Genetic Dissection of Stem Water-Soluble Carbohydrates and Agronomic Traits in Wheat under Different Water Regimes

Drought is a major environmental stress threatening wheat (Triticum aestivum L.) productivity worldwide. Although drought impedes wheat performance at all growth stages, it is more critical during the flowering and grain-filling phases and results in substantial yield losses. In this context, stem water-soluble carbohydrates (SWSC) were dissected at flowering and grain filling stages under drought stress (DS) and well-watered (WW) conditions using a population consisted of 116 wheat accessions in this research. The main goal was to dissect the genetic basis of water-soluble carbohydrates and the agronomic traits using association mapping approach and identify linked molecular markers. The results showed significant and positive correlations for stem water-soluble carbohydrates at grain filling (SWSCG) with accumulating efficiency of stem water-soluble carbohydrates (AESWSC) and grain filling efficiency at the late stage (GFEL). The accumulating and grain filling efficiency at grain filling stage could play an important role for SWSC especially under DS condition. Four favorable alleles for plant height (PH) and grain yield (GY) were identified in two water environments. Xbarc78-4A163 and Xbarc78-4A155 were variant alleles for PH which were identified in both water regimes. Whereas Xwmc25-2D151 and Xgwm165-4B191 positively linked with GY in WW. Although Xwmc420-4A121 and Xwmc112-2D215 were alleles for stem water-soluble carbohydrates at flowering (SWSCF) and SWSCG in DS but the frequency were < 5% so they were considered as rare alleles. These SSR markers which explained significant level of phenotypic variability for chosen traits could be used for selection of genotypes in wheat breeding programs through marker-assisted selection.


Introduction
Wheat is one of the three major cereals.Global annual production is 727.87 million metric tons in 2014-2015(USDA, 2016)).To ensure food for the rapidly growing world population, wheat production needs to double by 2050 (Alexandratos & Bruinsma, 2012).Further increases in wheat production depend on higher yields rather than an increase in cropping area (Araus et al., 2003).Declining water resources challenge this notion as water availability impacts heavily on crop yields (Kang et al., 2008).Among all the abiotic stress factors that limit crop productivity, drought is the most devastating one and the most difficult to breeders' efforts.Breeding efforts in the past, to improve drought tolerance has been hindered by its quantitative genetic basis and the poor understanding of the physiological basis of yield under water-limited conditions (Tuberosa & Salvi, 2006).Genomic approaches enable the identification and selection of chromosome regions harboring genes/QTLs (Quantitative trait loci) controlling agronomic traits and yield in crops (Collins et al., 2008;Cooper et al., 2009;Tuberosa & Salvi, 2006).Among such approaches, association mapping is increasingly being adopted as a method complementary to traditional bi-parental linkage mapping to identify genotype-phenotype associations i.e. molecular marker-trait associations (Sorrells & Yu, 2009;Waugh et al., 2009).
(Hordeum vulgare), and oats (Avena sativa) during the period from stem elongation to early phase of grain filling and serve as temporary carbohydrate reserves, commonly called the stem carbohydrate reserves (Blum, 1998;Gebbing, 2003).In general, WSC accumulate until 10~20 days after anthesis, and the reserved WSC can reach more than 40% of total stem dry weight in wheat (Rebetzke et al., 2008).
The contribution of WSC to final yield and kernel size is 10~20% of total grain weight under normal condition (Gebbing & Schnyder, 1999).Drought stress during grain filling, often involving not only water stress but also heat, inhibits current assimilation and damages photosynthetic organs, especially leaves.When photosynthetic activity is suppressed, the reserved WSC play a more important role in partial compensation of the reduced carbon supply.In addition, drought induced reserved WSC mobilization with higher efficiency, potentially contributing up to 70% of grain dry matter (Goggin & Setter, 2004;Rebetzke et al., 2008).Therefore WSC is a major contributor to wheat grain yield and grain size in all environments but especially where photosynthesis is compromised as occurs where water is limiting (McIntyre et al., 2012).Grain yield depends on carbon from two resources: flag leaf photosynthesis and remobilization of water-soluble carbohydrates, mainly fructans, from the wheat stems (Yang & Zhang, 2006).In term, high water-soluble carbohydrate content in the stem has been suggested as a selection criterion for use inbreeding.WSC QTL have been reported in rice (Nagata et al., 2002;Takai et al., 2005), wheat (Rebetzke et al., 2008;Yang et al., 2007), maize (Thévenot et al., 2005), barley (Teulat et al., 2001) and perennial rye grass (Turner et al., 2006).With the rapid increases in number of molecular markers, association analysis has become an important tool for dissection of complex traits (Bradbury et al., 2007).
The main objective of this research is to find out the favorable alleles responsible for expression of water-soluble carbohydrates and important agronomical traits by association mapping to help breeders to improve their breeding programs through marker-assisted selection.

