Engler and Prantl system of classification in plant taxonomy
2015. Greg Rebetzke. Integration of new tools for improving genetic gain of grain yield in water limited environments.
1. Integration of new tools for improving genetic gain of
grain yield in water-limited environments
CSIRO AGRICULTURE FLAGSHIP
Greg Rebetzke, Bill Bovill, David Deery, Jose Jimenez-Berni, Anton Wasson, Richard
James and Lynne McIntyre
3. 1. Trait value?
• Are the traits relevant to the target environment(s)?
• What is the value proposition?
• What are the trade-offs?
• Can I scale up from cell, single-plant and row to canopy?
• What do I give up in order to accommodate the new trait/germplasm?
2. Selection?
• How do I phenotype? Is it quick, cheap and reliable?
• Heritability and the correlation of phenotype with genotype?
• Genetic complexity/QTL/markers?
• Influence of genetic background/repeatable?
• Quality of donor germplasm?
• Correlated response?
3. Adoption?
• How does the gene/trait fit in the target farming system?
The questions a breeder is asking are very different to those
being asked by researchers further upstream
4. Grain
Yield
= Water
Use
x Water-use
Efficiency
x Harvest
Index
Biomass
The model for productivity under water-limitation:
T/ET - rapid canopy growth to shade the soil surface and restrict evaporation
DM/T - transpiration efficiency (or leaf-level WUE)
HI - C partitioning to the growing spike (grain number) and then to the grain (grain
size) (balance water use before and after flowering to optimise harvest index)
T/ET x DM/T
Drought resistance? No!
Water use efficiency as a breeding target. Yes!
7. Many potential traits to improve crop performance under
drought
New root
architecture
Stem CHO for large grain
Genes for better emergence
Transpiration efficiency
C13
C12 Glaucousness
Vigorous early growth
Reduced tillering
Developmental genes
Which traits where?
Trait dissection
Trait delivery
8. Need for repeatable phenotyping - controlled ‘managed’
environments (Managed Environment Facilities – ‘MEF’)
In Australia - three sites with two-three irrigation regimes
9. Which traits where? Quality phenotyping – controlled field
environments (Managed Environment Facilities – ‘MEF’)
10. Maximising water uptake by removing constraints to root
(and shoot) growth – mapping of soil conductivity
(Rick Graham NSWDPI)
11. Break crops to reduce root disease
Canola in the rotation
Wheat after Wheat Wheat after Canola
*** Reduce soil-borne diseases
(take-all, crown rot, CCN, root-
lesion nematodes)
(Kirkegaard CSIRO)
12. Random to selected lines – deriving greater benefit from
populations in selection of tails extreme for target trait
High-selected groupLow-selected group
13. ‘Traits’ germplasm
Germplasm Background(s) Number of lines
Canopy temperature Multiple 20 lines/tail
Development Single 13 near-isogenic pairs
Early vigour Multiple 10 lines/tail
Grain fertility Multiple 20 lines/tail
Grain size/screenings Single 10 lines/tail
Ear morphology Multiple 24 near-isogenic pairs
Reduced-tillering Multiple 20 near-isogenic pairs
Staygreen (leaf) Multiple 10 lines/tail
Stem carbohydrates Multiple 15 lines/tail
Transpiration efficiency Multiple 10 lines/tail
14. Box-plot of all entry means/variance for grain yield at each sampled MEF
(M = Merredin, N = Narrabri, Y = Yanco; Ir = irrigated, Rf = rainfed)
16. Grain yield of trait germplasm is comparable to that of
commercial checks at the MEF
Note that the traits are tested in germplasm developed in current or recent commercial wheat
backgrounds and that the trait lines used in the MEF are not selected for grain yield
