Simulation of biomass production - Cirad

1

Simulation of biomass production

Introduction

Dr. Ep Heuvelink - Wageningen University

Simulation of biomass production– scope of the course -

Growth:

yieldleaf area

light interception

photosynthesis

biomass increase

partitioning

Uptake water & nutrients

Development: quality

seed

young vegetative plant

generative plant

harvestable product / seed

germinationemergence

flowerinduction

Influencing factors: light (amount/quality/day length), temperature (level/day & night

regime), CO2, humidity, EC, water, nutrients, plant density, pruning.Interactions between climate factors and crop in a greenhouse. Not considered: organic matter mineralization, specific nutrients, pests & diseases, plant breeding.

2

Simulation of biomass production- contents -

Crop models: what are they and why useful/important?

General scheme of a photosynthesis-driven crop growth model

Details of processes: - light interception- photosynthesis and respiration- dry matter productiondry matter partitioning

(other subject - concept of sink strength)

Simulation of biomass production- contents (cont.) -

Examples from

- wheat

- fruit vegetables (tomato, sweet pepper, cucumber)

- cut chrysanthemum

Educational methods

- Theory: syllabus/hand-outs/examples/exercises

- Training assignments: LINTUL, TOMSIM

3

Simulation of biomass productionAbout models

What is a model:

Simplified representation of a part of reality (= a system)

Mechanistic models enable us to represent and combine knowledge in a generic way

Why use models:

Research - testing hypotheses

Greenhouse climate control

Economic decision making

Education

Decision Support System (DSS)}

Simulation of biomass productionabout models -

4

Model use in climate and fertigation control

Descriptive: black box, regression, statistic, empirical direct relation between input and output

Explanatory: physiological, mechanistic quantitative description of mechanisms and processes contains sub-models at least one hierarchical level

deeper than the response to be described.

Simulation of biomass production- two types of models -

5

Simulation of biomass production- example of black box model -

Data from experiments in climate chamber:Length of young tomato plants after 40 days of growing.

Model based on these data:

Length = 36.3 +4.48 * (Td-21) + 1.37 * (Tn-21)

Disadvantages: No extrapolation No new conditions Advantages: High predictive value

Td (°C)

Tn(°C)

Plant length (cm)

26 16 5224 18 5022 20 3420 22 3518 24 2816 26 2318 18 1824 24 5424 12 39

Leaf area index

Light interception

Photosynthesis Respiration

Dry mass production (Growth)

LeavesRoot Stem Flowers

Simulation of biomass production- Principles of crop growth model (mechanistic)-

6

Advantages:

extrapolation is possible

separated in modules

give insight in plant reactions (‘emerging behavior’)

Disadvantages:

development is very time consuming

(team of researchers)

only known effects are in the model

Simulation of biomass production- Example of Mechanistic model-

7

Simulation of biomass production– Compare two models -

Model 1: descriptive model: growth rate (g m2 d1) = 1 + 0.2 * LI

Model 2: explanatory model: Photosynthesis GPHOT (g m2 d1) = LI * LAI Maintenance respiration MAINT = 0.03 * TDWTotal Dry Weight TDW = TDW + 0.7 * (GPHOT - MAINT)Leaf Dry Weight LDW = 0.2 * TDWLAI = LDW * 0.02

Question: If plants are grown under the situations listed below: at a light intensity (LI) of 20 W m2

with an initial dry weight of 100 g m2

with an initial Leaf Area Index (LAI) of 0.4 m2 m2

Please calculate crop dry weight after 2 days of growth using the 2 models given above respectively, and then compare the differences.

Simulation of biomass production– Compare two models (results)-

Model 1: descriptive model: growth rate (g m2 d1) = 1 + 0.2 * LI = 1+0.2*20 =5 (g m2 d1)

crop dry weight after 2 days = 100 + 2*5 = 110 g m2

Model 2: explanatory model: Firstly to calculate results for day 1 and use these results for the calculation on day 2.

Results DAY 1 DAY2GPHOT 8 8.2MAINT 3 3.1TDW 103.5 107.1LDW 20.7 21.4LAI 0.41 0.43

Compare models: 110 g m2 (Model 1) VS 107.1 g m 2 (Model2)

8

Simulation of biomass productionLight interception

The processes: step by step - Light interception

Leaf area index

Light interception




9

Influence of Leaf Area Index (LAI) on the fraction of light intercepted by a tomato crop (k = extinction coefficient = 0.8)

Simulation of biomass production– Light on top of canopy -

Wavelength :

Global radiation : 300-3000 nm

PAR: Photosynthetically Active Radiation: 400-700 nm

Energy:

In day light PAR outside 0.5 x Global radiation

In a glasshouse: PAR 0.7 x PAR outside

Example:

20 MJ day1 global radiation outside would result in about

20 x 0.5 x 0.7 = 7 MJ (PAR) day1 inside a glasshouse.

