You are on page 1of 83

Modelling Complexity:

space, diversity (functional, species, size)


Fabien Lombard 5UM35 MEM
Outline

• Introduction : why to model complexity?

1) Plankton functional types (PFT)


2) Trait-based modelling
3) Adaptative dynamic approach
4) End-to-End modelling
5) Special case of the benthos: Integrating implicit
competition (tomorow)
Fasham model: a « simple » model

E =f(t) Production nouvelle


Absorption
Excrétion
Nd:DON P Nn:NO3

MLD
Absorption

Nr:NH4
Absorption

Absorption

Mortalité, égestion
Reminéralisation
Broutage Broutage

Broutage
Bacteria:B Z Detritus: D

Kz, M, sédimentation

Excrétion

Reminéralisation B attaché

NO30
Fasham model: a « simple » model

System of differential equations(Fasham et al., 1990)


Fasham model: main results
In agreement with data
Total NPP
1: La démarche
Why changing de modélisation
the modeling2: framework
Réponse du plancton à l’environnement
? Too simple ? 3: Les modèles biogéochimiques

Conclusion on NPZD Fasham et al. 1990

 most of the export takes place during detrainement (37-65%)


 main loss for phytoplankton is natural mortality (70%) and
the grazing is weak

-> the physic is more important than the biology in generating


the flux?
-> do we need a foodweb?
-> is it realistic?
1: La démarche
NPZD MODELS de modélisation
AT global2:scale
Réponse du plancton à l’environnement 3: Les modèles biogéochimiques

More recent models

SeaWiFS

NPZD

Chla
(mgChl/m3)
Nature is more abundant and extreme than models
Models usually fails in the antarctic zone (too high) and are too low
for the rest [Prentice et al., 2004]
NPZD MODELS AT global scale

More recent models

SeaWiFS

NPZD

Interanual variation of Chla by sattelite and seen by different NPZD


models

Nature is more stable than models


=> Stabilizing retro-actions?
[Prentice et al., 2004]
Why changing the modeling framework ? Too simple ?

Buffering capacity of
oceans?

Not enough complexity


to generate fluctuation
through feed backs and
non linear interactions

Prentice et al. [2001]


Why changing the modeling framework ? Too simple ?

That’s our view of the ecosystem… in the 80’s


Why changing the modeling framework ? Too simple ?
Why changing the modeling framework ? Too simple ?

« Plankton »
Why changing the modeling framework ? Too simple ?

« Plankton »

If god had consulted me before embarking on the creation, I would have suggested
something simpler.
Alfonso of Castile (15th century)
Why changing the modeling framework ? Too simple ?
New attemps in modelling pelagic ecosystem with more complexity

Central question :How to represent the marine biodiversity


in a model?

1) Plankton functional types (PFT)


2) Trait-based modelling
3) Adaptative dynamic approach
4) End-to-End modelling
A new attemp in pelagic ecosystem modelling with PFT and complex processes
A new attemp in pelagic ecosystem modelling with PFT and complex processes

PISCES Model (Aumont & Bopp 2006)


A new attemp in pelagic ecosystem modelling with PFT and complex processes
A new attemp in pelagic ecosystem modelling with PFT and complex processes

1. Have an important role in


biogeochemical cycles
2. Can be observed
3. Parameters are known
4. Are present in most oceans
A new attemp in pelagic ecosystem modelling with PFT and complex processes

An example with Planktom10

DIN: NP3 +NH4

Bact. Het. Macrozoopl.

PlankTOM10 model, a global marine biogeochemical model based on the representation of ten Plankton
Functional Types (PFTs), including six phytoplankton (pPFTs), three zooplankton (zPFTs) and bacteria.
PlankTOM10 also represents the full cycles of C, O2, P and Si and simplified cycles for Fe and N. It currently
comprises of 39 biogeochemical tracers

http://lgmacweb.env.uea.ac.uk/green_ocean/model/description.shtml
http://lgmacweb.env.uea.ac.uk/green_ocean/model/code_description/Tex/bgcbio/bgcbio.html
A new attemp in pelagic ecosystem modelling with PFT and complex processes

(incl. cocco’)

(e.g. diatoms)

Too high

Too low
A new attemp in pelagic ecosystem modelling with PFT and complex processes
A new attemp in pelagic ecosystem modelling with PFT and complex processes
N

“Just” adding jellyfishes helps to solve


quite a lot of issues
A new attemp in pelagic ecosystem modelling with PFT and complex processes
A new attemp in pelagic ecosystem modelling with PFT and complex processes
Several PFT models do exists
A new attemp in pelagic ecosystem modelling with PFT and complex processes
Are they better ? Atlantic
Skill assessment with the Taylor diagram Antarctic
Pacific

Monthly observation Annual observation Global

Di
ff er
en
ce
wi
th
ob
se
rv
ati
on
s

Difference in mass between model/observation

Taylor, K.E.: Summarizing multiple aspects of model performance in a single diagram. J. Geophys. Res., 106, 7183-7192, 2001 (also see PCMDI Report 55, http://www- pcmdi.llnl.gov/publications/ab55.html)
A new attemp in pelagic ecosystem modelling with PFT and complex processes
Nutriment, O2 Atlantic
Fe=0.7
Antarctic
Pacific
Global
NO3=0.96
O2 = 0.97

Phytoplankton

Pac=Atl=0.8

Zooplankton
Pac= 0.6
Atl=0.7

The higher is the trophic layer


The worst is the relationship between model
results and observation
A new attemp in pelagic ecosystem modelling with PFT and complex processes

Inter comparaison between models

Bopp et al., 2013, Multiple stressors of ocean ecosystems in the 21st century: projections with CMIP5 models
Biogeosciences, 10, 6225–6245, 2013
A new attemp in pelagic ecosystem modelling with PFT and complex processes
Inter comparaison
between models

This is a plankton mess!


