- International Livestock Research Institute, India
- University of Queensland, Australia
- Department of Employment, Economic Development and Innovation, Australia
- Directorate of Sorghum Research, India
The project was approved late July 2008 and activities have started as expected for the postrainy season 2008-09. We could not have a meeting with all project partners before the start of the season. However, we had a couple of strategic meeting with partners based in India (ICRISAT, NRCS, ILRI), in early September to ensure that planned activities for the Rabi season would proceed as planned in the project proposal. Essentially, we met to decide on locations and material to include in testing.
Field activities have been performed as planned. Details are described in the different project activities below. The kick-off meeting with all partners of the project has been held between 23-25 March 2009 at ICRISAT. This meeting was originally planned for January but was postponed because of the aftermath of the Bombay events. Minutes and program of the meeting have been circulated and are available upon request.
The project is well on track. Detailed analysis of data gathered so far and only recently collected needs to be further pursued. The use of the lysimetric system to assess pre-post anthesis water use, a key component of information that this project aims to produce, appears to fulfil really well the expectations and thorough analysis of data now needs to be done. Excerpts of these analyses and progresses made thus far are provided in what follows.
Activity 1.1. Complete the marker-assisted backcross development of single/multiple stay-green QTL (stgA, stgB, stg1, stg2, stg3, stg4) NILs in elite parents including sweet sorghum S35.
These materials were advanced prior to the 2008-09 trial and seeds were multiplied and dispatched to NRCS for multilocation trials reported below (Activity 1.3)
Activity 1.2. Test single/multiple stay-green QTL for stay-green expression in the field and grain/stover yield/quality at Patancheru under terminal drought conditions.
The single and multiple stay-green QTL introgression lines in two different genetic backgrounds (R16 and S35) were evaluated in two moisture regimes (irrigated and stress) during postrainy season of year 2008-2009 at two sites in India, Patancheru and Tandur . The stay-green QTL isolines in R16 genetic background were evaluated in a replicated 42-entry (including parents and checks) trial in an alpha-design with 4-row plots (of 6m length). The stay-green QTL isolines in S35 genetic background were accommodated in a replicated 56-entry trial with a field design similar to R16 genetic background trial. The trials were sown during second week of October 2008. The observations on Green Leaf Area (starting from 2 weeks before flowering till maturity) and other related traits on agronomic performance of these QTL isolines were recorded. The whole plant straw samples were harvested on plot basis, and grinding of these samples will be completed by second week of June 2009, for stover fodder quality traits. On the basis of results from this trial, best single QTL isolines in both genetic backgrounds will be identified for development of double QTL introgression lines.
Activity 1.3. Test selected single/multiple QTL isolines in multi-location trials of India
These test were performed at several locations by NRCS. During Rabi season (October-April 2009), two sets of material, 42 lines of R 16 based and 56 lines of S35 based genetic background stay green QTL introgression lines were evaluated for postflowering drought tolerance. The trials were conducted at three locations, Tandur (Andhra Pradesh), Rahuri (Maharashtra) and Bijapur (Karnataka) in 6x7 (R16 based genetic material) and 7x8 (S35 based genetic material) alpha designs with three replications. The sowing was done on Oct 3, Oct 17 and Oct 27, 2008 at Tandur, Bijapur and Rahuri, respectively. Two water regimes were imposed by withholding the irrigation during the post flowering period. Unfortunately, the trials conducted at Rahiri and Bijapur were abandoned before flowering due to severe shoot fly attack due to late sowing. Excess rains during first weak of October resulted in the flooding of fields, which did not allow us to prepare the field for sowing in time.
At Tandur, trial on the evaluation of S35 based genetic material was successfully conducted and the required data set were recorded. Following observations were recorded.
Days to 50% percent flowering and grain physiological maturity,
Plant height at Flowering and physiological maturity
Green leaf area estimation of top six leaves from flowering onward
Panicle number and dry weight,
Stover and grain dry weight
Stover Quality parameters- stover samples for quality analysis was handed over to ILRI.
As data analysis is yet to be completed, the results from this trial will be reported in the next report.
Activity 2.1. Model analysis of increased WUE (pre/post anthesis water use) stay-green under different climatic scenarios for both Australia and post-rainy season India
Soil depth measurements have been taken to feed into APSIM. The work on APSIM will fully start once the post doctoral fellow is recruited, which should take place towards the beginning of Year 2 of the project.
Activity 2.2. Assess variation for WUE/WU in the GCP-genotyped sorghum reference collection and in single/multiple stay-green QTL isolines from 1.1.
This work was carried out at ICRISAT
Evaluation of the full reference collection of sorghum for transpiration efficiency under well-watered and water-stressed conditions, and for the rate of water loss per unit of leaf area
We assessed all the accessions (375 + 3 controls) of the reference collection for transpiration efficiency (TE). The trial was planted on October 27th and the TE experiment was carried out between 8 and 27th December 2008. Plants were grown in large pots (11” diameters, containing 11 kg of Alfisol). Four seeds were planted in each pot, later on thinned to one seedling per pot at 2 weeks after sowing. Twelve plants (pots) were grown for each entry. Plants were grown under fully irrigated conditions until 6 weeks after sowing. At that stage, 4 replicated plants per genotype were harvested to evaluate plant biomass before imposing the treatment in the other 8 plants, which were used for the stress (water stress, WS) and control treatment (well-watered, WW), that is 4 replicated plants for each treatment. For that, the remaining plants were saturated with water. Since there were 378 entries in the trial, only one replication within each treatment could be handled per day and therefore the treatment imposition was staggered over 4 days. On each day, one replication within each treatment (WS and WW) were saturated with water, bagged with plastic bag around the stem the following morning, and weighed immediately after. Bagging prevented soil evaporation. Same procedure was followed for the 3 other replications. The procedure for each of the treatment was as follows:
1. Drought stress set: WS plants were exposed to progressive water stress by letting plant loose no more than 150 g per day during the first 4 days after imposing stress, no more than 100 g per day in the following 4 days, and no more than 75 g per days during the rest of the experiment. Since pots were weighed every 4 days only, this corresponded to a maximum water loss of 600 g over 4 days, 400 g in the subsequent 4 days, and then 300 g for any 4 days interval in the rest of the experiment.