Plant Materials and Field Trials
The plant material was a population consisted of 116 winter wheat genotypes, which were collected from the different regions of China, including landraces, advanced lines and modern cultivars released from 1940s to 2000s.The wheat accessions were planted in the experiment station of the Institute of Crop Science, Chinese Academy of Agricultural Sciences at Changping (116°13′E; 40°13′N), Beijing in September 25 th 2013 and harvested in the mid of June 2014.
The experimental unit was a plot consisted of four rows of 2 m length with row spaced 30 cm apart.Forty seeds were sown in each row.The field plots were subjected to two water treatments: full irrigation (well-watered, WW) and rain fed (drought stress, DS) recorded 163 mm of rainfall during 2013~2014 growing season.The WW plots were watered with 750 m 3 /ha (75 mm) at the pre-overwintering, booting and flowering stages.The amount of rainfall was insufficient during each corresponding period.A randomized complete block design with three replicates was used for both WW and DS treatments.Plants were scored for the following agronomic, physiological and developmental traits: days to heading (DTH), days to flowering (DTF), flag leaf area (FLA), chlorophyll content at flowering (CCF), chlorophyll content at grain filling (CCG), plant height (PH), peduncle length (PL), spike length (SL), length of second internode from the top (LSIT), length of internodes below second internode from the top (LIBSIT), number of spikes per plant (SP), number of spikelets per spike (NSPS), sterile spikelet at the top (SST), sterile spikelet at the middle (SSM), sterile spikelet at the base (SSB), number of grains per spike (NGS), grain yield per plant (GY), grain yield per plot (GYP) and thousand-grain weight (TGW).

Phenotyping of Stem WSC and TGW
The five main stems were cut from the soil surface at two morphological stages, viz.flowering and mid-grain filling (14 days after flowering).Samples were taken from the mid-part of second row of each plot.Leaf blades were removed and the stems with leaf sheaths were cut into two parts i.e. the stem and the spike.The fresh samples were killed at 105 °C for 30 minutes and then keep at 80 °C to dehydrate until a constant dry weight.Stem samples of each accession were chipped into 2~5 mm length.The stem WSC was determined by the near-infrared reflectance spectroscopy (NIRS) regression models (Wang et al., 2011).Briefly, at the first step, partial least square regression models for predicting WSC in the target parts of wheat were developed using selected wavelength regions, spectroscopic pretreatments and the latent variables included in each model.The total amounts of WSC (mg/g dry weight) in each sample were also measured by chemical assay (anthrone colorimetric assay), and used for the cross validation.The NIRS regression models were highly accurate in determination of the true values of WSC measured by chemical assay in the wheat organs tested, according to high coefficients of determination of the true values of WSC measured by chemical assay in the wheat organs tested, according to high coefficients of determination (R 2 > 0.992) and low root mean square errors of prediction (RMSEP < 0.228).