NB: Absorbed Intercepted

10

Light use efficiency (LUE, g MJ1) in winter: 3 plant densities

and - or + assimilation light for cut chrysanthemum

Simulation of biomass production– Dry matter production: a simple LUE model –

dW/dt = LUE (1 e k * LAI) I

dW/dt = growth rate [g(DM)m2 d1]

LUE = light use efficiency [g(DM) MJ1(PAR)]

K = extinction coefficient

LAI = Leaf area index

I = Photosynthetic Active Radiation (PAR) incident on crop

[MJ(PAR)m2 d1]

Assumes constant LUE !

11

Simulation of biomass productionPhotosynthesis

The processes: step by step - Photosynthesis

Leaf area index

Light interception




12

Simulation of biomass productionInfluence of Leaf Area Index (LAI) on crop photosynthesis

LAI amount of light intercepted amount of Photosynthesis

Simulation of biomass productionInfluence of CO2 concentration (ppm) on crop photosynthesis

CO2 Light Use Efficiency amount of Photosynthesis

13

Simulation of biomass productionInfluence of temperature (oC) on crop photosynthesis

Temperature Gross Photosynthesis (≠Crop growth/ yield)

Simulation of biomass production– Photosynthesis modules (1) –

Negative-exponential response curve

1

Non-rectangular hyperbola (not compulsory)

[ is between 0 and 1; determines the shape; ]

Rectangular hyperbola (not compulsory)

•

NB: Temperature and CO2 affect and Pgmax

14

Simulation of biomass productionPhotosynthesis - light response curves (values for )

Simulation of biomass productionLeaf photosynthesis: Pg = Pg,max (1-exp(- H/Pg,max))

Absorbed PAR

Le

af

ph

oto

syn

the

sis

ra

te

Shade leaf

Sun leaf

α

Pg,max

Pg,max

Sun Shade Effect on Pg,max

but not α Positions in the

canopy or type of plants

NB: H= Iabs

15

Simulation of biomass production– Resistance chain for CO2 transport and binding –

Total resistance = R(boundary layer)+ R(Stomata)+R(mesophyll)

Simulation of biomass production– Resistance chain for CO2 transport and binding –Inside leaves:

rm: Mesophyll resistance gm: Mesophyll conductance Г: CO2 compensation point

Internal CO2 concentration

Phot

osyn

thes

is

16

Simulation of biomass production– Photosynthesis modules (2) –

Biochemical model (Farquhar & Von Caemmerer type)Leaf photosynthesis is described as either Rubisco-limited (CO2) or electron transport rate limited(light) .

Canopy photosynthesis: big leaf approach exponential radiation profile (light extinction in crop)

and integration of leaf photosynthesis over the leaf layers

Influence of sink demandon photosynthesis ?

Cucumber leaves

Only reduced photosynthesis after 20 days when all fruits were removed

Marcelis (1991). J.Exp.Bot. 42: 1387-1392.

17

Simulation of biomass production– Effect of sink demand on photosynthesis? –

60 100 140 180

Day of year

0.0

0.4

0.8

1.2T

otal

dry

wei

ght (

g m

-2)

3 fruits per truss

7 fruits per truss

Influence of number of fruits per truss on total dry matter production in tomato. A reduction in fruit number (sink strength) by more than 50% did not influence total dry matter production. Since leaf area index was identical for both treatments, one can conclude that

effects of sink demand on leaf photosynthetic rate in commercial tomato production can be ignored.

Simulation of biomass productionRespiration

18

The processes: step by step - Respiration

Leaf area index

Light interception




Simulation of biomass productionDry matter production: explanatory crop growth model

dW/dt = Cf (Pgd - Rm)

dW/dt = Crop Growth Rate (g DM m2 d 1)

Pgd = Crop gross Assimilation Rate (g CH2O m2 d 1)

(In the model all leaves have identical photosynthetic properties)

R m = Maintenance Respiration Rate (g CH2O m2 d 1)

(Depends on Organ Dry Weight, Temperature, RGR)

Cf = Conversion Efficiency (g DM g1CH2O)

19

Simulation of biomass production– Maintenance respiration (Rm) –


Rm = Coef. dry weightCoef. = Coef. at 25°C Q10

{(temp.-25)}/10

Example:

Question: Assume at 25°C, Coef. is 0.014 g(CH2O)g1(DM) and Q10=2.0. What is Coef. at 15°C ? And at 20°C ?