2. Well-watered set: WW plants were maintained well-watered by re-adjusting pot weight close to field capacity on those days when the pots were weighed (every 4 days), and by adding water two days after weighing to bring back pots close to field capacity, based on transpiration data (from previous weighing intervals).
All plants were harvested when the transpiration of the drought set fell below 10% of the transpiration of the well-watered set. Harvest and transpiration data have been performed but data entry has still to be completed and should be reported during our next meeting.
Evaluation of a portion of the reference set (with relatively similar flowering time during Rabi) and selected stay-green QTL introgression lines for water uptake under stressed conditions in lysimeters (2.0-m long and 25-cm diameter tubes), and for the proportion of water used prior/after anthesis
A set of 210 entries, including 152 accessions from the reference collection + 58 marker-aided stay-green QTL introgressed lines in the background of S35 and R16 were assessed in large and long PVC cylinders (2.0 m long and 25 cm diameter), mimicking roughly the soil volume that sorghum plants would have at usual field planting densities. The trial has been sown on 20th Oct 08. Re-saturation of the soil profile was performed between Nov 13 and 15th, by adding 3 liters of water to each cylinder to ensure that all lysimeters were saturated at the time of imposing the stress. Immediately, after saturation, polyethylene bead were applied to the soil surface to prevent soil evaporation, at a rate of 600 mL, giving a layer of approximately 1.2 cm, sufficient to limit soil evaporation by about 80%. First weighing of the cylinders was taken on 19-20-21-22 November 08. In each case, one full pit was taken, corresponding to one and half replications in each of the treatment (these are 630 cylinders, which is the maximum that can be weighed accurately in a single day). Subsequently, weighing was done following the same sequence, so that the interval between weighing was the same for all treatments and replications. Two treatments were applied: a water stress treatment and a fully irrigated control. Procedure for each of these treatment was as follows:
1. Water stress treatment: No water was added from the time of saturation, i.e. 15th Nov 08, except 1000 mL applied during grain filling on each of 02nd and 09th Jan 09. A total of five weighing were taken for the water stress treatment, i.e. 21-22 Nov, 5-6 Dec, 19-20 Dec, 7-8 Jan, 19-20. Plants were physiologically mature at the time of the last weighing and thus were harvested soon after weighing. Plant parts were separated in leaf, stem, panicle and grain weights.
The data showed a range of variation in water extraction between 10193 g plant-1 to 15260 g plant-1, i.e. about 50% variation in water extraction between the highest and lowest water “extracters”.
One of the hypothesis of the project is that staygreen could be a consequence of differences in the pattern of water use, with staygreen material likely to be those using a relatively lower proportion of water during the pre-anthesis period, to favor more water availability for grain filling. We found a clear negative relationship across the 210 entries, between the amount of water taken before anthesis and that taken after anthesis, showing that genotypes taking up more water before anthesis took up less water after anthesis. How this relates to grain yield still needs to be analyzed.
The PVC experiment gave also the opportunity to assess TE across a fairly large range of genotypes, over a long period of time (from 4 weeks after sowing until maturity). TE value varied between 2.44 g kg-1 water transpired to 6.09 g kg-1 water transpired, and so showed a very large range of variation. A detailed analysis is need to investigate how differences in TE relate to the presence or absence of certain staygreen QTL. In short, in S35 background, few staygreen QTL introgression lines had TE above S35. Introgression lines with Stg 3, Stg4 or Stg B had all fairly high TE equal or superior to S35. In R16 background, most Stg B, and Stg3 and few Stg4 introgression lines had high TE, equal or superior to R16. In any case, there were entries from the sorghum reference collection with TE higher that the highest of the staygreen trial entries.
2. Well-watered treatment: Six cylinder weighing have been done on: 19-20 Nov 08, 3-4 Dec 08, 17-18 Dec 08, 5-6 Jan 09, 21-22 Jan 09, and 2-3 February. Plants were harvested soon after the last weighing. For genotypes maturing earlier, the head had been harvested before to avoid grain losses. However, all the stover biomass was harvested at same time. Analysis of the water extracted has been done until after the final weighing of the samples.
Total water use under well-watered conditions varied between about 10.5 kg to about 42.3 kg of water per plants from 4 weeks after sowing until maturity. The highest and lowest water users under well-watered conditions are reported below. Biomass and grain data have been performed and reveal a large range of variation. As for the DS treatment, the trial was an opportunity to assess TE over a long period of time (from 4 weeks after sowing until maturity) in a large set of entries. There also, the very interesting thing is that data revealed also a good range of variation in TE. Thorough data analysis is still needed and is planned for the next meeting in Australia during the National Conference on Sorghum (7-8 August 2009).
Activity 2.3. Conduct an ex-ante analysis of the economic benefits (increase in grain/fodder quantity/quality) of incorporating stay-green and/or improved pre/post anthesis water uses/WUE traits in existing sorghum varieties in Australia and post rainy season India
Monthly sorghum stover sampling from six sorghum stover traders in Hyderabad was taken up November 2008 to provide new baseline data for stover quality - price relationships. From November 08 to March 09, the average price per kg dry stover was 12.7 cents. For comparison from November 2004 to March 2005 the average stover price was 6.7 cents.
The project is fully underway. For the second postrainy season of the project, the trials in ICRISAT and at the partner locations were carried out almost rain-free, which allowed an excellent expression of the stay-green phenotype and an excellent discrimination of the stay-green introgression lines. From the six different QTL that have been introgressed from donor parent B35, it is fairly evident that StgA, StgB and Stg3 all contribute to a clear expression of the stay-green phenotype and some of the lines contribute to enhanced seed yield and biomass under terminal drought.
Interestingly, the 2009-10 season confirms the value of several StgA, StgB and Stg3 introgression lines across locations on trials at the partner’s stations. Due to this excellent proof of concept of the value of several stay-green QTL, discussions during the most recent planning meeting centered on: (i) developing a di-allele scheme to combine beneficial QTLs; and (ii) choosing six popular varieties from Rabi sorghum (North Karnataka, West Andhra Pradesh, Maharashtra) to initiate the introgression of beneficial QTLs into these varieties.
