european agroforestry federation | euraf - evidences and … · 2012-10-17 · meine van noordwijk...

Evidences and explanations for the unexpected high productivity of improved

temperate agroforestry systems

1rst EURAF Conference, 9 October 2012

Session 1

Christian Dupraz , Grégoire Talbot

INRA, Montpellier, France

Trees and Agriculture : a change of paradigm ?

Trees inside parcels

Trees around parcels

Around parcels

Inside parcels

5

Innovative Agroforestry

Dual approach : long-term experiments + models

Controlled Experiments

Agroforestry Agri-voltaism

8

2002 : Wageningen seminar

2005 End of SAFE project

2004 : Montpellier Seminar

The Hi-sAFe model : 2002-2011

2007

2011

Christian Dupraz (Inra) Grégoire Vincent (Ird)

Nick Jackson (Nerc) Harry Ozier-Lafontaine (Inra)

Hervé Sinoquet (Inra) Alain Fouéré (Inra)

Martina Mayus (Wageningen) François Bussière (Inra) Isabelle Lecomte (Inra)

Meine Van Noordwijk (Icraf) Betha Lusiana (Icraf)

Benoit Courbaud (Cemagref) Wopke van der Werf

(Wageningen) Hermann Van Keulen

(Wageningen) Gerry Lawson (Nerc)

Jean-Claude Poupa (Inra) François de Coligny (Inra)

Rachmat Mulia (Inra)

Objectives Hypotheses

Specifications

To mix or not to mix : trees and crops/animals…

Forest

Agroforestry

1 ha

Agriculture 0.8 ha

0.6 ha

LER = 1.4 ha

Land Equivalent Ratio (LER) (Mead and Willey, 1980)

Mixture Separation

1996

1997 2002

2005

2003

2009

2010

1.2 to 1.6 Land Equivalent Ratio

Poplars-winter cereals

14 years

A 1.4 LER means that a 100 ha agroforestry farms produces as much crop and tree products as a conventional 140 ha farm where trees and crops are separated

Part of a new green revolution Increased efficiency of natural production factors (light, water, natural nitrogen)

Measured LER for agrivoltaic systems with various densities of PV panels

1.3 to 1.7

Dupraz et al, 2011, Renewable Energy 36: 2725-2732

2003

2009

Solving the tree shade shape riddle

17

21 June

21 December 21 March / 21 September

Average irradiation under an isolated tree

Integrated irradiation under a tree stand

Light capture (40 years)

Light interception Land Equivalent Ratio : Walnut trees : 0.73 Winter wheat : 0.66 Light LER = 1.39

0 10 20 30 40 0 10 20 30 400 10 20 30 40

020

4060

8010

0 Pure crop Agroforestry Pure trees

% o

f inc

iden

t rad

iatio

n,

Year

Walnut trees Winter wheat Not captured

Water budget (40 years)

Water capture Land Equivalent Ratio : Walnuts : 0.71 Wheat : 0.84

Water LER = 1.55

Inputs: Rain Water table

Outputs Wheat transpiration Walnut transpiration Understorey transpiration Soil evaporation Drainage Run-off

A AF F

Wat

er fl

uxes

(mm

/yea

r)

0 20

0 40

0 60

0 80

0

A AF F

Inputs Outputs

22

Pure walnut grove Walnut wheat Agroforestry

Root profiles of 14 year old walnut trees (November 2009)

Experimental measures

Tree roots systems influenced by crop competition

23

Agroforestry tree

12 years

Forest tree

6 ans

Model predictions

0 10 20 30 40

02

46

8

0 10 20 30 40

02

46

8

Wheat yields predicted

Blé précoce Blé tardif

Année Année

Yiel

d (t

.ha-

1 )

• Pure wheat : high variability in yields

Témoin agricole

Model predictions

0 10 20 30 40

02

46

8

0 10 20 30 40

02

46

8

Wheat yields in agroforestry

AF Late walnut AF Early walnut

Témoin agricole

Année Année

• AF : yield and variability (more impact of early walnut)

Yiel

ds (

t.ha-

1 )

Early wheat Late wheat

Model predictions

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0 0.5 1 1.5 2Tree crowding index

Rel

ativ

e cr

op y

ield

Vézénobres_1996 Vézénobres_1997 Restinclières_East-West Restinclières_North-South

Research for action : guidelines for improved AF systems

Looking for the best tree density

Experimental measures

LER interpretation

• LERs can be simulated with process-based models • The simulations allow to decompose the LER in various

interacting effects • This is a way to hierarchy factors of success in agroforestry

and agrivoltaism

Talbot, 2011, Ph. D. thesis (Env. Model. Soft. In press)

LER interpretation for a walnut-wheat AF system

Land Equivalent Ratio : Walnuts : 0.76 Wheat : 0.52

LER = 1.28

Modelling the 3 options concurrently

7m 7m

4m 4m

1m

Pure crop Pure tree

Same soil Same climate

Same management 40 years

Spontaneous understorey

4m

9m 13m

1 m

Decomposing the LER

• LER= Rytree + Rycrop

• RY =π[relative indexes] • Exemple of decomposition :

