• Explore the Conservation Blueprint

southatlanticlcc.org/blueprint

southatlanticlcc.org

6 6

Questions? *6 to unmute