Transcript Integrated Plant Phenotype

Use of ecophysiological approaches and
biophysic plant modelling
in determination of complex phenotypic traits
and analysis of interactions GxE
Pr. Jérémie LECOEUR
Professor of Plant Biology
Director of Plant Science Department
Montpellier SupAgro
1. Context
Context : a need to understand the building of the plant phenotype
The plant phenotype is always a complex object resulting from the spatial
and temporal integration of various biological processes  « Integrated
Plant Phenotype »
Integrated plant phenotype: Plant traits resulting of the integration of the
major plant functions in response to environment.
An example of an « Integrated Plant Phenotype »:
The architecture of the At rosette
This integrated phenotype results from:
• organogenesis
• morphogenesis
• carbon metabolism…
in interaction with the environment
Col
se
rot
Picture
Corresponding
Virtual plant
Context : a need to understand the building of the plant phenotype
The plant phenotype is always a complex object resulting from the spatial
and temporal integration of various biological processes  « Integrated
Plant Phenotype »
Genotype x
x
Environment
=
 Responses
=
=
Response
Integrated plant phenotype: Plant traits resulting of the integration of the
major plant functions in response to environment.
=
genotype 2
genotype 1
Environment
Phenotype
Context : a need to understand the building of the plant phenotype
Choice of the plant representation
fruits
Ecophysiological modelling
Organ populations in relation with environment
through correlative relationships
Leaves
=
roots
Genetic modelling
Mainly statistical approaches
phenotype = G + E + GxE + e