The lysimetric facility used to evaluate a number of traits related to plant water use and water extraction has revealed a wealth of variation in the germplasm for transpiration efficiency (over 50% range of variation) and a range of variation for water extraction from 12 to 15 litres per plant. This provides new sources of variation for critical traits that can be fed into breeding programs. Interestingly, certain stay-green QTL enhance either transpiration efficiency or water extraction, although this depends on the genetic background where they are introgressed. For instance, Stg1 and Stg2 enhance water uptake in an S35 background (which has limited capacity to extract water), while there is no QTL that improves water extraction in an R16 background, which is already endowed with high water extraction capacity. Therefore, stay-green QTL can be dissected for the different traits they control at the gene level. However, a prior knowledge of these traits in target cultivars for introgression is critical.
The postdoctoral fellow has been recruited and is at present undertaking training in APSIM in Queensland, Australia. Initial work will consist of mapping yields across Rabi sorghum states in India and extrapolating soil conditions prevailing at each test site. This mapping will then be used to test the effects of specific QTL on yield across a range of locations.
The project is still in full swing. For the third post-rainy season of the project, trials have been repeated at ICRISAT and at three partner locations during 2010-11, using a subset of the most promising entries tested in the 2009-10 season, i.e. 18 entries in S35 and 18 entries in R16 backgrounds. Harvest of the trials has been completed and fodder quality assessment is in progress. These same entries have also been tested in the lysimetric system to evaluate precisely the pattern of water extraction and its relation to yield and leaf development parameters. Some of these entries have also been tested in pot studies in a controlled environment facility to evaluate traits related to the plant’s control of water use under well-watered and water stressed conditions. The introgression of stay-green QTLs into six popular varieties from the Rabi sorghum main track (North Karnataka, western Andhra Pradesh, Maharashtra) is in progress. Characterization of the Rabi sorghum environment has been done to elicit the type of stress that is most common in order to better target the choice of traits needed for adaptation. This has involved the collection of climatic data from Rabi sorghum districts and productivity data from these same districts, as well a thorough APSIM simulation. In a nutshell, Rabi sorghum primarily undergoes flowering and post-flowering terminal drought, which is fully in line with the project’s objectives. Efforts have been made and are still on-going in Year 3 to deliver publications, with one paper accepted in Functional Plant Biology, and one paper currently under review (Crop and Pasture Science). A meeting has also recently been held to discuss other potential publications in view of the value of the data collected since the beginning of the project. We have agreed to have one manuscript prepared to report the result of the 44 introgression line entries (+ checks) in the S35 background and one from the 31 introgression line entries (+ checks) in the R16 background, across the DSR and ICRISAT locations using data for the period 2009-10. This manuscript will be drafted by Dr HS Talwar, Principal Scientist at DSR. A second manuscript will report the detailed yield trials carried out at ICRISAT and will relate expression of the stay-green phenotype and its relationship with yield, using the 2009-10 data and possibly the 2008-09 data for the S35 background. This manuscript will be drafted by Dr Santosh Deshpande, and will report, in particular, the clear expression of the stay-green phenotype in StgA, StgB and Stg3 introgression lines. A third manuscript, headed by Dr Michael Blummel, will report on the assessment of fodder quality data in different locations and under different water regimes at each location. An ex-ante assessment of the value of introgression of stay-green QTLs has been executed and a fourth manuscript is coming up. A meeting was held in Queensland in April 2011 to update project partners on progress. We have also discussed the possibility of a Phase 2 for this project, capitalizing on the great progress achieved in Phase 1. An outline of the ideas that have been exchanged will be sent to the ACIAR program manager as the basis for an initial dialogue.
The project is now completed. A no-cost extension has been granted in order to carry out a workshop to disseminate the findings of the project. This workshop was initially planned for the last semester of Year 4. However, while the project has, on the whole, achieved excellent outputs, the experimental data from Year 4 could not be fully analysed by the time of the workshop. Therefore, it was proposed to have the workshop just after the end of the project. The purpose of the workshop will now go beyond the scope of the current project and will be open to a much larger community of sorghum scientists, especially with a view to linking project activities to the CGIAR Research Program on Dryland Cereals (CRP 3.6), and also as a basis for drafting a proposal for a potential second phase to this project.
In Year 4, postrainy trials were repeated to assess the most promising introgression lines from Year 1 and 2, i.e., 12 entries in S35 and 13 entries in R16 background, which were also tested in Year 3, plus parental lines and checks. These were tested at ICRISAT and at three partner locations in 2011-12. Again, very promising introgression lines were identified, and some confirmed the previous year’s assessment, both for grain yield and for stover yield. The capacity to positively alter both grain yield and fodder is very important as both components now share almost an equal importance in the sorghum value chain. Harvest of the trials has been completed and fodder quality assessment is underway. The testing of these same entries has also been repeated in the lysimetric system to evaluate precisely the pattern of water extraction and its relation with yield, as well as leaf development parameters. Several dry-down experiments in the glasshouse to assess specific water saving traits have also been performed in some of these lines under well-watered and water-stressed conditions. A few traits appear to clearly distinguish staygreen introgression material, e.g., the capacity to develop smaller leaf area and the capacity to restrict transpiration under high VPD conditions. The introgression of stay-green QTLs into six popular varieties from the postrainy (Rabi) season sorghum main areas (North Karnataka, West Andhra Pradesh and Maharashtra) is now at the BC1F1 stage. We have also fingerprinted the staygreen introgression lines, in R16 and S35 backgrounds, with Diversity Array Technology (DArT) markers. This had helped to saturate the staygreen QTL regions, and also helped to assess the recurrent parent recovery. In all, we have data on 669 DArT clones with known physical position on the sorghum physical map for all staygreen introgression lines. A publication is under review in Field Crops Research related to the characterization of the postrainy season sorghum environments that segregates these environments into sub-areas characterized by different stress types. The results clearly shows that the postrainy season sorghum area (about 5 M ha) should not be taken as a single block, but as sub-blocks characterized by different stress patterns, therefore requiring specific breeding for precise drought-adaptive traits. Using APSIM and based on the finding of traits discriminating staygreen introgression lines from recurrent parents, crops simulation modelling efforts have been made to test the effect of several water saving traits (those identified from physiological studies), one of them having an important effect, especially under harsh conditions. The publications that were mentioned in the previous report are progressing well. A review article will be forthcoming in a CABI book in 2013. Furthermore, there is a plan for additional papers on the lysimetric assessments and the trait-based assessments of introgression lines and germplasm, to capitalize on the most important physiological traits that are affected by staygreen introgression. This should give the basis for evolving from the use of staygreen QTL in breeding to the use of specific, well-characterized mechanisms. Our current target is to have these papers prepared for the workshop, which is set for 19-21 February 2013.