21. Trait value for grain yield at each MEF (2010-14)
Trait value is calculated as the percentage change in mean yield of lines selected for the trait
relative to sister lines that lack the trait at each MEF location in one to four years of testing
Trait
C
anopy
tem
perature
C
arbon
iso.discrim
ination
Stem
carbohydratesPresence
ofawns
Leafwaxiness
Leaf-rolling
Early
vigourR
educed
tillering
Trait(yield)value(%)
-10
0
10
20
Merredin
Narrabri
Yanco
22. Trait value for grain yield at each MEF (2010-14)
Trait value is calculated as the percentage change in mean yield of lines selected for the trait
relative to sister lines that lack the trait at each MEF location in one to four years of testing
Trait
C
anopy
tem
perature
C
arbon
iso.discrim
ination
Stem
carbohydratesPresence
ofawns
Leafwaxiness
Leaf-rolling
Early
vigourR
educed
tillering
Trait(yield)value(%)
-10
0
10
20
Merredin
Narrabri
Yanco
24. Dynamic phenotyping
Pt = G×Et (×M)
with t: time
Time constant
Seconds Hours Days Season
Canopy
Temperature
Canopy
Height
Crop
Yield
25. Phenomics from the pot to the paddock
Thermal
Lidar /Colour 3-D
Chlorophyll fluoresence
Hyperspectral
High spatial or temporal
resolution
26. Real-time / hourly continuous measurements of canopy temperature using ArduCrop
Wireless Sensor Network (US$300 per sensor)
27. Diurnal patterns of canopy temperature as a surrogate for
stomatal behaviour. Yanco MEF 2014 (GRDC)
28. Above (well above) the canopy measuring canopy temperature/leaf area
FLIR SC645 thermal camera
Calibrated < 0.05 deg C sensitivity,
2% accuracy, 640x480 pixels, 0.7 kg
29. Linking phenotypes to traits and genetic architecture: Canopy
temperature and water use
“Old way” h2<0.1
“New way” h2>0.6
Using in irrigated
trials for photosynthetic
screens, and drought trials
for water use
30. Wireless data logger and radio transmitter for monitoring soil moisture
and soil temperature
33. Early vigour (leaf area) and water use efficiency
Soil
evaporation
rapid early growth slow early growth
Soil
evaporation
Plant
transpiration
Plant
transpiration
34. Genetic complexity - an example with early vigour -
Partitioning of water use
Fertility
treatment
LAI
(lai.days)
Yield
(t/ha)
Water use
(mm)
Evaporation
(mm)
Transpiration
(mm)
High
63N, 20P
3.1 5.6 366 173 193
Low
8N, 10P
1.4 2.8 363 259 104
(David Hall, DAFWA)
Esperance 2001, 380 mm in-crop rainfall
35. Traits for Greater Early Vigour
Embryo
size
Leaf lamina
thickness
Seedling
vigour
High vigour
New physiological typesCurrent cultivars
Low vigour
36. Global Survey for Early Vigour
Entry Mean leaf width
(mm)
Leaf area
(cm2)
Jing Hong
(China)
6.3 14.3
Kharchia
(India)
6.2 14.2
V743/Oligo
(Israel)
5.9 / 6.3 11.1 / 14.6
Glenlea/Roblin
(Canada)
5.7 / 5.8 12.0 / 12.2
CC-CIMMYT
(Mexico)
5.6 13.9
Janz
(Australia)
4.5 7.4
Where available, pedigrees indicate coancestry among lines is low
37. Trait
value
0
Cycle of Selection
1 2 3 4
Recurrent selection for genetic gain (accumulating
favourable additive genetic effects)
Existing
value
Target
value
38. Genetic covariances and variances
Cov (a,e) = 2ae2
A + 2ä+ë2
D + (2äe+2aë) D1 + äë D2
Var (S0 families) = 2
A + 2
D
Var (S0:1 families) = 2
A + 0.25 2
D + 1 D1 + 0.125 D2
Var (S1:2 families) = 1.5 2
A + 0.125 2
D + 2.5 D1 + 0.563 D2
Var (S families) = 2 2
A + 0 2
D + 4 D1 + D2
Where 2
A are 2
D are the additive and dominance genetic variances, D1 is the
covariance of an additive effect of an allele with its dominance deviation and D2 is
the variance of homozygous
dominance effects
39. Genotypic variation and covariation for early vigour
Parameter h2 ra_LFA RSG_LFA
(%)