Answer: 0.007g(CH2O)g1(DM) and 0.010g(CH2O)g1(DM)

Simulation of biomass production– Maintenance respiration (Rm) cont. –


Rm is assumed to have priority above growth

Rm results from maintenance processes likeprotein turnover and ion fluxes over membranes

If Rm > Pgd then Rm is set equal to Pgd; growth is zero.

20

Simulation of biomass production– Growth respiration –


Cf = growth conversion efficiency (g DM g1 CH2O)

Growth respiration depends on chemical composition

(costs of turning sugars into fats, organic acids, etc., & costs for ion uptake)

Simulation of biomass production– Glucose requirements –

Average amount of glucose required (G; gCH2O g1DM) to produce 1 gram of the following biochemical groups

GCarbohydrates 1.24Proteins 2.70Lipids 3.11Lignin 2.17Organic acids 0.93Minerals 0.05

21

Simulation of biomass production– Glucose requirements (cont.) –

Glucose requirement (G; gCH2O g1DM) and conversion efficiency (Cf; gDM g1CH2O)

G CfYoung leaf(carbohydrates + proteins)

1.656 0.60

Oil rich seed(fats and proteins)

2.572 0.39

Woody stem(carbohydrates + lignin)

1.569 0.64

Root sugar beet(carbohydrates)

1.271 0.79

NB: G * Cf = 1

Simulation of biomass productionDry Mass Production

22

The processes: step by step - Growth

Leaf area index

Light interception




Light use efficiency (LUE, g MJ1) in winter: 3 plant densities

and - or + assimilation light for cut chrysanthemum

23


dW/dt = LUE (1-e k LAI) I

dW/dt = growth rate ( g DM m 2 d 1)

LUE = light use efficiency ( g DM MJ 1PAR)

k = extinction coefficient



(MJ m 2 d 1)


Light use efficiency (LUE, g MJ-1) in summer at 3 light levels

(shade screens) for cut chrysanthemum

A fixed value for the LUE, will not work

over the whole

season/year, because LUE is depend on LIGHT LEVEL

24

Simulation of biomass production– Simulated light use efficiency for crop with LAI = 3 –

–

Very high light

intensity will give

lower LUE because of

light saturation



dW/dt = Crop Growth Rate (g DM m-2 d-1)

Pgd = Crop gross Assimilation Rate (g CH2O m-2 d-1)


R m = Maintenance Respiration Rate (g CH2O m-2 d-1)

f (Organ Dry Weight, Temperature, RGR)

Cf = Conversion Efficiency (g DM g-1CH2O)

25


Why can we expect that the relative effect of temperature on crop growth is larger at low light than at high light ???

In other words: crop growth rate changes with a different percentage when temperature rises from 15 to 25oC at low light, compared to high light levels. Why ?


Temp Low light intensity High light intensityPg Rm dW/dt=Cf(Pgd Rm) Pg Rm dW/dt=Cf(Pgd Rm)

15°C 10 4 6*Cf 30 4 26*Cf25°C 10 8 2*Cf 30 8 22*Cf

% decrease (6-2)/6=66% % decrease (26-22)/26=16%

At higher temperature gross assimilation rate (Pg) doesn’t change much, but 10°C difference will lead to twice as much as respiration rate (Rm).

Light intensity doesn’t have effect on Rm, but will influence Pg. From the table, we can see that crop growth rate changes with a

different percentage (66% at low light and 16% at high light) Rm is not depend on light level. At low light intensity, Rm is the main

part of gross assimilation. You can choose any number (logical), many numerical examples are

possible, but they will give you the same conclusion

26

Simulation of biomass production– ‘Homework’ question –

Assume that salinity stress increases maintenance respiration rate.

Why can this explain that elevated CO2 reduces salt stress in wheat?