Activity 1.1. Complete the marker-assisted backcross development of single/multiple stay-green QTL (StgA, StgB, Stg1, Stg2, Stg3, Stg4) NILs in elite parents including sweet sorghum S35
This activity was completed at the end of Year1. The most promising of these are now being used in the diallele crossing program, whose purpose is to pyramid together several staygreen QTLs.
Activity 1.2. Test single/multiple staygreen QTL for staygreen expression in the field and grain/stover yield/quality at Patancheru under terminal drought conditions
A trial including a subset of the most promising entries from the 2009-10 season, i.e. 18 entries in S35 and 18 entries in R16 background, which had been tested in 2010-11, has been conducted. Harvest of the trials has been completed and fodder quality assessment is in progress. The stress level this year was mild (about 20% reduction in grain yield in both backgrounds), which was explained by cultivation in a fairly deep soil and a slightly later imposition of stress than in the previous year. However, in both backgrounds there were again several introgression lines with performance superior to the recurrent background (e.g. S35SG06014, a Stg3 introgression in S35 background, K359W, K369, K382, K456 in R16 background). We are now running a complete analysis of these data using the two years in which this same set was evaluated.
There was also substantial variation in the stover biomass in this repeated trial, although the reduction in stover biomass under terminal water stress was about 30% in the S35 background sub-trial (stover yield reduction higher than grain yield reduction), but only about 20% in the R16 background sub-trial (same reduction than for grain yield). In S35 background, introgression lines S35SG06026, S35SG06028, S35SG06038, S35SG06014, S35SG06001 had stover yield significantly higher than recurrent parent (30-60%). In R16 background, the same introgression lines having yield superior to R16 (K359W, K369, K382, K456) also had higher stover yield under terminal water stress (10-20%). Given the importance of stover in the sorghum value chain, these results indicated that gains can be made from both the improvement of grain and stover yield. The fact that several introgression lines in the two backgrounds are “double winners” is very encouraging and confirms results from the earlier years. The increase in stover biomass in the S35 background was also very substantial (over 30%).
To deepen the understanding of the putative mechanisms of stover/grain yield drought adaptation (undoubtedly impacting their quality), the leaf area (LA) developmental characteristics were assessed in 20 genotypes (10 each from R16 and S35 backgrounds; out of that in 6 genotypes detailed leaf observations allowing their model parameterization were gathered) under well-watered (WW) and water-stressed (WS) regimes in 2011-2012 field experiment. We also gathered further evidence of genotypic differences in the LA developmental patterns under WW and additional genotypic plasticity under WS. Under WW several genotypes developed smaller LA (6028, 6038, 6040 [S35 background]); K606 (R16 background), while others had larger LA (extreme case 6026 [S35 background]) compared to their respective senescent parents. We found the genotypes further differed in their reaction to WS in terms of LA development - some genotypes we found clearly restrict leaves growth upon drought exposure (K359 white [R16 background], 6026, 6008, 7001 [S35 background]) while others did not (both senescent parents - R16, S35). Very similar patterns were found in detailed LA study in lysimetric system and, moreover, there the total LA developed under WS related strongly (negatively) to plant’s grain yield, indicating clearly that, under these conditions of stress, developing a smaller leaf area was beneficial for yield under fairly severe terminal stress.
The available range of LA identified across the different introgression lines, leading to a wide range of plant water usage, therefore provides us an arsenal of plant material that have potentially the promise to be specifically adapted across a fairly broad spectrum of environmental conditions. This is intended to be further used for precision breeding purposes - i.e. matching particular developmental pattern with given agro-ecosystem (this will be tested with model; Activity 2.1.)
Stover quality data from the 2011-12 trial have been generated and data are being analysed. Current efforts are focusing on the publication of the data from the initial two years. In short, there appears to be no trade-off between fodder quality and quantity, in both genetic backgrounds, indicating that varietal improvement for fodder quality can be done along with grain and fodder productivity enhancement. This year’s data confirmed significant positive effects of several staygreen QTL on fodder quality traits, and also confirmed that these effects depended on the genetic background (same results were obtained for drought adaptation mechanisms). For instance, there is a significant 2-3 unit increase in in-vitro digestibility due to StgB and StgA introgression in S35 background, but almost no change in any of the staygreen introgression in R16 background. An opposite effect was found for stover nitrogen (another fodder quality trait), with no effect of any staygreen QTL in S35 background, but a significant effect of introgression in several QTL in the R16 background.
Activity 1.3. Test selected single/multiple QTL isolines in multi-location trials of India
The trials have been repeated at three partner locations (Tandur, Bijapur, and Hyderabad) using selected 12 entries in S35 and 13 entries in R16 backgrounds, plus parental lines and checks (18-entries trials in both cases). The trials were sown under stress and non-stress treatments at all three locations. In this season the trials under stress conditions were replicated six times, and three times under non-stress conditions. The trials were carried out almost rain-free at all three locations. All the required observations (phenology, plant height, green leaf area retention at different growth stages during post-flowering growth period and yield components) were recorded at all three locations.
The variations between genotypes and water-treatments were significant for all the yield components (grain yield, stover yield and total dry matter), but interactions between treatments (well-watered and water-stressed) and genotypes (QTLs) were non-significant at all the locations under both genetic backgrounds. On the other hand, variation in genotypes/QTLs differs significantly across locations and their interactions with locations (environments) were also significant under both stress and non-stress environment in both the genetic backgrounds. This implies that same QTL may not work in all locations. Our results confirmed the finding of previous year that GLAR (green leaf area retention at physiological maturity) and yield components were significantly higher for the R16 background. Our phenotypic data clearly indicate that stgB is the key QTL for yield components under water stress in the R16 genetic background and stgA, stgB and stg1 are the key QTLs for yield components under water stress in the S35 genetic background. It seems that the QTL stgB is common in both, hence less interaction with genetic backgrounds. Data clearly indicated about a 10% advantage in grain yield (GY) and total dry matter (TDM) accumulation by staygreen QTLs introgression in R16 background under both WW and WS conditions. On the other hand, about an 8% advantage was recorded in GY and TDM by staygreen QTLs introgression in the S35 background under WS, but no advantage under WW conditions.