Response
Process based models (crop models)
géno 2
géno 1
Environment
« Virtual plants »
Set
of
phytomeres
with
connections with matter flows
topological
Context : a need to understand the building of the plant phenotype
The plant is a complex system = a large number of sub-units with the
same organisation and topological connection resulting in a network
Purslane plant
Cell protein tree
(d’après Jeong, 2003)
The same level of complexity could be find at organelle, cell, tissue…
Context : a need to understand the building of the plant phenotype
Theory of the increase in scientific progress through combinatories
of conceptual and technic artefacts (Lebeau, 2005)
A postulate ?
«The only way to make significant progress in understanding the
genotype - environment interaction is to associate several scientific
disciplines»
The needed scientific disciplines would be:
- genetic and genomic,
- plant biology and plant physiology,
- ecophysiology and biophysic
- applied mathematics,
2. Advances in Ecophysiology
Step 0 : Characterization of the physical
environment at plant boundaries
Advances in Ecophysiology
The absolute necessary to take into account the physical environment
Systematic characterization of plant microclimate
In field
The minimum data set includes
temperature, radiation and atmospheric
humidity, wind speed and rainfall
In growth chamber
To allow the comparison between experiments and the establishment of
trial network typologies or a future use of models
Advances in Ecophysiology
To be as close as possible to the microclimate sensed by the plant or
by its organs
First use of modelling: to estimate the environmental variables
instead of measuring them.
To model the energy, radiative and water balances….
(from Rey, 2003; Lhomme and Guilioni, 2004 and 2006; Chenu et al., 2005 and 2007; Louarn et al., 2007)
Advances in Ecophysiology
To be as close as possible to the microclimate sensed by the plant or
by its organs
To identify the environmental variables quantitatively related to
plant development and growth.
For instance, what is the
organogenesis on At?
variable well related
to
the
0.10
-1
Vitesse d'initiation (°Cj )
0.10
Incident PAR
0.05
0.00
0
2
4
6
8
10
12
-2 -1d-1)
Incident PAR (mol m-2
PAR incident (mol m j )
Light quality
(R/FR - Blue)
0.05
Densité de flux de photons
-2 -1
-1
(µmol m s nm )
Vitesse d'initiation (°Cj-1)
radiative
Absorbed
PAR
2.5
2.0
10.8 mol m-2 j-1
5.2 mol m-2 j-1
2.5 mol m-2 j-1
1.5
1.0
0.5
0.0
400
600
800
1000
Longueur d'onde (nm)
0.00
0 0.22 0.44 0.6 6 0.8 8 1.0 101.2 12
0.0
1.4
-2 -1-1d-1)
Absorbed
PAR (mmol
plt-1
PAR incident
(mol m
j )-1
PAR absorbé (mmol plte j )
0.01
0.1
1
10
-1 -1scale)
Absorbed
PAR
PAR absorbé
(mmol
plte(log
j ) [échelle log]
(from Chenu et al., 2005)
Advances in Ecophysiology
To be as close as possible to the microclimate sensed by the plant or
by its organs
A lot can be done by using standard bioclimatological indicators…
Thermal time,
Cumulative solar radiation,
Photothermal coefficient,
Climatic water balance…
Step 1 : Ecophysiologic diagnosis of the
phenotypic variability
To dissect the genotype – environment interaction
Advances in Ecophysiology
Second use of modelling: formalization of plant – environment
interaction to identify unknown phenotypes
Analysis of a panel of wild types and their mutants in At
Wild type
mutants
Col
Ws
Ler
se
3.5
ron
rot
(from Chenu et al, 2007)
Dij
Advances in Ecophysiology
Second use of modelling: formalization of plant – environment
interaction to identify unknown phenotypes
0.15
0.15
Col
Ws
Ler
Dij
0.10
0.10
0.05
0.05
0.00
se
3.5
All wild type
0.01
ron
0.1
1
10
0.00
0.001 0.01
0.1
1
10
0.10
0.05
0.00
0.01
0.1
1
10
0.01
0.1
1
10
0.10
0.05
0.00
0.001 0.01
0.15
0.1
1
10
All genotypes
Comparison wild types vs corresponding mutants
Col / se / rot
Ws / 3.5
0.15
Ler / ron
Génotypes
0.10
0.10
0.05
0.05
0.00
0.001 0.01
0.1
1
10
0.01
0.1
1
10
0.01
0.1
1
10
-1 -1
Absorbed PAR (mmol plte j ) [log scale]
(from Chenu et al, 2007)
0.00
0.001 0.01
0.1
1
10
0.8
0.6
0.4
0.2
0.0
0
10
20
30
40
1.0
0.8
0.6
0.4
b
0.2
0.0
0.0
0.2
Température des feuilles (°c)
For instance, leaf expansion…
Vitesse relative d'expansion
Tournesol
0.4
0.6
0.8
1.0
FTSW
1.0
0.8
0.6
0.4
c
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1.0
0.8
0.6
0.4
0.2
d
0.0
0
Vitesse relative d'expansion
Vigne
10
20
30
40
PARa (m-2 mol j-1)
FTSW
1.0
0.8
0.6
0.4
e
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1.4
1.2
1.0
0.8
0.6
0.4
f
0.2
0.0
0.0
0.2
0.4
0.6
0.8
RER (mm2 mm-2 °Cj-1)
Laitue
Vitesse relative d'expansion
FTSW
0.06
0.04
0.02
g
0.00
1.0
0.1
FTSW
0.2
0.3
0.4
RER (mm2 mm-2 °Cj-1)
PARa (mol j-1)
Arabidopsis
thaliana
0.050
0.045
0.040
0.035
h
0.030
0.0000.0010.0020.0030.0040.0050.006
PARa (mol j-1)
Haricot
Vitesse relative d'expansion
Response curve families
Rayonnement
absorbé
Vitesse relative d'expansion
a
1.0
Vitesse relative d'expansion
Pois
Vitesse relative d'expansion
Establishment of consistent
relatioship betwen plant and
environment variables
Déficit hydrique
édaphique
Température
1.0
0.8
0.6
0.4
i
0.2
0.0
0.0
0.2
0.4
0.6
FTSW
0.8
1.0
Advances in Ecophysiology
Second use of modelling: formalization of plant – environment
interaction to identify unknown phenotypes
Columbia
Serrate
Vini = aini log(PARa) + bini
This approach allowed to identify a new
involvement of the Serrate gene in plant
organogenesis.
(from Chenu et al., 2007)
G
GxE
G
Advances in Ecophysiology
Time consuming ecophysiological measurements require
« industrial phenotyping » or a large field trail network
It will be necessary to increase by 10 to 100 the number of
characterized experimental situations
(From Joined Unit LEPSE – INRA / SupAgro, 2006 report)
Step 2 : To quantify the impact of the
observed phenotypic differences
