The Atlantis Ecosystem Modeling Framework

1
The Atlantis Ecosystem Modeling Framework

• Isaac Kaplan, Chris Harvey, Phil Levin, Jason
  Link, Howard Townsend
• NOAA NMFS
• Beth Fulton CSIRO Australia

2

• Ecosystem models can improve our understanding of
  interactions between species, climate, fishing,
  and habitat.
• The Atlantis ecosystem model (Fulton et al. 2004)
  is a strategic tool used to1. synthesize this
  information 2. simulate possible ecosystem
  responses3. identify key processes that govern
  ecosystem condition

3
3-dimensional structure of model
Daily oceanographic fluxes (water, heat,
salt) into and out of each box are controlled by
a ROMS oceanographic model
Biogeochemistry
Hydrographic submodel
Climate and oceanography
4
Atlantis Models

• Fully Developed
• SE Australia
• Port Phillip Bay, Aus.
• Mostly Completed
• Westernport, Australia
• Northeast US
• California Current
• Early Stages
• Torres Strait
• Coastal NSW
• Clarence River NSW
• SW Australia
• Ningaloo Marine Park
• SE Tasmania
• Central California

5
S.E. Australia
Cal. Current
Beth Tasmania
N.E. U.S.
Santa Cruz
Long Island
6
  What are the key features/equations/functions/
assumptions of the model?
7
Primary producers
Detritus
Light
Nutrients
Disease, Lysis, Stress, Fouling
Primary producer
Space
Grazer B
Grazer A
Grazer C
8
Invertebrate consumers
Detritus
Nutrients
Food availability
Disease, Oxygen limitation
Prey A
Invertebrate
Prey B
Prey C
Predator B
Predator A
Predator C
9
Vertebrate consumers (age-structured)
Nutrients
Detritus
Prey availability
Disease, Oxygen limitation
Gape limitation
Vertebratei Reserve Structure
Predator A
Prey A
Predator B
Prey B
Predator C
Prey C
Reproduction
10
Vertebrate predation and reproduction
Holling type II
Beverton Holt
Prey Consumption per Predator
Recruits
Prey abundance
Adults
OR Ricker, condition-dependent Beverton Holt,
constant, lognormal, dependent on prim producers
(Chla), dependent on all plankton, Bev-Holt with
lognormal variation added, Bev-Holt with
prescribed recovery, linearly dependent on
maternal condition, pupping/calving a fixed
number per adult spawning, or forced timeseries
of recruitment
OR Holling type I, Holling type III, Ecosim,
Size specific Holling type III, min-max
11
Spatially Explicit Output
Washington
Oregon
Small flatfish (Pleuronectids Dover sole, rex
sole, sanddab) (mg N/m3)
Cape Mendocino
Santa Cruz
Santa Barbara
Year 1
Year 40
12
Spatially Aggregated Output
13
What are the data requirements?
14
Atlantis biological parameters

• For each species
• Abundance per area
• Individual growth rates, length-weight
  conversions
• Max age, and age-at-maturity
• General habitat preferences
• Dispersal and/or migratory characteristics,
  within and outside model
• Diets
• Recruitment parameters (e.g. Beverton Holt,
  Ricker)

• For ecosystem map
• Habitat distribution (bottom type biogenic
  habitat)

• For calibration
• Historical catch and abundance data

15

• What key data gaps have been identified? Are
  these data gaps informing monitoring efforts?
• Biomass estimates nearshore and pelagic
  species, lower trophic levels (jellyfish,
  benthos, etc.)
• Diets
• Bycatch rates
• Fleet costs (US) and social factors governing
  fleet behavior

16
What is the model being used for?Is it used for
management?
17
Management Strategy Evaluation
18

• S.E. Australia-Fleet Restructuring
• Ecosystem-scale MSE
• Pros and cons of 5 kinds of management strategies
• Results balanced use of levers is most
  effective
• Led to refinement of questions (incentive
  systems, size of spatial management zones etc)
• S.E. Australia- Indicators
• MSE tested hundreds proposed indicators (with
  fisheries dependent and independent "data")
• Resultsneed a suite, cover range of species with
  different characteristics
• Will shape reporting and decision rules (in
  tiered harvest strategies)

Central California Considering interactions
between state and Sanctuary management
19
What are the strengths and weaknesses of the
model?
20
Atlantis

• Pros
• Flexible options for predation, reproduction,
  growth, gape limitation
• MSE (monitoring, assessments, indicators,
  economics, management)
• Nutrient handling, and interfaces with ROMS
  oceanography output
• Migrations out of region

• Cons
• Build time (6 months)
• Run time (hours- days)
• Lacks balancing routines
• Cumbersome parameterization

21
Model development and improvement

• Output graphics code for total biomass time
  series, biomass per region, and weight-at-age
• Diet handling
• Improved user interface and parameter input

Weight-at-age
22
Funding

• NOAA NMFS
• NOAA Fisheries and the Environment (FATE)
• NOAA NMFS Economics Program
• Moore Foundation
• Packard Foundation
• Contacts
• Beth Fulton Beth.Fulton_at_csiro.au
• Isaac Kaplan Isaac.Kaplan_at_noaa.gov
• Jason Link Jason.Link_at_noaa.gov

23
(No Transcript)
24
Weight-at-age
Weight-at-age relative to expectation from von
Bertalanffy
Age class 1
Age class 10
25
Atlantis publications are peer reviewed