Mean leaf
width
0.84 ± 0.11** 0.57 ± 0.10** 92
Mean leaf
length
0.67 ± 0.16** 0.43 ± 0.09** 64
Number of
leaves
0.39 ± 0.11** -0.37 ± 0.16** -10
+ Based on F2:4 - F2:6 parent-offspring covariance
44. Higher vigour germplasm with greater leaf area.
Developed originally to improve WUE and weed competitiveness
Commercial
variety
Early vigour
selection
Advanced vigour
selection
….above ground differences – also reflected below ground!
Higher vigour germplasm with greater leaf area
Developed originally to improve WUE and weed competitiveness
45. Greater root biomass and root length of
advanced vigour selections
Genotypes
Janz
W
estonia
W
yalkatchem
Vigor18
50-4
37-6
38-19
92-11
Rootbiomass(g/plant)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Commercial cultivars
vigor18
Recurrent selections
Genotypes
Janz
W
estonia
W
yalkatchem
Vigor18
50-4
37-6
38-19
92-11
Rootlength(m/plant)
0
10
20
30
40
50
60
70
Commercial cultivars
vigor18
Recurrent selections
Commercial
cultivar
Advanced vigour
selection
Advanced vigour
selections
50 – 90% greater
root biomass
30 – 60% greater
root length
59. Alternative dwarfing genes allow selection of reduced
height, long coleoptile progeny
Janz
HM14aS
134mm
81mm
60. 1 2 3 4 5 6 7
Genetic dissection of coleoptile length†
Integration of multi-population, multi-environment mapping
A B D
Group
(C/H = Cranbrook/Halberd, MAGIC = Baxter/Chara/Westonia/Yitpi)
Cranbrook/Halberd MAGIC (4-way)
† QTL common at two or soil temperatures
61. 1 2 3 4 5 6 7
Genetic dissection of coleoptile length†
Integration of multi-population, multi-environment mapping
A B D
Group
(C/H = Cranbrook/Halberd, MAGIC = Baxter/Chara/Westonia/Yitpi)
Cranbrook/Halberd MAGIC (4-way)
† QTL common at two or soil temperatures
Yitpi +13mm
Baxter +5mm
Baxter +6mm
Westonia +6mm
Chara +7mm
Yitpi +6mm
Westonia +7mm
62. 1 2 3 4 5 6 7
Genetic dissection of early growth†
Integration of multi-population, multi-environment mapping
A B D
Group
(4-way MAGIC = Baxter/Chara/Westonia/Yitpi)
Kukri/Janz MAGIC (4-way)
† QTL at two air temperatures
Yitpi +23mm
Chara
+11mmWestonia
+15mm
Chara
+17mm
Chara
+13mm
Yitpi
+16mm
Yitpi
+15mm
Yitpi
+11mm
Westonia
+13mm
Chara
+8mm
65. How do we overcome the constraint on seedling growth with the
green revolution dwarfing genes?
240
260
280
300
320
0 1 2 3 4
Number of GA-insensitive Rht alleles
Celllength(µm)
Tall Doubled-
dwarf
(Keyes et al. 1989)
(r2 = 0.99)
69. Critical need to better link whole crop physiology for delivery to commercial breeding
programs:
1. Are traits relevant for the challenge being addressed?
2. How do we prioritise one trait over another (trait value - Managed Environments?)
3. Are there cheap, reliable, population-friendly high-throughput phenotyping methods?
4. Can phenotype be replaced with breeder-useful, linked markers in selection?
5. Can we deliver adapted germplasm containing key traits for crossing and validation?
6. In moving from traits singly, are there crop-gene models capable of assessing trait x trait
combinations?
In summary -