Advances in Ecophysiology
Third use of modelling: to analyse the consequences of multi-trait
differences on integrated plant phenotypes
The sensitivity analyses allow to rank the traits in term of their
quantitative effects on the integrated phenotype.
An example: phenotypic variability in light interception in sunflower during
seed development.
Among a panel of 20 genotypes, the following phenotypic differences were
observed:
- plant leaf area,
- individual leaf area,
- leaf number,
- leaf size distribution along the stem,
- blade angle,
- duration of leaf life.
Advances in Ecophysiology
Third use of modelling: to analyse the consequences of multi-trait
differences on integrated plant phenotypes
Virtual sensitivity analysis of
light interception to various
phenotypic traits
Changes in
position of the
largest leaf on
the stem
Changes in
plant leaf
area
Average virtual
plant
(from Casadebaig, 2004)
Changes in
leaf number
Virtual plot at flowering (6.6 plants m-2
cv Heliasol)
Estimation of light
interception
1.2
Y = 0.98 X + 0.01
R2 = 0.993
0.8
0.6
0.4
0.2
0
Fraction of radiation intercepted
Estimated
1
0
Sunflower virtual plant
cv Heliasol
1.0
0.8
0.6
 ei
0.4
0.2
0.0
0.2
0.4
0.6
0.8
Measured
1
Days
Ei
1.2
(from Rey, 2003; Casadebaig, 2004)
Advances in Ecophysiology
Third use of modelling: to analyse the consequences of multi-trait
differences on integrated plant phenotypes
Sensitivity analysis
(from Casadebaig, 2004)
200
Leaf number
Position of the largest leaf on the stem
Plant heigth
Duration of leaf life
100
150
Blade angle
50
(in % of average plant)
Changes in light interception
Plant leaf area
-400
-200
0
200
400
Evaluated ranges of variation in observed traits
(in % of the average value)
A hidden trait affecting the light interception was identified: the distribution
of leaf sizes along the stem
Relative leaf irridi
Advances in Ecophysiology
85
Emerging properties at plant Exp.
level
in At?
1
80
Exp. 2
Exp. 3
Y = -0.78 X + 92; r² = 0.618
The changes in organogenesis, organ expansion
and morphology lead to
unexpected property: the life irradiance is 75
improved
in 4response
to
reductions
in
0
2
6
8
10
12
14
16
incident light
Incident PAR (mol m-2 d-1)
Relative leaf irridiance (%)
90
85
Exp. 1
Exp. 2
Exp. 3
80
Y = -0.78 X + 92; r² = 0.618
75
(relatively to the standard treatment)
Effect on light interception efficiency
10
95
8
6
4
2
0
-2
Overall
effects
Petiole
length
Leaf
shape
Phenology
delay
-4
0
2
4
6
8
10
12
14
16
Nature of effects analyzed
Incident PAR (mol m-2 d-1)
(b)
standard treatment)
erception efficiency
10
8
6
4
(adapted from Chenu et al., 2005)
Change with time in trophic competition inside the grapevine shoot
• 3 phases
0C
6C
A
0.25
B
F
Q/D ratio (arbitrary units)
0.20
0.15
1
2
3a
1
V
2
3b
'GRENACHE N'
0.10
2- Increase in trophic
competition due to rapid
production of new sinks
0.05
0.00
C
0.25
D
F
0.20
0.15
1
2
3a
1- decrease in trophic
competition due to the
increase in sources
1
V
2
3b
'SYRAH'
0.10
3-(0C)- Decreasein
trophic competition due
to the end of secondary
axes development
0.05
0.00
0
200
400
600
800 1000 1200
0
200
400
600
Thermal time from budburst (°Cd)
800 1000 1200
3-(6C)- Increase trophic
competition due the
second growth phasis of
clusters
Relationship between axis development and trophic competition
Probability to maintain the
development
Relationship between Q/D values and the probability of end of secondary axes development
Primary axes
P0 secondary axes
P1- P2 secondary axe
1.0
0.8
0.6
0.4
0.2
A
Sigmoidial adjustment
Syr 0C
Syr 6C
Gre 0C
Gre 6C
B
C
0.0
0.00 0.05 0.10 0.15 0.20 0.250.00 0.05 0.10 0.15 0.20 0.250.00 0.05 0.10 0.15 0.20 0.25
Q/D ratio (arbitrary units)
• Primary axes are not affected by the trophic competition
1.0
• Secondary axis are affected by the trophic competition
• A single sigmoidal relationship P=f(Q/D).
• A difference in sensitivity according to the type of axes
0.8
0.6
0.4
Primary axis
P0 secondary axis
P1-P2 secondary axis
0.2
0.0
0.00
0.05
0.10
0.15
0.20
0.25
Relationship between axis development and trophic competition
Relationship between Q/D values and the probability of end of secondary axes
development according to their type and size
1.0 A
0.8
1-5
0.6
1-5 leaves
0.4
(0.31g)
0.2
0.0 B
1.0
0.8
6-10
6-10 leaves
(2.87g)
0.6
0.4
0.2
0.0 C
1.0
0.8
0.6
0.4
0.2
11-16
11-16 leaves
(10.21g)
P0 secondary axes
P1-P2 secondary axes
P1-P2 sigmoid adjustment
P0 sigmoid adjustment
0.0
0.00 0.05 0.10 0.15 0.20 0.25
Q/D ratio (arbitrary units)
3. The front of « modelling experiences »
Step 3 : To model the impact of genotypic
variability on the plant phenotypic plasticity
To associate various kind of models to predict the integrated plant
phenotypes
The front of modelling experiences
To evaluate the genotype performances
The biophysical modelling approaches are now enough tried and tested to be
revisited to predict the genotype – environment interaction.
The available modelling approaches (not exhaustive):
- biophysical balances,
- crop models,
- ecophysiological descriptions of regulations and signals in plants,
- 3D architectural plant and canopy models,
- mathematical models to estimate parameters in complex systems…
The front of modelling experiences
To evaluate the genotype performances
Construction of dedicated models
Flow chart of potential yield estimation in sunflower
Input data
Phenology
Rendement en graines
MSgraine = MScapi x HI_grainegen
Fin
Temps thermique depuis la levée
t i
TTi (Tmoy Tbase)dt
levée
Données climatiques
Tmoy
Tmin
Tmax
PARi
oui
TT>TT_M3gen
non
Architecture (3D)
Light interception (3D)
Biomass production
Biomass partitioning
Phénologie
TT_E1gen
TT_M0grn
Biomasse du capitule
TT_F1gen
TT_M3gen
TTi<TT_E1
TTi<TT_M3
Nombre de feuilles à fin expansion
MScapii = 0
0,632
MScapii
1 TTiTT_E1gen
774