• RY = Relative density x Size Memory x Light competition x Water competition x Harvest Index

Decomposing the relative yield of trees

Interception light: • a : interception capacity • i : competition

Conversion into biomass: • p : leaf aging • w : water stress

Harvest ratio • ha : ABG/BG allocation • hs : Trunk allocation

yind a i p w ha hs

Relative Yield per tree

-0.4 0.0 0.4 0.8


yind a i p w ha hs

Relative Yield per tree

-0.4 0.0 0.4 0.8


19 m x 19 m 19 m x 11 m 11 m x 9 m 11 m x 5 m 7 m x 7 m

yind a i p w ha hs

-0.4 0.0 0.4 0.8

Interception capacity (leaf area, leaf duration)



yind a i p w ha hs

-0.4 0.0 0.4 0.8

Light competition • On the tree line space essential



yind a i p w ha hs

-0.4 0.0 0.4 0.8

C Allocation to trunk: • Negative • Large canopies -> C allocated to branches



Decomposing the crop yield

Light Interception : • a : interception capacity • i : competition by trees

Conversion to biomass : • p : phenology • n : nitrogen stress • s : light saturation • t : temperature stress • w : water stress

Harvest index • g : grain filling duration • u : thermal stress • f : grain number

-40 -30 -20 -10 0 10

y a i p n w s t g u f

0.0 0.5 1.0 1.5 2.0 2.5 3.0

-1.0

-0.5

0.0

0.5

1.0

1.5

Log(

Ren

dem

ent r

elat

if)

LAI peuplement noyers (m2.m-2)

Relative yield


Relative Yield

0.0 0.5 1.0 1.5 2.0 2.5 3.0-1

.0-0

.50.

00.

51.

01.

5Lo

g(R

ende

men

t rel

atif)


-40 -30 -20 -10 0 10

Walnut Early Late

Whe

at

Late

Ear

ly



Light Competition : • Major and negative effect • Strong under early walnut

0.0 0.5 1.0 1.5 2.0 2.5 3.0

-1.0

-0.8

-0.6

-0.4

-0.2

0.0


Log(

i) 44

-40 -30 -20 -10 0 10


Decomposing the crop yield Walnut

Early Late

Whe

at

Late

Ear

ly

Water Stress Thermal Stress during grain filling

-40 -30 -20 -10 0 10



Early Late

Whe

at

Late

Ear

ly

Interception capacity • Delayed senescence

TA

AF

-40 -30 -20 -10 0



Early Late

Whe

at

Late

Ear

ly

Number of grains

-40 -30 -20 -10 0 10



Early Late

Whe

at

Late

Ear

ly

If the model is right…

Advantages in AF Drawbacks in AF Neutral

Trees Larger canopies

Less light competition

Few trees

Invest C in roots

Water stress

WinterCrops

Reduced water stress

Longer life of leaves

Reduced N stress

Reduced heat stress

Reduced cropped area

Less light

Harvest index (?)

Integrated over 40 years

But what dynamics during the 40 years?

0,0

0,3

0,5

0,8

1,0

1,3

1,5

1,8

2,0

0 5 10 15 20 25 30 35 40Years

Land

Equ

ival

ent R

atio

AnnualIntegrated

Association type Plant Cycles Soil depth

Root segre-gation

Water body in summer

Probable LER

Walnut-cereal on plains with a water table

Almost complementary

Deep High Yes 1.5

Sorbus-wheat Lagged Medium Intermediate No 1.4

Prunus-Medicago Close Deep Intermediate No 1.3

Prunus-sunflower Synchronous Shallow Poor No 1.2

Populus-maize Synchronous Poor No Yes 1.1

Evergreen tree and summer crop

Superposed Any Variable No 1.0

Never demonstrated so far <1.0

A synthesis of probable LERs of temperate AFs

Conclusions

• LERs of AFs may be extremely high (part of a second green revolution?)

• Process-based models can help in interpreting productivity results of AF systems at various time steps.

• Winter crops better in AF (temperate latitudes-specific)

Pending : more virtual experiments

11 m x 9 m 100 stems/ha


7 m x 7 m 200 tiges/ha (forest control) 19 m x 11 m

50 stems/ha


Tree density and plantation design

Pending : more virtual experiments

AF and Climate change : are crops in AF protected against climate change hazards ?

Pending : System management optimisation assisted by modelling

Tree line orientation North-South / East-West

133 trees.ha-1 64 trees.ha-1

10 10

10

10

6

6 6

6

NS NS EW EW

Pruning height : 10 m versus 6 m

Tree density

Pending : Designing enhanced and innovative agroforestry systems

Biomass production in agroforestry Various tree managements (pollarding, root trenching, precision fertilisation…) Mixtures of tree species

european agroforestry federation | euraf - evidences and … · 2012-10-17 · meine van noordwijk...

Documents