Genotyping by SSR Markers
Seedling leaves were used as the experimental material.Leaf samples were collected from one hundred and sixteen genotypes.Experiment set in vitro and twenty seeds placed in each petri plate.First leaves were collected for DNA extraction.DNA extracted by DNA Quick Plant System Kit (Tiangen Bio.Co. LTD, Beijing, China).Selection of 92 SSR markers based on evenly distributed on different chromosomes and/or linked with WSC or investigated agronomic traits.These SSR markers were found on chromosome 1A, 1D, 2A, 3B, 4A, 4B, 5A, 6B, 7A, 7B and 7D, respectively.SSR analysis was conducted using 92 primer pairs directed to the amplification of di-and tri nucleotide microsatellite loci, originally developed in bread wheat within five different research programs, WMS Wheat Microsatellite (Röder et al., 1998), WMC Wheat Microsatellite Consortium (Gupta et al., 2002), BARC USDA-ARS Beltsville Agricultural Research Center (Song et al., 2005), CFA and CFD (Guyomarc'h et al., 2002;Sourdille et al., 2001).Among 92 primers, sixty primer pairs for wheat microsatellite loci were synthesized by Sangon Biotech (Shanghai) Company.For each primer pair, the sequence of original forward primer was redesigned by adding a universal M13-tail (5'CACGACGTTGTAAAACGAC-3') to their 5' ends.Thirty-two primers were used as fluorescent markers.The universal M13 primers were labelled with different fluorescence dyes i.e. blue and green.
The PCR was conducted in a total volume of 15 μl, containing 20 ng genomic DNA, 10 X supplied PCR buffer including 1.2 mM of MgCl 2 , 2 mM of dNTPs, 5 unit/μl of Taq DNA polymerase, 2 μM of primer.Amplification reactions were conducted using a Senso Quest lab cycler.Gradient PCR was used to determine optimal annealing temperature for each primer pair.The M13 primers were labelled with different fluorescent dyes i.e. blue and green, allowing done PCR again with the following reaction mixture containing 162 μl M13 dye (AB Applied Biosystem, USA), PCR buffer 9 μl, Taq DNA polymerase 7.2 μl.Added 2.1 μl of above mixture in 96 well plate that already amplified at above mentioned conditions.This mixture was further amplified under the following conditions 94 °C for 5 min; 16 cycles of 94 °C for 45 s, 56 °C for 45 s and 72 °C for 45 s; followed by a 10 min extension at 72 °C.The PCR products were analyzed by electrophoresis in 2% agarose gel.The PCR were carried out separately for each microsatellite and the mixture of PCR products of three different markers with different dyes was made for simultaneous detection of the amplified alleles.Sequencing of 92 simple-sequence-repeat (SSR) loci has done by ABI3730 DNA analyzer.The PCR products were analyzed by Gene Mapper Software.The polymorphism information content (PIC) values (Botstein et al., 1980) were calculated using Power Marker software v3.25 (Liu & Muse, 2005).Population structure was estimated by STRUCTURE v2.3.2 using data from 92 SSR markers.Twenty subpopulations (k = 1 to 20) were set with a burn-in period of 50,000 iterations and a run of 500,000 replications of Markov Chain Monte Carlo after burn in.The Δk method was applied according to LnP(D) in STRUCTURE, and the output and result were estimated (Pritchard et al., 2000).

Marker-Trait Association and Statistical Analysis
Association between markers and traits was calculated using a general linear model (GLM) method in TASSEL v2.1 (Yu et al., 2006).The population structure matrix (Q) obtained from the STRUCTURE software and relative kinship matrix (k matrix) derived from the unlinked marker data estimated by TASSEL v2.1 were combined to covariate in the association tests to reduce false positive rate.The significant marker-trait associations were declared by P ≤ 0.01 and the magnitude of the allele effects were evaluated by R 2 -marker.Analysis of variance (ANOVA) was conducted using SPSS 16.0.Pearson's correlation coefficient among the traits under two water regimes were calculated by SPSS 16.0.Broad sense heritability (h 2 B ) was computed by QTL IciMapping (http://www.isbreeding.net/).