2.83
 MS i
MScapi >= MSi x HI_capigen MScapi = MSi x HI_capigen
NF<NFfinalgen NF = Phyllochronegen x TTj
NF=NFfinalgen
Biomasse aérienne totale
MSi= MSi-1+dMSi
Rang de la dernière feuille morte
TTi>TT_M0gen
NFmorte  NFfinal
gen
TT_M3 genTTi

 

 TT_M3 genTT_M0 gen 
Production journalière de biomasse
dMSi = eb x ei x PARi
Surface foliaire de la plante produite
jNF
aSFgen
dj

bSFgen  j 
j0
1exp(4c SFgen 
)

 aSFgen 
Efficience biologique
ebi = ebpoti x FTi
SFplante_produite 
Surface foliaire sénescente
k NFmorte
aSFgen
SFplante_sénescence 
dk


k 0
1exp(4cSFgen bSFgen k )
 aSFgen 
Facteur thermique
FTi = 1 – 0,0025 (0,25 Tmin + 0,75 Tmax –25)²
Efficience biologique potentielle
TTi<TT_E1gen,
TTi<TT_F1gen,
Surface foliaire de la plante
SFplante = SFplante_produite – SFplante_sénescente
(adapted from Lecoeur et al., 2008)
Indice foliaire
LAI = dens x SFplante
TTi<TT_M0gen,
Efficience d’interception
ei = 1 – exp ( - kgen x LAIi)
TTi<TT_M3gen,
ebpoti = 11 TT-F1 -TT_E1  (e -1)
ebpoti =
ebpoti = ebgen
TT TT_M0
ebpoti = ebgen x 0,38  exp ( 2 ( 1 -  TT_M3
-TT_M0
TT_F1gen - TTi
gen
bgen
gen
i
gen
gen
gen



The front of modelling experiences
To evaluate the genotype performances
Construction of dedicated models
Flow chart of potential yield estimation in sunflower
Input data
Phenology
Rendement en graines
MSgraine = MScapi x HI_grainegen
Fin
Temps thermique depuis la levée
t i
TTi (Tmoy Tbase)dt
levée
Données climatiques
Tmoy
Tmin
Tmax
PARi
oui
TT>TT_M3gen
non
Architecture (3D)
Light interception (3D)
Biomass production
Biomass partitioning
Phénologie
TT_E1gen
TT_M0grn
Biomasse du capitule
TT_F1gen
TT_M3gen
TTi<TT_E1
TTi<TT_M3
Nombre de feuilles à fin expansion
MScapii = 0
0,632
MScapii
1 TTiTT_E1gen
774


2.83
 MS i
MScapi >= MSi x HI_capigen MScapi = MSi x HI_capigen
NF<NFfinalgen NF = Phyllochronegen x TTj
NF=NFfinalgen
Biomasse aérienne totale
MSi= MSi-1+dMSi
Rang de la dernière feuille morte
TTi>TT_M0gen
NFmorte  NFfinal
gen
TT_M3 genTTi

 

