This document discusses a project between researchers in Australia and Mali to develop drought-tolerant sorghum varieties for Africa and Australia. The project focuses on enhancing the "stay-green" trait which allows plants to maintain green leaves during drought. Researchers conducted studies in Australia comparing stay-green and senescent sorghum varieties under irrigated and drought conditions. Promising stay-green lines from these studies were used to develop new populations and introgression lines in Mali to select for improved yield, stay-green and local adaptation. Evaluation of developed varieties was conducted across multiple countries in Africa. The project also involved training African scientists in sorghum breeding and physiology.

1. Working together with the
Queensland Government
Developing drought-adapted sorghum
germplasm for Africa and Australia
University of
Queensland
Andrew Borrell
David Jordan
Queensland
Government
Barb George-Jaeggli
IER, Mali
Sidi Bekaye Coulibaly
Niaba Teme
Mamoutou Kouressy
CIRAD, Mali
Michel Vaksmann

2. Working together with the
Queensland Government
Outline
1. The „stay-green‟ trait
2. Overview of GCP project
3. Physiological studies in Australia
4. Stay-green germplasm for Mali
5. Conclusions

3. Working together with the
Queensland Government
Maintaining green
leaves and stems when
water is limiting during
the grain-filling period
1. The “stay-green” trait
Stay-green Senescent

4. Working together with the
Queensland Government
Crop productivity in water-limited environments is
regulated by:
Drought adaptation
Stay-green affects all three processes.
a) the extent of water capture by the crop (T),
b) the efficiency with which the crop
exchanges water for CO2 via transpiration in
producing biomass (TE), and
c) the fraction of the total biomass that ends up
in the grain (HI).

5. Working together with the
Queensland Government
Higher yield
Increased water use during grain filling
Increased water availability at flowering
Increased water
accessibility
(roots)
Reduced water
use at flowering
Higher plant
water status
Increased
growth rate
Increased
TE
Delayed leaf
senescence
Smaller
plant size
„Low tillering‟
mechanism
„Small leaf‟
mechanism
Modified leaf
anatomy
Driving T
Driving TE
Driving HI
Emergent
consequences
Increased N
uptake
Increased
stem strength

6. Working together with the
Queensland Government
2. Overview of GCP project
Development and evaluation of drought-
adapted sorghum for Africa and Australia

7. Working together with the
Queensland Government
African partners (Phase 1)
Mali
Institut d‟Economie Rurale (IER)
CIRAD, Bamako, Mali

8. Working together with the
Queensland Government
African partners (Phase 2)
Niger, Sudan, Ethiopia,
Kenya & Uganda

9. Working together with the
Queensland Government
Objective 1. Determine the impact of changes
in plant height and photoperiod sensitivity
on the expression and value of stay-green
Objective 2. Develop Malian germplasm
enriched for stay-green QTLs
Objective 3. Training of African scientist(s) in
the physiology and breeding of drought
adaptation
Objectives (Phase 1)

10. Working together with the
Queensland Government
Objectives (Phase 2)
Objective 1: Evaluating in Africa the material
produced in Phase 1
Objective 2
• Training in Australia for visiting African
scientists on sorghum crop improvement
• Training in Africa by visiting Australian
scientists on sorghum crop improvement
Objective 3: Evaluation of African germplasm for
known stay-green regions

11. Working together with the
Queensland Government
Knowledge gaps
Interaction between stay-green and plant height is
unknown.
Interaction with photoperiod response is unknown.
Value of the trait will be influenced by the types of
drought experienced in Mali.
Will stay-green be useful in
Mali ?
Challenges
Value of stay-green is more obvious in hybrids
compared with inbred lines.
Variation in flowering time complicates conventional
selection.
Selecting for stay-green while retaining local
adaptation (PPS, grain quality, pest resistance etc).

12. Working together with the
Queensland Government
3. Physiological studies in Australia

13. Working together with the
Queensland Government
Creating „managed‟ environments
Irrigated Control
(HD & LD)
Water-Limited Treatment
(HD & LD)
Low density
(mild drought)
High density
(severe drought)

14. Working together with the
Queensland Government
Genotypes
Tall & stay-green
(R931945-2-2)
Short & stay-
green
(R931945-2-2)
Tall & senescent
(R955343-1)
Short &
senescent
(R955343-1)

15. Working together with the
Queensland Government
Seasonal variation
2009: Post-flowering drought
(stopped irrigating 2 weeks
before flowering)
2010: Pre-flowering
drought (stopped
irrigating 3 weeks
before flowering)
2011: Post-flowering
drought (stopped
irrigating 3 weeks
before flowering)

16. Working together with the
Queensland Government
Retention of green leaf area
0
5000
10000
15000
20000
25000
Greenleafarea
(cm2
/m2
)
T S T S T S T S
R931945-2-2
(stay-green)
R955343-1
(senescent)
R931945-2-2
(stay-green)
R955343-1
(senescent)
High Low
2009
The stay-green pair retained more green leaf area at maturity than the senescent pair
under both high and low density treatments. There were no significant differences in
green leaf area between tall (T) and short (S) isolines of the stay-green pair.