• Journal Articles
• Fulton, E. A. 2004. Biogeochemical marine
  ecosystem models II the effect of physiological
  detail on model performance. Ecological Modelling
  173371-406.
• Fulton, E. A., and A. D. M. Smith. 2004. Lessons
  learnt from a comparison of three ecosystem
  models for Port Phillip Bay, Australia. South
  African Journal of Marine Science 26219-243.
• Fulton, E. A., A. D. M. Smith, and C. R. Johnson.
  2003a. Effect of complexity on marine ecosystem
  models. Marine Ecology Progress Series 2531-16.
• Fulton, E. A., A. D. M. Smith, and C. R. Johnson.
  2003b. Mortality and predation in ecosystem
  models is it important how these are expressed?
  Ecological Modelling 169157-178.
• Fulton, E. A., A. D. M. Smith, and C. R. Johnson.
  2004. Effects of spatial resolution on the
  performance and interpretation of marine
  ecosystem models. Ecological Modelling 17627-42.
• Fulton, E. A., Smith, A. D. M., and Punt, A. E.
  2005. Which ecological indicators can robustly
  detect effects of fishing? ICES Journal of
  Marine Science, 62 540-551.
• Technical Documents
• Brand, E. J., I. C. Kaplan, C. J. Harvey, E. A.
  Fulton, A. J. Hermann, J. C. Field and P.S.
  Levin. In press.  A Spatially Explicit Ecosystem
  Model of the California Currents Food Web and
  Oceanography. NOAA Tech Memo. NMFS-NWFSC.
• Fulton, E.A., Fuller, M., Smith, A.D.M. and Punt,
  A., 2005. Ecological Indicators of the Ecosystem
  Effects of Fishing Final Report. Australian
  Fisheries Management Authority Report, R99/1546.
• Fulton, E.A., Slater J., Smith, A.D.M. and Webb,
  H., 2005. Ecological Indicators for the Impacts
  of Fishing on Non-Target Species, Communities and
  Ecosystems Review of Potential Indicators.
  Australian Fisheries Management Authority Report,
  R99/1546-A.
• Fulton, E. A., Smith, A. D. M., and Punt, A. E.
  2003. Indicators of the Ecosystem Effects of
  Fishing Case-Study 1 Temperate Bay Ecosystem.
  Milestone Project Report, CSIRO Marine Research,
  Hobart.

The code base is not (but is the product of 10
years of work by many at CSIRO)
26
Fisheries age-specific catch
Carrion
Effort
Catch
Fleet A
Availability
Fleet B
Fish, age i
Catchability
Fleet C
Selectivity
27

• Key elements for each presentation      What
  is/has/will the model be used for?    What are
  the data requirements?    What key data gaps
  have been identified?    Are these data gaps
  informing monitoring efforts?    What are the
  key features/equations/functions/assumptions of
  the model?    What are the strengths of this
  model?    What are the weaknesses of this model?
     Has the model been published in the peer
  reviewed literature?    Has the model software
  been through a formal peer review?    Have the
  model outputs been through a formal peer review?
     How portable is the model software package?
     What remains for model development/improvement
  /enhancement?    Has/is/will the model outputs
  be used in LMR management?

28
Improved Diet Handling(Jason Link and Robert
Gamble)
29
Atlantis Model of the California Current
30

• Model food web has 56 functional groups,
  including
• 3 primary producer groups
• 2 bacteria groups
• 3 infaunal invertebrate groups
• 9 epifaunal invertebrate groups
• 5 pelagic invertebrate groups
• 21 fish groups
• 3 seabird groups
• 6 marine mammal groups
• 2 detritus pools

31
California Current Ecosim model Field (2004,
2006)
32
Fisheries
To date PacFIN catch per region,
1981-2004 Future work Fleet dynamics calibrated
to historical catches, and limited by spatial or
seasonal closures, quotas or effort limits
Groundfish trawling effort off central Oregon
Coast, 2003
33
Model Dynamics
34
Outputs with No Fishing
Chloro- phyll a (mg N/m3)
Year 40
Year 1
35
Outputs with No Fishing
Deep corals and anemones (mg N/m3)
Year 1
Year 40
36
Outputs with No Fishing
Small flatfish (Pleuronectids Dover sole, rex
sole, sanddab) (mg N/m3)
Year 1
Year 40
37
Scoping
Identify goals of EBM and
threats to achieving goals
Develop ecosystem indicators
and targets
Risk Analysis
Characterization of
Characterization of
susceptibility to
resiliency to
perturbation
perturbation
Data Acquisition
Monitoring of
Ecosystem Indicators
Assessment of ecosystem status
relative to EBM goals
Management Actions
38
Testing Ecosystem Indicators

• Changes in ecosystem pressure (e.g., temperature)
  or ecosystem attribute of interest (trophic
  structure)

Indicator
Indicator
Attribute
Attribute
39
Project 1 Indicators of Fishing Impacts

• A first cut
• For fished species, remove a fixed amount of
  biomass annually from standing stock
• We show removals as a of baseline biomass
• After 25 years, examine changes in ecosystem
  structure
• What indicators reveal this change?

40
(No Transcript)
41
(No Transcript)
42
Project 2 English Sole and Canary Rockfish
Current Management
Observations of Stock
Stock Synthesis II Assessments and Parameter
Estimation
Real World Ecology
2 yr cycle
Implementation error
Decision Rule
43
Management Strategy Evaluation
Climate Indicators in Assessments and Management
Decisions (w/ Ian Stewart, NWFSC)
Observations of Stock
Stock Synthesis II Assessments and Parameter
Estimation
Ecological Indicators
Atlantis Ecology and Fishing Simulator
2 yr cycle
Implementation error
Decision Rule
44
Atlantis

• Management Strategy Evaluation Test monitoring,
  indicators, assessments and regulations
• Integrating ecosystem information
• Strategic planning for fisheries management

45
Contacts

• Phil Levin phil.levin_at_noaa.gov
• [phone]
• Isaac Kaplan isaac.kaplan_at_noaa.gov
• [phone]