Activity 1.4. Saturate key QTL regions on sorghum linkage groups with easily used markers
To saturate the target staygreen QTL regions, all of the ILs evaluated in first two years of the project were fingerprinted using DArT assays. Thirty-eight staygreen ILs in R16-background, and 43 ILs in S35 background, along with parents, were genotyped with standard DArT assay available at Diversity Array Technology Pvt. Ltd. in Australia. A total of 721 DArT clones were scored for staygreen ILs in both R16 and S35 background, and 669 DArT clones for which physical map positions on the sorghum genome is used for further interpretations. For staygreen ILs in R16 background, the range for monomorphic DArT clones was 37-48%, with a maximum of 6% missing data. The range for B35-allele introgression in R16 background ILs is between 0 to 17%; with recurrent parent recovery (for R16 alleles) in the range of 35 to 52%. Similarly, for S35 background ILs, the range of monomorphic DArT clones was 36 to 39%, with maximum of 6% missing data. The B35-allele introgression ranged from 2 to 19%, with recurrent parent recovery for S35 alleles in the range of 42 to 58%. Currently, we are in the process of developing haplotype/allelic distribution map for all of the staygreen ILs, and this DArT data along with polymorphic SSRs will be further utilized to analyse the effects of different introgressions on staygreen expressions.
To further characterize the QTL introgression regions related to different staygreen QTLs at the sequence level, we have submitted all of the ILs genotyped with DArTs to the Institute of Genomic Diversity (IGD) for Genotyping-By-Sequencing (GBS) analysis under a NSF-BREAD project with Ed Buckler’s Laboratory at Cornell University. The data from this GBS analysis will identify SNPs and PAVs (present-absent variations). This data will further help us identify SNPs and candidate genes in the staygreen QTL regions. The SNP data along with field and lysimetry characterization can further be utilized for genomic selection for improved drought tolerance in postrainy season sorghum.
Activity 1.5. Initiate the introgression of key QTLs onto the most widely grown postrainy season season sorghum varieties
The most promising staygreen QTL ILs (both in R16 and S35 background) shortlisted based on both field and lysimetry evaluations were crossed in half-diallel crossing scheme to generate double QTL introgressions. For R16 background, 7 ILs (K359-white, K375, K260, K271, K382, K456 and K369) and 8 ILs in S35 background (S35SG06003, S35SG06028, S35SG07003, S35SG0632, S35SG06014, S35SG06005, S35SG06010 and S35SG06011) were crossed during the postrainy season 2011 and rainy season 2012 to generate F1 seed in diallel scheme to produce 21 and 28 crosses in R16 and S35 backgrounds, respectively. These F1s are being advanced in current rainy season 2012 to generate F2 seed for all combinations. These F2 seed along with their respective staygreen IL parents along with F1s will be field evaluated during the 2012-13 postrainy season to understand gene interactions and genetics of staygreen expression as a function of double QTL introgressions. Resources permitting, we will also try to use some of the remnant seed from selected F2 populations to identify double QTL homozygotes.
Staygreen QTL introgressions in six popular postrainy season sorghum varieties (CSV22, Parbhani Moti, M35-1, Akola Kranti, Parbhani Jyothi, and DSV5) has been initiated. The most successful introgressions in both R16 (K359-stgB, K260-stg3, K271-stg4) and S35 background (S35SG06008-stgA and S35SG06040-stg1) were used as donors for introgression of all the six validated staygreen QTLs. These QTLs are being introgressed into six popular postrainy season varieties, e.g., M 35-1, Parbhani Moti=SPV1411, Phule Vasudha, and CRS1 by marker-assisted backcrossing using 3 to 5 SSRs flanking these QTLs. Currently, the crossing program is at the BC1F1 stage, and with input from the Directorate of Sorghum Research (DSR), ICAR, Hyderabad we will add three more genotypes (Phule Yashoda, Parbhani Moti and DSV5) into the backcrossing scheme during the 2012 postrainy season. Currently, we are assessing the polymorphism of additional SSRs in staygreen QTL regions which will help for faster recovery of recurrent parents by BC2 or BC3 generation.
Activity 2.1. Model analysis of increased WUE (pre/post anthesis water use) staygreen under different climatic scenarios for both Australia and postrainy season India
Last year, we reported that the postrainy season sorghum area was, in fact, compounded of several sub-regions characterized by different patterns of drought, from very severe to mild. The paper reporting this analysis is under review in Field Crops Research. This year we have initiated the simulation of trait effects on yield in these different sub-regions. The knowledge of drought stress environments across main postrainy season sorghum tract (Field Crops Research, under review) along with improved understanding of drought tolerance mechanisms (Activity 1.2.) was used to virtually introgress these mechanisms into common postrainy season cultivar (M35-1) and model its effect across these environments. Traits of our main interest were modelled, and out of them two showed promise of improving yield under postrainy conditions in the target zone. These were analysed in detail, i.e., with dynamics of LA development and tissues hydraulic conductivity.
To model the effect of differences in LA development on yield, the APSIM coefficients of LA parameters for M35-1 (TPLA production coefficient and TPLA inflection ratio) were altered. The modification in these parameters used the experimental data gathered in some introgression lines and were then within the magnitude of differences in LA developmental dynamics described between R16 and one of the most successful stg introgression line; K359white (stgB). The APSIM outputs indicated the introgression of mechanism altering the dynamics of leaves expansion rates would improve the grain production in the terminal drought environments in which stress occurs before flowering while there would be production trade-off in the environments with later drought stress onset or stress-free environments.