17. Working together with the
Queensland Government
Grain yield
0
100
200
300
400
500
600
700
800
Grainyield(g/m2
)
T S T S T S T S
R931945-2-2
(stay-green)
R955343-1
(senescent)
R931945-2-2
(stay-green)
R955343-1
(senescent)
High Low
2009
The stay-green pair exhibited higher grain yield than the senescent pair under both high
and low density treatments. Grain yield was higher in the tall (T) than short (S) isoline of
the stay-green pair under both densities. Hence, the „tall stay-green‟ combination yielded
particularly well under post-flowering drought.

18. Working together with the
Queensland Government
Lodging resistance
-200
-150
-100
-50
0
50
100
Deltastemmass(g/m2
)
T S T S T S T S
R931945-2-2
(stay-green)
R955343-1
(senescent)
R931945-2-2
(stay-green)
R955343-1
(senescent)
High Low
2009
Under stressed conditions, the stay-green pair remobilized less stem reserves during grain filling than
the senescent pair under both high and low densities. Less stem reserves were mobilized in the tall (T)
than short (S) isolines in the stay-green pair, while the reverse was true for the senescent pair. This
suggests that stay-green should provide much-needed lodging resistance in tall sorghums under post-
flowering drought.

19. Working together with the
Queensland Government
Stay-green
Senescent
0
50
100
150
200
250
300
350
16 30 44 65 73 79 86 93 100 107 121 137
Cumulativewateruse(mm)
Days afteremergence
Low Density (2011)

20. Working together with the
Queensland Government
4. Germplasm with enhanced stay-green
and adaptation for Mali
The DEEDI sorghum breeding program has been
selecting for the stay-green trait in sorghum for more than
30 years while simultaneously selecting for grain yield.
In Australia, 3-dwarf hybrids with stay-green have been
shown to increase grain yield & grain size, and decrease
lodging in crops subjected to post flowering drought.
Choosing a stay-green donor
• Options: B35 or an elite line from the DEEDI program
• Chose R931945-2-2 as a well characterized line from
the DEEDI program with high SG and good yield potential
and other potentially useful traits (e.g. midge resistance).
Stay-greenSenescent
R2
= 0.218
0
1
2
3
4
5
6
7
8
2 3 4 5 6 7 8 9
Stay-green rating (1=SG 9=dead)
Yieldt/ha
Data from 20 breeding trials (934 hybrids)
conducted between 2005 and 2008

21. Working together with the
Queensland Government
Produce populations based on elite
material from Mali (Phase 1)
Prior to the commencement of the GCP project, DEEDI
had developed 33 F2 populations each based on a cross
between a diverse set of germplasm lines from Mali with
the elite DEEDI line (R931945-2-2).
In consultation with Malian breeders, four of the 33
populations were chosen for:
• backcrossing to the recurrent parent and
selection for four SG QTL;
• development of RIL populations; and
• production of F1 hybrids with CMS lines
contrasting in SG.

22. Working together with the
Queensland Government
Evaluation in Africa of material
produced in Phase I (Phase 2)
Seed from selected BC1F2 will be used by Malian
collaborators for final stage selection and evaluation.
QTL-enriched lines will be evaluated in about 12 trials per
year (6 countries x 2 sites per country) over 2 years.
Target countries will include Mali, Niger, Sudan,
Ethiopia, Kenya and Uganda.
Five plants from each of 15 backcross-derived
introgression lines have been selected (13 lines from
F2_R04021-2/PI609084 and 2 lines from F2_R04003-
2/PI585749).
Hence about 75 genotypes x 2 reps x 2 water regimes
(WW & WD) will be evaluated per site, depending on
irrigation capacity for WW treatment.
Evaluation will focus on yield, height, maturity, stay-
green and grain quality.

23. Working together with the
Queensland Government
Evaluation of African germplasm for
known stay-green QTL (Phase 2)
Four RIL populations containing between 170 and 400 lines (totaling 917
individuals) will be phenotyped at the F4 generation (2012). Lines were selected for
height and PPS similar to the recurrent parent. QTL mapping studies will be
conducted in Phase 2.
Seed of 6 F1 hybrids produced. Trials will be conducted to compare the yield and
stay-green of the F1 hybrids at 2 sites in Mali during 2012.