But no one knows what
will be the truth

Bopp et al., 2013, Multiple stressors of ocean ecosystems in the 21st century: projections with CMIP5 models
Biogeosciences, 10, 6225–6245, 2013
A new attemp in pelagic ecosystem modelling with PFT and complex processes
This is a plankton mess!

Simulated patterns that agree well with observations have relatively high correlation and low RMS errors
and be close to std=1.
A new attemp in pelagic ecosystem modelling with PFT and complex processes

GONÇALVES LELES et al 2018


Traditionally Modern view

Flynn et al 2013
GONÇALVES LELES et al 2018
A1:new
La démarche
attemp de modélisation 2: Réponse du
in pelagic ecosystem plancton à l’environnement
modelling by resolving 3: Les modèles biogéochimiques
functional types with
trait based modelling
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

The major ecological axes that define ecological niches of plankton are physical
environment, resources, and natural enemies (grazers and parasites).
For each of these axes a whole hierarchy of traits exists, from the subcellular to
population level, that allow plankton to survive and reproduce in the environment.
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

Kiorboe et al 2018
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

Kiorboe et al 2018
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

Phytoplankton

Litchman and Klausmeier


2008
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

Zooplankton

Litchman et al 2013
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

All plankton + benthos + necton

Martini et al 2021
Large interrest
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

Most traits in organisms are not independent but are instead interrelated. Correlations between
and among traits often represent trade-offs and complicate assessment of the selective pressures.

Correlations among traits can also lead to a reduced number of potential strategies characterized
by different trait values, where certain combinations of trait values become impossible. A trade-
off arises when a trait that is advantageous for one function confers a disadvantage for others.
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

Traits trade off for zooplankton

Barton et al 2014
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

Traits trade off for zooplankton

Conley et al 2018
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling

Size is a master trait


“Size has a remarkably great influence on the
organisation of animal communities”
Elton, Animal Ecology, 1926

“For every type of animal there is a most convenient


size, and a large change in size inevitably carries with
it a change of form”

Haldane, On Being the Right Size, 1928


Andersen et al 2016
Andersen et al 2016
Andersen et al 2016
Testing different functional responses

And refuge from predation


Testing different functional responses

And refuge from predation


A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling
Adaptive dynamic approaches
A new attemp in pelagic ecosystem modelling with trait based and adaptative
dynamic modelling
A new attemp in pelagic ecosystem modelling with trait based and adaptative
dynamic modelling

Certainly an effect of the chlorophyll / biomass ratio


A new attemp in pelagic ecosystem modelling with trait based and adaptative
dynamic modelling
A new attemp in pelagic ecosystem modelling with adaptative dynamic modelling
A new attemp in pelagic ecosystem modelling with adaptative dynamic modelling
A new attemp in pelagic ecosystem modelling with adaptative dynamic modelling
A new attemp in pelagic ecosystem modelling by resolving functional types with
trait based modelling
A new attemp in pelagic ecosystem modelling with adaptative dynamic modelling
A new attemp in pelagic ecosystem modelling with adaptative dynamic modelling
End to end models
A new attemp in pelagic ecosystem modelling with adaptative dynamic modelling
A new attemp in pelagic ecosystem modelling with End to End models
A new attemp in pelagic ecosystem modelling with End to End models
A new attemp in pelagic ecosystem modelling with End to End models
A new attemp in pelagic ecosystem modelling with End to End models

Types of models

1) Individual
Based Model (for example OSMOSE)
or DEB models structured in size (for
example APESCOM)

2) PFT models for plankton (PISCES,


PLANKTOM, …)

3) Physical model (for examples


ROMS, OPA, …) for transport and
physical conditions

Cury et al., 2008


A new attemp in pelagic ecosystem modelling with End to End models

using size structure for fishes


A new attemp in pelagic ecosystem modelling with End to End models
A new attemp in pelagic ecosystem modelling with End to End models
A new attemp in pelagic ecosystem modelling with End to End models

Everett et al under press


A new attemp in pelagic ecosystem modelling with End to End models
TD
Objectif

1) Integrate a 2D model
1) Code for the 2d difusion of a single tracer using a A-grid
-Remember that 2D is actually 1D in one direction + 1D in the
second direction
-Remember that there is two limit conditions in the first
dimension … and the same in the second dimension

2) Add an NPZD model within this 2D framework

3) Add several nutrients / phytoplankton / zooplankton as you wish


(take a look at the hidden slides in there, they contains parameters)

You might also like