The evidence of altered tissues’ hydraulic conductivity within sorghum germplasm was gathered (similar to that observed in pearl millet [Kholova et al. 2010]). Further investigations showed that this mechanism appears to be linked to lower root hydraulic conductivity which probably limits leaves’ hydraulic conductivity especially in conditions of high transpiration demand (high VPD; noon hours). This mechanism was also altered in M35-1 APSIM model in two ways - through virtual variation in kl constant (affecting root hydraulics) and maximizing the rates of transpiration (generation of circadian VPD trajectory with imposed limit on transpiration rates when crossing the set VPD threshold). Virtual manipulation of kl benefited the grain production in the most severe pre-flowering terminal drought environments while limited transpiration rates benefitted grain yields across all environments; simulated combination of both these traits, also, resulted in grain yield benefits across all environments.
Detailed dissection of virtual crop growth revealed that yield improvement from the different investigated mechanism came from water saving (few mm) during vegetative growth, which allows its utilization during later crop growth and ultimately result in an extended period of grain-filling. Generated outputs are being worked out into the form of a publication.
Activity 2.2. Assess variation for WUE/WU in the GCP-genotyped sorghum reference collection and in single/multiple stay-green QTL isolines from 1.1
The paper reporting the lysimetric assessment of staygreen introgression lines is now accepted in Functional Plant Biology.
Building on the earlier evaluation of sorghum germplasm, a number of germplasm entries varying for water extraction in previous lysimetric trials have been re-tested (20 entries) and most of them confirmed their higher water extraction capacity. Initial work, inconclusive so far, has been attempted to assess possible root architecture differences in these lines. Another 20 germplasm entries from previous screening, contrasting for TE, have been re-tested in the lysimetric system and the large genotypic contrast for TE has been confirmed. Then a standard dry-down experiment to assess traits related to the control of plant water use has been completed with these lines, which shows that all but one of the germplasm entries with high TE shows an earlier and sharper decline in the rate of transpiration under outdoor conditions and high VPD than the germplasm entries having low TE. Additional work is underway (although not planned earlier) to confirm this important finding (repeat assessment of the VPD response, analysis of plant hydraulics).
Additional experiments have been repeated in the lysimetric system to assess water use efficiency and water extraction under terminal drought in the following set of entries. Among the parents of mapping populations, ICSV745 and PB15881-3 showed good contrast for water use efficiency (higher in ICSV745 and lower by about 2 units in PB15881-3), which opens the possibility to map QTL for TE. The knowledge obtained above on the germplasm contrasting for TE also suggest that these two parental lines contrast for the VPD response mechanisms, which would allow a more focused phenotyping. These two lines also contrasted for the total water extracted (lower in ICSV745 and higher by about 2 units in PB15881-3). The trial, testing popular varieties/landraces adapted to the postrainy season sorghum conditions, showed that overall, the TE level of these lines was relatively low (around 5 g biomass per kg water transpired), except for Parbhani Jyoti which has a slightly higher level. In comparison, many introgression lines in S35 background have TE above 6 g biomass per kg water transpired. In contrast, it seems that the water extraction capacity of postrainy season adapted landraces and varieties is higher than that in S35 background. Therefore, the focus of future crop improvement efforts in the postrainy season materials probably needs to be on water use efficiency.
Fifteen introgression lines in the background of S35 have also been tested under different water treatments in the lysimeters (no water from 3.5 weeks after sowing), normal (last irrigation at 5 weeks after sowing) and mild water stress (last irrigation at 5 weeks after sowing plus 2L of water [equivalent to 40 mm rain] at 10 days after flowering). We have confirmed the superiority of three StgB introgression lines in the trial (S35SG07001, S35SG07002, S35SG06011b) but also one each of Stg1, Stg2, and Stg3. It appears that the yield decrease, especially in recurrent background S35 was explained by decreased grain filling.
Activity 2.3. Conduct an ex-ante analysis of the economic benefits (increase in grain/fodder quantity/quality) of incorporating staygreen and/or improved pre/post anthesis water uses/WUE traits in existing sorghum varieties in Australia and postrainy season India
The ex-ante analysis has been completed toward the end of Year 3. A manuscript is currently being prepared.
This is the summary of the final report for phase 1 of this project, which will be extended for 4 years (2013-2017).
Why the work was done? Post-rainy sorghum is an important commodity for about 5 M households in South India and both grain and stover residues now play an almost equally important role in this value chain, the price of stover being about two-thirds of the price of grain. Moreover stover price is linked to its quality. The main constraint to post-rainy sorghum production is water. The project aimed at genetic solutions to this limitation, looking at the whole value chain of sorghum, to improve both grain/stover productivity and quality.
What was achieved? The results of the project showed that the introgression of stay-green QTLs can improve grain yield, stover yield, and stover quality traits, without having trade-offs between these traits. It showed some exciting synergistic associations between stover productivity and stover quality. Based on this, backcross introgression lines have been initiated in the background of six different Rabi-adapted and farmer-preferred sorghum lines towards the development of varieties containing drought adaptive traits. Prediction of trait effects on yield has been proved possible, which would therefore greatly help to refine the choice of most promising breeding targets.
What impacts has the project had or is it likely to have in the future? The first impact will come from the absence of trade-offs between stover quality attributes and grain/stover productivity, and that breeding for both aspects can be done together. The second impact comes from the proof of concept that stay-green QTL can increase quality and productivity, which gives credit to a biotechnological approach for a complex constraint (drought), an area where successful examples are rare. Third, we have identified germplasm variants for critical traits which will be used in the breeding program. Last, but not least, the predictions on the effects of traits will probably affect how breeding decisions are taken.