24. Working together with the
Queensland Government
Variables & constants are
initialised in model or read
from spreadsheet (can
also read APSIM met files
etc to prototype)
Variables & constants are
initialised in model or read
from spreadsheet (can
also read APSIM met files
etc to prototype)
Can compare runs &
monitor all variables using
graph tools (or via APSIM
tools, with VenLink)
Can compare runs &
monitor all variables using
graph tools (or via APSIM
tools, with VenLink)
At present, VenLink to
APSIM is only for complete
modules, but Dean
Holzworth is going to
enable Vensim to work as
a sub-module, e.g. crop
process
At present, VenLink to
APSIM is only for complete
modules, but Dean
Holzworth is going to
enable Vensim to work as
a sub-module, e.g. crop
process
Great tool for prototyping
ideas before programming
APSIM module
Great tool for prototyping
ideas before programming
APSIM module
Genomics Simulation Modelling
Data management
Phenotypic data
Integrated
information that is
better able to
address complex
quantitative traits
Environmental
characterisation
Enhanced genetic gain
Training on linking breeding,
molecular & physiological aspects

25. Working together with the
Queensland Government
Cost-effective HTP whole
genome profiling technology
Applied to QTL and association
mapping studies
Genomics: What do you need?
Good quality genetic linkage maps
Applied to structured and
unstructured germplasm sets
Knowledge of genomic regions controlling
key traits
Put this is context of existing knowledge of
other QTL and major effects genes in order
to develop more effective breeding
strategies

26. Working together with the
Queensland Government
Stg1

27. Working together with the
Queensland Government
Stg1 QTL region
67 markers across
376 genotypes (AYT)
With such dense marker data sets, we
can look at the haplotypes in selected
QTL regions. This example focuses on
Stg1 on SBI-03. We can identify which
haplotype class is derived from stay-
green lines vs those from senescent lines.
B35 derived
haplotype
(stay-green)
Principal component analysis

28. Working together with the
Queensland Government
5. Conclusions (Physiology)
1) Height did not counteract the benefits of stay-green. On
the contrary, under stress, the tall version of the stay-
green pair yielded significantly more than the short
version.
2) The yield advantage in the tall version of the stay-green
pair was due primarily to larger grain size.
3) Stem mass during grain filling increased in the stay-
green pair and decreased in the senescent pair,
highlighting the role of stay-green in lodging resistance,
particularly in the tall version of the stay-green pair.
4) Tall stay-green line extracted more water during grain
filling than the tall senescent line.

29. Working together with the
Queensland Government
Conclusions
(Population development)
1) About 75 QTL-enriched lines will be evaluated in 12
trials per year (6 countries x 2 sites per country) over 2
years.
2) Four RIL populations containing between 170 and 400
lines (totaling 917 individuals) will be phenotyped at the
F4 generation.
3) Yield and stay-green of six F1 hybrids will be compared
at 2 sites in Mali.

30. Working together with the
Queensland Government
Conclusions (Training)
1) Intensive training for two Malian
scientists in Australia on physiological
methods and breeding techniques
related to drought adaptation will occur
in Feb 2012 (Phase 1).
2) Workshop in Africa on „breeding for
drought‟ and „physiology of drought
adaptation (Phase 2).
3) Six African scientists will work with the
sorghum program in Australia for 3
months each (Phase 2).

31. Working together with the
Queensland Government

32. Working together with the
Queensland Government

33. Working together with the
Queensland Government
Canopy development
0.00
0.20
0.40
0.60
0.80
1.00
1.20
Tillers/plant
Tall Short Tall Short
R931945-2-2 (stay-green) R955343-1 (senescent)
2009
The stay-green pair reduced tillering compared with the senescent pair. There were no
significant differences in tillering between tall (T) and short (S) isolines of the stay-green
pair.

34. Working together with the
Queensland Government
Implications for the grains
industry

35. Working together with the
Queensland Government
Stay-green
Senescent
0
50
100
150
200
250
300
350
16 30 44 65 73 79 86 93 100 107 121 137
Cumulativewateruse(mm)
Days afteremergence
High Density (2011)
Stay-green
Senescent
0
50
100
150
200
250
300
350
16 30 44 65 73 79 86 93 100 107 121 137
Cumulativewateruse(mm)
Days afteremergence
Low Density (2011)

36. Working together with the
Queensland Government
Leaf senescence
0
10
20
30
40
50
60
SPAD(FL-2)
T S T S T S T S
R931945-2-2
(stay-green)
R955343-1
(senescent)
R931945-2-2
(stay-green)
R955343-1
(senescent)
High Low
2009
The stay-green pair retained greener leaves (higher SPAD) during late grain filling than
the senescent pair under both high and low density treatments. There were no significant
differences in leaf greenness between tall (T) and short (S) isolines of the stay-green pair.

37. Working together with the
Queensland Government
Stay-green
Senescent
0
50
100
150
200
250
300
350
16 30 44 65 73 79 86 93 100 107 121 137
Cumulativewateruse(mm)
Days afteremergence
High Density (2011)

38. Working together with the
Queensland Government
Grain size
0
5
10
15
20
25
30
Grainmass(mg)
T S T S T S T S
R931945-2-2
(stay-green)
R955343-1
(senescent)
R931945-2-2
(stay-green)
R955343-1
(senescent)
High Low
2009
Under stressed conditions, the stay-green pair exhibited larger grain size than the
senescent pair, particularly under low density. Grain size was generally larger in the tall (T)
compared with the short (S) isolines of the stay-green pair.