Canopy temperature as field phenotyping trait for rainfed-lowland rice breeding

program for drought tolerance

A. Audebert

RCI Project

• Objectives : to develop drought-tolerant cultivars with high yield potential in normal years and good yield under drought and other major stresses for each target environment.

• Sponsor : Generation Challenge Programme (GCP)• Target countries : Burkina, Nigeria and Mali• Target environment : Rainfed lowland ecosystem• Duration : 4 years• Partners : CIRAD, IRD, IRRI, INERA, IER, NCRI, CIAT and AfricaRice

Field phenotyping for drought tolerance

• Based on Infra-red thermography.

– Canopy temperature give an indication of the leaf surface cooling capacity by transpiration along environmental conditions

– Could be use as a trait for phenotyping • (indirect evaluation of drought)

– This trait depending of• Environmental conditions

– Air temperature– Wind speed– Solar radiation– Evaporative demand (VPD)

• Sol water conditions– Humidity / available

• Plant characteristics– Surface of canopy– Plant Architecture– Water status management

Difficulties to solve• Environmental conditions highly variable• Quick plant reaction

– Wind– Radiation

• Low equipment (1 camera and 1 technician)• Impossible to have one unique picture for the whole experiment

– Helicopter, plane– Drone

• Optimum 3-4 lines per image– 500 lines -> 160 pictures– 1 image per 30 sec– Tc canopy temperature

• Time for measurement – About 3 hours– Environmental stability

How to control the environment variability and compare results ?

Normalizing canopy temperatures

• Quantifying the water stress with standardizing canopy temperature by evaporative demand (CWSI)– Ta (Air temperature)– VPD (Vapor pressure deficit)– CWSI (Crop water stress index)

• Need simultaneously measurement of the evaporative demand – Weather station– Psychrometer measurement

• Humid and dry temperature

min)(max)(

min)()(

TaTsTaTs

TaTsTaTsCWSI

Phenotyping experiment

• Dry season 2012-2013• Field experiment

– Villavicencio station “Santa Rosa “(Colombia)• 250 varieties tested with 2 reps

– 230 Mars lines (IR64 * B6144-F-MR-6-0-0)– 10 controls repeated twice

• 3 row of 3 m long• Stress period 3 weeks (5/01-25/01/2013)

• Reproductive stage• Design

– Alpha lattice 8 sub-Blocs with 2 replications– 2 treatments– Complete randomization– 5 control varieties repeated

• Measurements – Canopy temperature (IR thermography camera)– Soil humidity with Aqua Pro system– Microclimatic data with Davis weather station

Methods• Soil humidity

– 60 AquaPro tubes,• Distributed in the field

• Weather data– Davis station (Vantage Pro 2)

• 1 minute delay

• Canopy temperature– NEC TH9100 M

• Human height• 500 pictures• Image analysis with Image processor

Field experiment results

0 5 10 15 200

5

10

15

20

25

30

35

Date

07/01 14/01 21/01 28/01 04/02 11/02

Top

-soi

l hum

idity

(%

)

4

6

8

10

12

14

16

Soil heterogeneity

Experimental design to control the soil heterogeneity

Soil desiccation

Phenotyping Results

CWSI

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Fre

quen

cy (

%)

0

4

8

12

16

20

Tc-Ta (°C)

-4 -3 -2 -1 0 1 2 3 4

Fre

quen

cy (

%)

0

5

10

15

20

25

Better lines with low temperatureE207, E153, E14, E110, E31, E16, E79, E167, E59, E242, E130, E217, E139, E239, E220, E30, E163, E57, E231, E127

Perspectives : Association studies

• QtlsUsing a mixed model with correction for structure and kinship, the association study detected some markers

• Based on ajusted values• Software Tassel or other• Determination of LOD• Qtls determination (P-Value)

• MARS

Conclusions

• High-throughput phenotyping with thermographic camera on field condition is possible– Could be improved by using drone

• The use of the CWSI allows to compare cultivars between them during the phenotyping time

• Good diversity is observed for transpiration• QTls could be determine with association studies

RCI project• Phenotyping is on going

– 3 countries (PhD students)

• Association study will be done further with SNPs

Thank you