What future actions might be required? Recent re-sequencing of introgression lines from the first phase, plus additional knowledge from fine-mapping work on different stay-green QTLs, would provide a fairly precise, if not gene-based, marker for very precise selection of the on-going introgressions. From the first phase, a much better knowledge of the mechanisms underlying the expression of a stay-green phenotype has been gained, namely: (i) improved water extraction; (ii) improved transpiration efficiency (TE), seemingly due to the capacity of high TE genotypes to restrict water losses under high VPD conditions; and (iii) reduced leaf area. However, a key finding is that the effect of stay-green introgression is highly dependent on genetic background, so that the development of stay-green varieties still requires a step of testing to secure the right phenotype. That step anticipates an evolution from using “stay-green QTL” to using QTLs involved in the key underlying mechanisms, and/or combinations of these to optimize production. A very exciting development in the first phase has been the identification of large genetic variation in the germplasm for these underlying traits, beyond levels currently available in stay-green QTL donor parents. In addition, the analysis of weather patterns in the Rabi sorghum area shows a clear clustering of sub-zones with different stress severity patterns. Simulation of the effect of some of the underlying traits predicts both increased yield and increased yield resilience. This calls for specific breeding targets for these different zones, where specific combinations of traits and management practices are needed. The second phase would capitalize on this by using identified germplasm variants for key traits to harness the genetics of these traits and develop promising pre-breeding materials. Backcross nested association mapping (BCNAM) populations in the background of a few Rabi-adapted varieties with varying germplasm appears to be a well fitted, long term approach to sorghum breeding. The use of crop simulation modeling will be expanded to other regions facing similar constraints, to guide the choice of key traits and/or management options. This more strategic part would be aligned to the sorghum-related Product Lines of the CGIAR Research Program (CRP) on Dryland Cereals, and aims at a global integration of sorghum breeding efforts targeted to water-limited environments, ensuring a better coordination of efforts across regions.
This project in an extension to a previous 4-year phase where the proof-of-concept was established of the value of introgressing staygreen quantitative trait loci (QTL) for improving both the grain and the stover productivity of sorghum during the post-rainy season in India. An important aspect of that work was that several QTLs had a beneficial effect on some stover nutritional quality (for cattle feed), without having trade-offs with grain and stover productivity. This showed the potential to improve different components of the sorghum value chain at the same time. However, the staygreen QTL did not have the same effects in all genetic background where they were introgressed, implying the need to check intermediate introgression products.
The objectives of the second phase are:
1. Develop an application phase in which several staygreen QTL will be introgressed into the most popular cultivars currently used by farmers in the post-rainy sorghum track of India.
2. Expand the work undertaken in the first phase to the West African region with regards to the environmental characterization of stress scenarios, and for the development of new breeding populations - Backcross Nested Association Mapping (BCNAM) populations - involving the most popular cultivars of the postrainy sorghum area of India as recipients and new donors of beneficial traits identified in the first phase.
The second objective is supported by the CGIAR Research Program on Dryland Cereals (CRP DC)
In objective 1, much of the work has so far focused on refining the QTL interval of critical staygreen QTL. It was initially planned to introgress 6 different staygreen QTLs (Stg1, Stg2, Stg3A, Stg3B, Stg4, StgA (=StgC)), into 6 different genetic backgrounds (M 35-1, Parbhani Moti=SPV1411, Phule Vasudha, CRS1, CRS4, RSLG262). At a meeting held in March 2014 it was decided to concentrate the introgression efforts only to Stg3A and Stg3B QTL (on sorghum chromosome 2, SBI-02), i.e. those QTLs showing the most promising effects in the first phase project. This would allow development of more introgression lines, thereby facilitating background selection. Work has also been carried out to refine the QTL interval for these two critical QTLs, using sequence data from the first phase. Most of the introgression work in the first phase had been accomplished using SSRs and had a large QTL confidence interval (indicating large genomic region) for most of target QTLs. These regions were further saturated by genotyping the introgression lines (ILs) in R16- and S35-genetic backgrounds with additional SSRs and DArT markers. Additionally these ILs were genotyped using genotyping-by-sequencing (GBS) approach and the target regions were further saturated with SNPs. Utilizing the extensive phenotyping and characterization data generated on the IL sets, we refined the QTL intervals for all of the QTLs except stgC (on SBI-01), which still spans across the centromere). For Stg3a and Stg3b, the interval is refined to 5.6Mbp and 2.2Mbp (genomic regions spanning 56.0 Mbp to 61.6Mpbp and 69.6Mbp to 71.8Mbp), respectively.
Under objective 2, we have started the parameterization of several cultivars adapted to the West and Central Africa (WCA) region, using existing evaluation data. These will be used as reference cultivar for the characterization of stress scenario in the region. The next step was to assemble weather data from a north-south transect in Mali where we expect the weather scenario to vary considerably. We also generated weather data from these locations where observed weather data are available, using Marksim weather generator. Then we compared the APSIM outputs from observed data to APSIM outputs from Marksim-generated weather data. We also initiated the development of new genetic populations - BCNAM populations, using 20 donors identified in the first phase and using the same recipient parents as those used in the introgression of staygreen QTL. During post-rainy season of 2013-14, at least 2 plant - plant crosses were executed between recurrent parents (Parbhani Moti = SPV1411 for Indian post-rainy season adaptation and CSM63E for Southern-Sahelian Zone adaptation) and 20 common set of donor accessions selected on basis of geographic-racial diversity. These were further backcrossed with male-sterile version of recurrent parent (currently at BC1F1-stage for CSM63E and F1-stage for Parbhani Moti=SPV1411). The long term purpose of these populations is two-fold: (i) because they use donors that “contains” traits shown to be the most promising to improve grain and stover productivity in the post-rainy sorghum track, we anticipate to derive improved breeding lines; (ii) we intend to map QTL for these traits, shown to be among the underlying mechanisms of the staygreen trait. Both of these activities in the second objective are also the object of the training of a PhD student from Mali, with the WCA part of the work supported by the USAID project aligned with Dryland Cereals CRP.
An inception meeting was held in March 2014, where representatives of other projects dealing with sorghum improvement under drought were invited. Discussions were held on collaborations and synergies between projects ranging across South-Asia, Eastern and Southern Africa, and West and Central Africa regions.
In the previous phase of this project the proof-of-concept was established of the value of introgressing staygreen (stg) quantitative trait loci (QTL) for improving both the grain and the stover productivity of sorghum during the post-rainy season in India. We established that several QTLs had a beneficial effect on some stover nutritional quality (for cattle feed), without having trade-offs with grain and stover productivity. This showed the potential to improve different components of the sorghum value chain at the same time. However, the staygreen QTLs did not have the same effects in all genetic background where they were introgressed, implying the need to check intermediate introgression products.
The objectives of the second phase are:
1. Apply the findings of the first phase by introgressing two linked staygreen QTLs into the most popular cultivars currently used by farmers in the post-rainy sorghum track of India. This objective is supported by ACIAR.
