3rd Africa Rice Congress
Theme 2: Intensification and diversification
Mini symposium: determinants of agricultural productivity in Africa’s rice-based systems
Author: Dingkuhn et al.
Developer Data Modeling Mistakes: From Postgres to NoSQL
Th2_Integrating Physiology, Crop Modeling and Genetics
1. GRiSP
Integrating Physiology, Crop
Modeling and Genetics
to Tackle Thermal
Stresses in
Rice:
The RIDEV Approach
Michael Dingkuhn (IRRI/CIRAD), Julie Mae
Pasuquin (IRRI), Cecile Julia (CIRAD), Richard
Pasco (IRRI), Jean-Christophe Soulie (CIRAD)
funded by GIZ, AfricaRice, CCAFS and CIRAD
Context of GRiSP Global Rice Phenotyping
Network
2. GRiSP Rationale
Thermal adaptation is fundamental for agro-ecological fit
Temperature governs rice phenology and spikelet fertility
Climate change is changing thermal environments
Accuracy of crop models is still poor re: thermal effects
We need…
Better predictive tools to map climate change impact
Better understanding of adaptive traits: Physiology &
Genetics
3. GRiSP
History: The 1990s research at WARDA
Thermal constraints to irrigated rice in Senegal
Effect of sowing date on crop duration and sterility
Days to flowering
• Thermal and photoperiod effects on phenology
• Chilling causes spikelet sterility
% sterility
Sowing date
Sowing date
% sterility
Tw(min) at booting
1995 development of RIDEV predicting phenology and thermal
sterility as risk analysis and decision aide for cropping calendars
4. New study on rice phenology and sterility
GRiSP
responses to T
(Thesis of Cecile Julia & ongoing CIRAD/IRRI/CCAFS project)
Emphasis on microclimate
NEW
NEW
Meristem T => phenology
Floodwater T => chilling stress at microspore stage
Panicle T => heat stress at anthesis
Time of day of anthesis (TOA)
RIDEV v.2 to characterize genetic diversity
5. Philippines - Hot and dry season 2009
36
32
28
24
20
16
12
8
01/03
32
28
24
20
16
12
11/03
21/03
31/03
10/04
20/04
8
10/05
30/04
20/05
Senegal - Cold and dry season 2010
44
40
36
36
Temperature (°C)
Phenology
TOA
Panicle transp. cooling
19/06
France - Temperate Summer 2009
40
Temperature (°C)
Traits observed
5
4
3
2
1
0
44
09/06
VPD (KPa)
4 environments
30/05
Date
Date
DS Philippines
HDS Senegal
CDS Senegal
Temp. summer France
VPD (KPa)
36
5
4
3
2
1
0
40
Temperature (°C)
Temperature (°C)
IR64
IR72
Sahel108
Chomrong
(N22 failed)
44
40
4 genotypes
44
5
4
3
2
1
0
VPD (KPa)
Scope of study:
Senegal - Hot and dry season 2010
VPD (KPa)
GRiSP
5
4
3
2
1
0
32
28
24
20
16
32
28
24
20
16
12
12
8
15/01 25/01 04/02 14/02 24/02 06/03 16/03 26/03
Date
8
01/08
Tair Max
Tair Min
Twater max
Twater min
11/08
21/08
31/08
Date
10/09
20/09
30/09
6. GRiSP
Results
Time of day of anthesis (TOA)
shows adaptive plasticity
Warm nights advance TOA =>
Escape from midday heat
Humid days advance TOA =>
Escape from heat caused by absence
of transpiration cooling
Mean air temp (min) during last 7d before anthesis (oC)
7. GRiSP
Panicle temperature: IR imagery in the field
Pan2
Pan1
Flagleaf4
Flagleaf1
Flagleaf2
Flagleaf3
Leaf5
Pan3
Pan4
ca. 4900 IR observations on in-situ panicle T
Microclimate recording
% sterility observed at maturity
8. Relative humidity or vapor pressure deficit is the
main determinant of Ta-Tp difference
GRiSP
14
Senegal cool-dry
season
Senegal hot-dry
season
France summer
12
14
10
TD (observed) [°C]
c
Example: Senegal cool-dry season
12
TD=Ta-Tp (°C)
10
y = 1.45x - 0.99
R² = 0.79
1:1
8
6
4
2
0
-2
8
-4
-4
6
-2
0
2
4
6
8
TD (predicted) [°C]
10
12
14
Model prediction (sim:obs)
4
Panicle cooler
than air
2
0
Panicle warmer
than air
-2
-4
0
1
Humid
2
3
4
VPD (kPa)
5
6
Arid
7
9. GRiSP The panicle is warmest not in the hottest, but in the
most humid environment
(b)
Air and Panicle Temperature at TOA (calculated)
32
Temperature (°C)
30
28
26
24
22
PHIL_DS
Phils
SEN_HS
Sen.-hot
Site
SEN_CS
Sen.-cool
FR_HS
France
10. Temperature induced spikelet sterility
Chomrong
100
IR64
S108
IR72
(c)
90
80
Sterility (%)
GRiSP
70
Disaggregate observed
sterility into its components
Incomplete panicle exertion
Chilling at microspore stage
Heat at anthesis (at TOA)
60
50
40
30
20
10
0
Phils
PHIL_DS Sen.-hot
SEN_HS Sen.-cool
SEN_CS
Site
France
FR_HS
11. Incomplete panicle exertion
GRiSP
occurred in cold-night environments
explained some of observed sterility
Chomrong