Please answer this question by yourself, and you can refer to the answers from previous question.


dW/dt = LUE (1-e -k LAI) I

dW/dt = growth rate ( g DM m-2 d -1)

LUE = light use efficiency ( g DM MJ -1 PAR)

k = extinction coefficient



(MJ m-2 d-1)


Be aware of differences in dimensions

27

Simulation of biomass production- Be aware of differences in units !!! -

For example:A radiation use efficiency of 3.0 g MJ-1 PAR

is the same asA radiation use efficiency of ?? g MJ-1 global

radiation

1 PAR = 2 global radiation

Global radiation PAR LUE2 1 31 0.5 1.5

Simulation of biomass production- Leaf area development -

Function of plant developmental stage (temperature)[works reasonably well in the field, not in greenhouses]

Predicted from simulated leaf dry weightSLA needed (Specific Leaf Area = leaf area/leaf mass; cm2 g-1)• constant• function of plant age/physiological age• function of season• environmental conditions

28

Simulation of biomass production- Leaf area development (cont.) -

In model LINTUL:When LAI < 0.75 m2 m-2 and TSUM < 330 oCdLeaf growth exponential and RGRL depends on

temperature

ElseSLA = constant and increase in LAI calculated as

SLA increase in leaf dry weight (g m-2)

Simulation of biomass productionSeasonal effect on SLA (Leaf area/Leaf mass)

(regression on 12 experiments)

29

Simulation of biomass production- From fruit dry weight to fresh weight -

Fig.4. Time course of measured(symbols) and simulated (lines) DMcontent of harvestable cucumberfruits during a growing season in agreenhouse at a daily temperatureof 23.2±0.4°C. DM content wassimulated as a function of fruit age,temperature and fresh weightaccording to Marcelis (1992b) (——)or a positive linear relationship tothe source/sink ratio of the past 5days was added to the formula (----);this relation was derived from theexperiment shown. Experimentalconditions and calculation ofsource/sink ratio are described byMarcelis (1992a, 1994).

Simulation of biomass productionCompare model with experimental data

30

Simulation of biomass production- Validation of dry matter production model -

Measured (symbols) and simulated (lines: model TOMSIM) dry matter production for tomato in Dutch greenhouses at 5 planting dates. Relevant model input: daily outside radiation integral, hourly greenhouse temperatures and CO2 concentrations, crop leaf area index.

(Heuvelink, 1995, Annals of Botany 75: 369-379).

Symbols and lines correlate

with each other Dry matter production is predicted well for all these

situations using the model

Simulation of biomass production- Measured and simulated total dry weights -

Logic patterns? Beginning:Plant density↑LAI ↑ light intercepted ↑ Crop growth rate ↑

In the end:LAI is not a limiting factor; amount of light intercepted are almost the same no big difference among crop growth rates parallel

31

Simulation of biomass production- Examples of simulation of crop growth -

0

4

8

12

16

50 70 90 110 130

Day of year

Dry

we

igh

t (g

/ p

lan

t)

32 plants / m2

64 plants / m2

From: J‐H Lee, E. Heuvelink & L. Ortega, unpublished

Note: y-axis is per plant! Chrysanthemum

Total dry weight (g m2) = plant dry weight (g plant 1) * plant density (plants m 2)

Salinity Effects on Fruit Yield in Vegetable Crops: a simulation study

Ep Heuvelink, Cecilia Stanghellini and Menno Bakker

Wageningen University and Research Centre

32

Salinity Effects on Fruit Yield in Vegetable Crops- via plant-water relations -

Increase fruit dry matter contentBut reduce fresh yield (less dilution with water)

Reduce leaf expansionNot enough turgor to extend leaf area

Less LAI, less light interception, less yield

Stomatal closure Stomata partly close, to reduce transpiration

Higher resistance for CO2, less Photosynthesis

Aim: Determine salinity impact on fruit yield of single and combined effects

Approach: Simulation study (model TOMSIM)

Nutrients in high salinity solution

reduce the water potential

at soil level, and consequently in the whole plant

(via osmotic potential of the soil solution)

Salinity Effects on Fruit Yield in Vegetable Crops - Crop growth: principles of the simulation model -

Leaf area index

Light interception



LeavesRoot Stem Fruits

SLA

DMC Yield

Model with NO salinity stress

33

+

Salinity Effects on Fruit Yield in Vegetable Crops - Crop growth model: influence of salinity -

Leaf area index

Light interception



LeavesRoot Stem Fruits

SLA

Salinity

DMC Yield

Negative effect Positive effect

+

+

Salinity Effects on Fruit Yield in Vegetable Crops-Quantifying salinity effects -

Fruit dry matter content (DMC)5% at EC=2 dS m-1; linear increase by 0.2% per dS m-1

Leaf Area Expansion- SLA according to seasonal pattern- decrease 8% per dS m-1 (from threshold 3 or 6 dS m-1)

Stomatal resistance50 s m-1; increase with EC over the range 1 to 10 dS m-1

by either a factor 2 or a factor 4

Li et al. (2001); Li & Stanghellini (2001)

34

Salinity effects on fruit yield in vegetative cropsDry matter production: explanatory crop growth model


dW/dt = Crop Growth Rate (g DM m2 d 1)