 TT_M3 genTT_M0 gen 
Production journalière de biomasse
dMSi = eb x ei x PARi
Surface foliaire de la plante produite
jNF
aSFgen
dj

bSFgen  j 
j0
1exp(4c SFgen 
)

 aSFgen 
Efficience biologique
ebi = ebpoti x FTi
SFplante_produite 
Surface foliaire sénescente
k NFmorte
aSFgen
SFplante_sénescence 
dk


k 0
1exp(4cSFgen bSFgen k )
 aSFgen 
Facteur thermique
FTi = 1 – 0,0025 (0,25 Tmin + 0,75 Tmax –25)²
Efficience biologique potentielle
TTi<TT_E1gen,
TTi<TT_F1gen,
Surface foliaire de la plante
SFplante = SFplante_produite – SFplante_sénescente
(adapted from Lecoeur et al., 2008)
Indice foliaire
LAI = dens x SFplante
TTi<TT_M0gen,
Efficience d’interception
ei = 1 – exp ( - kgen x LAIi)
TTi<TT_M3gen,
ebpoti = 11 TT-F1 -TT_E1  (e -1)
ebpoti =
ebpoti = ebgen
TT TT_M0
ebpoti = ebgen x 0,38  exp ( 2 ( 1 -  TT_M3
-TT_M0
TT_F1gen - TTi
gen
bgen
gen
i
gen
gen
gen



The front of modelling experiences
To evaluate the genotype performances
Estimation of a productivity index from the genotypic traits
Observed productiviy index
180
y = 0,997x
r² = 0,934
RMSE = 16,4%
160
140
120
100
80
60
40
20
20
(from Lecoeur et al., 2008)
40
60
80
100
120
140
160
180
Estimated productivity index
A simple biophysic model allows to take into account from 80 to 90% of the
observed phenotypic variability in potential yield among a panel of 30 genotypes.
The front of modelling experiences
To evaluate the genotype performances
A sensitivity analysis allowed to quantify the impact on plant productivity of the
C1 C
genotypic traits
Total
Biomass partitioning
Photosynthesis
Yield coefficient of variation
Architecture
Phenology
Harvest index
Biomass allocation to capitulum
All the major functions
contributed
to
the
productivity variability.
Radiation use efficiency
Light interception coefficient
Area of the largest leaf
Position of the largest leaf
Plant leaf area
Leaf number
Phytomere production rate
M3 thermal time
M0 thermal time
F1 thermal time
Classical ANOVA detected
only the contribution of the
harvest index
(from Lecoeur et al., 2008)
E1 thermal thermal
0.000
0.050
0.100
0.150
0.200
Genotypic traits or trait sets
0.250
0.300
Reminder : first setting of the biomass partitioning model (Greenlab)
Objective : to understand the genotype variability of harvest index
Fitting on experimental data on 4 genotypes
Leaf
area
Sink strengths : petiole < leaf < stem < capitulum
0,45 < 1,00 < 1,07 < 3000
Leaf
biomass
biomass production and partitioning
along growth cycles
Actually,
we
are combining SunFlo (crop model) with
petiole sink
variation
GreenLab (FSPM) in order to analyse the genotypic
sink
variability of harvest index
0.020
0.015
0.010
0.005
0.000
internode
blade
15
petiol
10
Flow er
5
0
cycle of expansion
89
81
73
65
57
49
41
33
25
9
17
1
1
sink value
Leaf
strength
biomass production
Biom Tot
20
7
13
19 25
31 37 43
49 55
61 67 73 79 85
91 97 103 109 115 121 127 133
grow th cycles
(d’après Rey et al., 2006)
The front of modelling experiences
First attempt in combining genetics modules and crop model to test the
potentialities of a virtual breeding on index
Sunflo, a crop model including :
• A description of plant compartiments (vegetative parts, reproductive parts,
roots),
• A description of main processes (organogenesis, morphogenesis,
photosynthesis, biomass partitioning),
• Responses to temperature, solar radiation and water availability.
• Each genotype is described by a set of 15 to 20 traits
Quantitative Genetics Modules :
• Estimation of genetic correlation between phenotypic traits,
• Estimation of heritabilities,
• Choice of selection pressure on the traits according
environnement,
the
target
Applying several selection cycles resulting in population with new phenotypic
characterics. The performance of each new genotype is tested in various
environnement. This leads to estimate the potential genetic progress.
3. Potentialities and present limitations
Conclusions
Potentialities
The past 10-20 years plant modelling could be now an effective tool to analyse
and model the genotype – environment interaction:
• Estimations of microclimate variables
• Modelling plant responses to environment
• Ranking plant traits in term of quantitative impact on phenotypic
variability
• Predictions of integrated plant phenotypic
The links between concepts and methologies from various disciplines may
increase the progress in understanding integrated plant phenotypes.
Conclusions
Present limitations
• Low spreading of the biophysical modelling culture.
• Heavy cost of phenotypic information.
• Lack of applied mathematic adapted to complex systems.