Panicle exsertion (%)
160
IR64
S108
IR72
Last grain
Neck node
(b)
140
120
100
Sterile
fraction of
panicle
caused by
non-exertion
80
60
40
PHIL_DS
Phils
SEN_HS
SEN_CS
Sen.-hot
Sen.-cool
Site
FR_HS
France
12. GRiSP 2. Chilling effect at microspore
stage on sterility
(Tmeristem = Twater)
100
Phil-ds
Sen-cs
90
Sen-hs
Fr-hs
Sterility (%)
80
70
60
50
40
Chomron
30
20
10
0
12
14
16
18
20
22
24
26
28
T water (min) at microspore stage (°C)
3. Heat effect at flowering
stage on sterility (Tp at TOA)
13. GRiSP
Conclusions from experimental study
Rice has highly effective adaptations to thermal stresses:
Avoidance
Transpiration cooling of panicle
Good panicle exertion (long peduncle)
Escape
Time of day of anthesis (TOA) and its adaptive plasticity
Tolerance
To cold, as shown for cv. Chomrong
Heat tolerant check cv. N22 failed (seed problems)
Heat stress more likely in warm-humid than hot-dry climates!
14. GRiSP A new modeling tool RIDEV V.2
Simulator of…
Phenology incl. microclimate & photoperiod effects
G and E effects on TOA
Sterility caused by…
Chilling effects on microsporogenesis (water Tmin)
Chilling effects on panicle exertion (air Tmin)
Heat effects on pollination (Tpanicle at TOA)
Prediction (forward mode)
Climate change impact mapping, plant type optimization
Agronomy (crop calendar; optimization)
Heuristic parameterization of genotypes (reverse mode)
Phenomics (extraction of genotypic parameter values
from experimental data)
15. GRiSP
Outlook
Use of RIDEV for Phenomics/GWAS
Indica GWAS panel (>200 acc., ORYTAGE project)
Field-phenotyped for phenology and sterility in 12 environments:
6 sowing dates in Senegal
3 altitudes x 2 years in Madagascar
Extraction of genotypic response parameters across
environments (Heuristics):
Cardinal temperatures Tb and To
Thermal duration of phenological phases
Photoperiod-sensitivity
Chilling sensitivity of microsporogenesis
Chilling sensitivity of panicle exertion
Heat sensitivity of anthesis
Association study using GBS and 700K Oryza SNP chip
Graph 1:Sowing date in the Sahel strongly affects crop duration even in modern varieties (e.g., WS vs. Hot-dry season)Graph 2:Sowing in September-November caused near-total sterility (cold). Sterility sometimes also high for sowing in Feb-Mar (heat)Graph 3:Much of the sterility could be explained with minimum T(water) at booting stage (ca. 2 wk before flowering)RIDEV:A 1st model simulating this was developed in 1995 and extensively used by NARS and ARC for risk analyses and crop calendar planning
Weakness of old RIDEV:No consideration of TOA and T(panicle)Need for new study focusing on micro climate and heat
Important:This is hours after sunrise- T-dependent shift in TOA by up to 4h (e.g., from 9 am (warm) to 1 pm (cool))
Manual scene takingManual image analysisLongitudinal and transversal gradient analyses
Observations at different times of dayRH or VPD is main determinant of Panicle-air T differenceRs, wind, solar angle etc. also have effectsThese effects were combined in a multiple regression model and used in RIDEV to predict panicle T for any given time of day.That model was compared with a mechanistic energy balance model (IM2PACT, Japan) which gave the same results
Highest spikelet sterility in Senegal cool seasonLowest in Senegal hot season (!)We wanted to disaggregate this sterility into 3 different fractions (causes)
Panicle exertion – relative position of the panicle neck node to the top of the enclosing leaf sheath (broken line); relative position of lowest spikelets on the panicle (solid line)Short-strawed high yielding rice – increased risk of incomplete panicle exertion (short peduncles)Sahel 108 – selected by breeders in Senegal aiming at avoiding bird damage and improving light use through panicles hidden deep in the canopy
Graph 1:At minimum water temperature below 18-20 oC, sterility increases, except the tolerantChomrongCold tolerance: involves anti-oxidative enzymes protecting the tissues, production of more pollen to increase probability of successful pollinationGraph 2:If cases of cold-sterility are taken out of the analysis, the remaining cases show a good correlation of sterility with heat at anthesis. But only if PANICLE T (not air!) and TOA (not Tmin or Tmax or Tmean) are used as reference.