2. Expand the work undertaken in the first phase in India to the West African region with regards to the environmental characterization of stress scenarios, and for the development of new breeding populations - Backcross Nested Association Mapping (BCNAM) populations. These populations will involve the most popular cultivars of the postrainy sorghum area of India as recipients and new donors of beneficial traits identified in the first phase. This second objective is supported by the CGIAR Research Program on Dryland Cereals (CRP DC).
In objective 1, we have refined the genetic interval for QTLs stg3A and stg3B: five SNPs for stg3A and 21 SNPs for stg3B are now shortlisted for developing monoplexes KASPar SNP markers assays and these markers will be used for the introgression of these 2 critical QTLs in the different Rabi-adapted backgrounds. The introgression work is in good progress and some introgressions have reached BC2F1 stage. Under this objectives field trials have also been carried out to test the effect of staygreen introgression, nitrogen treatment and plant density on grain and stover yield and on grain and stover quality, the purpose being to better understand GxExM interactions and generate data for model validation.
Under objective 2, we have almost finalized the parameterization of several cultivars adapted to the West and Central Africa (WCA) region, using existing evaluation data. Then we have initiated the characterization of stress scenario in the region. We have also started the trait assessment in a large set of parental lines of many BCNAM populations developed in Mali. Last but not least, we have tested the relevance and suitability of Marksim-generated weather data and these appear to be suitable to simulate changes in genetic and agronomic characteristics, which would allow us to perform simulation in grids of generated weather, without to depend on observed weather data (which is always a limitation).
A meeting was held in April 2015 in Brisbane, to update the group on the project’s progress and this report is a reflection of the discussion and progress reported at that recent meeting. The meeting participants also updated each other of other projects dealing with sorghum improvement, for the sake of continuing a good coordination of efforts.
Annual Report 2016
This project in an extension to a previous 4-year phase where the proof-of-concept was established of the value of introgressing genome segments conferring the capacity to maintain green leaf area under terminal water stress conditions (staygreen quantitative trait loci, QTL) for improving both the grain and the stover productivity of sorghum during the post-rainy season in India. An important aspect of that work was that several QTLs had also a beneficial effect on multiple parameters of the nutritional quality of the crop residue, important for cattle feeding, without having trade-offs with grain and stover productivity. Higher stover quality confers a price premium and also lead to higher milk productivity or weight gain. This work therefore showed the potential to improve different components of the sorghum value chain at the same time. We have also found that the staygreen QTL did not have the same effects in all genetic background where they were introgressed, implying the need to check intermediate introgression products. Finally, a thorough characterization of the stress patterns in the target regions has shown highly variable patterns of stress, requiring specific breeding/agronomic solutions to specific stress scenarii. The second phase is therefore focused on introgressing the staygreen QTL segments into the background of sorghum cultivar adapted to the target region in India, a work that is done in partnership between ICRISAT and IIMR (Indian Institute for Millets Research), and using crop simulation modelling to better target genetic and agronomic management packages for specific stress scenarios. The Objectives of the second phase are:
1. Develop an application phase in which several staygreen QTL will be introgressed into the most popular cultivars currently used by farmers in the post-rainy sorghum track of India. This first Objective is supported by ACIAR.
2. Expand the environmental characterization of stress scenarios to the West African region, and develop new breeding populations - Backcross Nested Association Mapping (BCNAM) populations - involving the most popular cultivars of the postrainy sorghum area of India as recipients and new donors of beneficial traits identified in the first phase. This second Objective is supported by the CGIAR Research Program on Dryland Cereals (CRP DC).
In Objective 1, much of the work has so far focused on refining the QTL interval for critical QTL. Five SNPs for stg3A and 21 SNPs for stg3B shortlisted for monoplex KASPar SNP assay are used for the introgression of these 2 critical QTLs in the different Rabi-adapted backgrounds. The introgression work is in good progress and some introgressions have reached either BC3F1 or BC4F1 stage. Under this objective field trials have also been carried out to test the effect of staygreen introgression, nitrogen treatment and plant density on grain and stover yield and on grain and stover quality, the purpose being to better understand the Genetic x Environment x Management interactions and generate data to validate the models. As a new project activity, grain nutritional aspects are being investigated, showing here also the beneficial effects of staygreen introgression on grain quality. We have also refined the ex-ante analysis carried out in the first phase by using modelling outputs as an input to the economic analysis, and we are further segregating these outputs into grain and stover productivity advantage of the staygreen introgression to better target farming communities focused on either grain or stover production. As such, we have shown that targeting specific segments of the sorghum value chain (grain or stover) brought higher economic return than using a single technology for the entire region.
Under Objective 2, we have finalized the parameterization of several cultivars adapted to the West and Central Africa (WCA) region, using existing evaluation data. Then we have undertaken the characterization of stress scenario in the region. We have also assessed a large set of parental lines of many BCNAM populations developed in Mali for the capacity to restrict transpiration under high vapor pressure deficit, i.e. a trait that underlies staygreen expression.
No meeting was held in early 2016 given the recent interactions between project participants. Rather, it has been chosen to group the annual meeting with another sorghum project meeting (USAID-Climate Resilient Sorghum Innovation lab) dealing with related issues for the sake of continuing a good coordination of efforts.
Sorghum grown in India in the post-rainy season (Rabi) relies on residual soil moisture, and the crop is commonly exposed to terminal drought stress. But there is a ready market for its high-quality grain and stover (used as fodder on dairy farms). Steps to improve productivity while maintaining quality offer an attractive opportunity for sorghum farmers to improve their incomes. Genetically improving the efficiency of using stored soil moisture is a prime target to maximise grain/stover production and quality of Rabi sorghum. This project aims to achieve this through the application of DNA sequences known as quantitative trait loci (QTLs). The project scientists will introduce marker-assisted introgression of stay-green QTLs into sorghum lines, enhancing both the quality and the quantity of grain/stover of postrainy sorghum. They will also use modelling to identify the key physiological traits involved in a higher, more stable yield across water-limited environments of India and Australia, and the key stay-green QTLs contributing to these traits. The publicly available QTL isolines lines developed in this project will be the basis of new varieties to be bred in a subsequent phase.