Pgd = Crop gross Assimilation Rate (g CH2O m2 d 1)


R m = Maintenance Respiration Rate (g CH2O m2 d 1)

(Depends on Organ Dry Weight, Temperature, RGR)

Cf = Conversion Efficiency (g DM g1CH2O)

Salinity Effects on Fruit Yield in Vegetable Crops- the simulation model: Dry Matter partitioning -

Regulated by the sinks (fruit trusses and vegetative units)

Sink strength = demand for assimilates = potential growth rate

One common assimilate pool

More information is given in Chapter 5, dry matter partitioning

35

Experiments for Salinity Effects simulation settings

Crop: Start January 10; flowering of first tomato truss

End: September 7 (day 250)

Plant density 2.5 plants m-2

No side shoots retained; no decapitation

All trusses 7 fruits (no abortion simulated)

Greenhouse climate: 20°C; 350 ppm CO2;

average radiation The Netherlands

Experiments for salinity effects Results (1)

0

10

20

30

40

0 2 4 6 8 10 12

EC (dS/m)

Fru

it Y

ield

(kg

/m2 )

DMC

RS2

RS4

Both these two factors: Increased fruit dry matter content and stomatal resistance have modest impact on yield (2-3 % yield loss

per dS m-1)

Impact on fruit yield by Increased dry matter content (DMC) or Stomatalresistance (RS2 - factor 2; RS4 - factor 4)

36

0

10

20

30

40

0 2 4 6 8 10 12

EC (dS/m)

Fru

it Y

ield

(kg

/m2 )

DMC

SLA3

SLA3+DMC

RS2

RS4

Strongest salinity effects result from effects on leaf expansion (reduced SLA) (12-17% yield loss per dS m-1)

Experiments for salinity effects Results (2)

Impact on fruit yield by reduced SLA • 8% per (dS m-1)• threshold 3 dS m-1

Influence of Leaf Area Index (LAI) on the fraction of light

intercepted by a tomato crop (k = extinction coefficient = 0.8)

•Higher EC at low LAI has bigger impact on light interception. •To reduce yield loss caused by high EC, high LAI is preferred

37

Delay leaf picking

Increase plant density

What can a grower do to mitigate salinity

effect via leaf expansion ?

Salinity Effects on Fruit Yield in Vegetable Crops- how to mitigate salinity effect via leaf expansion -

2 Hypotheses:

• To verify the hypotheses, following experiments were conducted ...

Salinity Effects on Fruit Yield in Vegetable Cropsmitigate salinity effect - influencing Leaf Area Index (LAI)

0

1

2

3

4

1 3 5 7 9 11EC (dS/m)

LA

I (

m2 /m

2 )

control

delayed leaf picking

increased plantdensity

• LAI is drastically declined with high EC level, because of decreased SLA• Increased plant density and delayed leaf picking gave relatively higher

LAI

38

Salinity Effects on Fruit Yield in Vegetable Cropsmitigate salinity effect – effect on fruit yield

0

10

20

30

40

0 2 4 6 8 10 12

EC (dS/m)

Fru

it Y

ield

(kg

/m2 )

control

delayed leaf picking

increased plantd itIncreased plant density

• Delayed leaf picking and increased plant density has positive

effect on reducing yield loss caused by high EC.

Conclusions:

Salinity Effects on Fruit Yield in Vegetable Crops- conclusions -

Increased fruit dry matter content or stomatal resistance

modest impact on yield (2-3 % yield loss per dS m1)

Delayed leaf picking or increased plant density mitigate

this negative effect

Strongest salinity effects result from effects on leaf

expansion (12-17 % yield loss per dS m 1)

Within the ranging of EC between 1 and 11

39

Simulation of biomass productionConcluding remarks

Simulation of biomass productionSimulated effect of number of fruits per truss in tomato

0

1

2

3

4

5

0 5 10 15

# fruits per truss

Dry

we

igh

t (k

g/m

2)

or

LA

I or

Fra

cti

on av_LAI

W_total

W_fruit

fraction_fruits

1: There’s an optimum number of fruit per truss 2: The number of fruits per truss defines partitioning fraction3: More fruits per truss less to vegetative parts Lower LAI less photosynthesis less total dry weight

40

Interaction (feedback) between assimilate production and assimilate partitioning

Simulation of biomass production- Conclusions on explanatory models -

Strong features: light interception gross photosynthesis

Weak features: leaf area development maintenance respiration organ initiation/abortion dry matter content product quality

Documents

Simulation of biomass production - Cirad