Drought Stress Induces SWSC
On the basis of the ANOVA, the SWSC of 116 genotypes at flowering and grain filling stages were highly affected by the DS.The mean square value of SWSCG under drought stress condition was significantly higher than that under well-watered condition, but the traits such as AESWSC, GFEE, GFEL, and TGWG showed non-significant differences (Table 1).SWSC at flowering stage ranged from 140.75 to 298.5 mg/g whereas SWSC at grain filling stage ranged from 140.63 to 346.37 mg/g under drought stress (Figure 1).It is reported that stem water-soluble carbohydrates are a major carbon source for grain filling under drought stress (Zhang et al., 2016).In addition, stem water-soluble carbohydrates buffer wheat grain yield against drought stress for photosynthesis during the grain filling stage (Li et al., 2015).In the present study, moderate to high broad sense heritability for SWSC at both stages under WW and DS conditions were estimated ranged from 0.76~0.78 to 0.71~0.80(Appendix 1).
Figure 1.Frequency distribution of SWSC at flowering and grain filling stages under DS and WW condition Note.DS: drought stress; WW: well-watered; SWSC: stem water-soluble carbohydrates; SWSCF: stem water-soluble carbohydrates at flowering stage; SWSCG: stem water-soluble carbohydrates at grain filling stage.

SWSC Correlated with Other Traits
Correlation coefficients among all traits under the two water regimes are given in Table 2.The traits associated with SWSC at two different growth stages such as SWSCF and SWSCG were highly significant correlation with each other under WW and DS conditions.SWSCF showed a highly significantly negative correlation with AESWSC under both water conditions.While, SWSCF had a significantly positive relationship with thousand-grain weight at grain filling (TGWG) and maturity (TGWM) under DS condition.In WW condition, SWSCF were also significantly positive correlation with TGWG and TGWM but had smaller values of correlation.However, SWSCG exhibited a highly positive correlation with AESWSC, TGWM and grain filling efficiency at late stage (GFEL) in both environments.On chromosome 2D, four traits such as GY (DS), LIBSIT (DS and WW), PH (DS) and SWSCG (DS) were tagged by SSR markers Xwmc25 (18 cM) and Xwmc112 (28 cM).The LIBSIT (DS and WW) and TGWG (DS) were associated by Xgwm285 (61 cM) and Xbarc164 (70 cM) on chromosome 3B.LIBSIT (DS), LSIT (DS and WW), PH (DS), PL (DS), SWSCF (DS), TGW (WW) were tagged by the Xwmc420 (7 cM) and Xgwm397 (18 cM) on chromosome 4A.Xbarc78 (71 cM) was associated with PH and LIBSIT also on chromosome 4A in both water conditions.On chromosome 4B, SP (DS) and TGW (DS and WW) and GYP (WW) were identified by markers Xgwm149 (31 cM) and Xgwm165 (28 cM).The LIBSIT, TGWG and TGW were separately tagged by three markers Xgwm293 (52 cM), Xgwm415 (56 cM) and Xgwm304 (61 cM) on chromosome 5A under DS condition.

Favorable Alleles for Different Traits
Four of the 17 loci had significantly favorable allelic effect on multiple traits were identified in DS and WW conditions (Table 3).Most of the traits had < 5% frequency of the SSR marker allele in both conditions.It was considered as a rare allele.Under DS condition, 163 bp allele of Xbarc78 on chromosome 4A (Xbarc78-4A 163 ) showed positive effects on LIBSIT and PH, respectively.This allele had a frequency of 5.17% on both traits.Under WW condition, two marker alleles Xwmc25-2D 151 and Xgwm165-4B 191 separately increased GY and GYP, but Xbarc78-4A 155 showed a negative effect on PH.These favorable alleles associated with important traits that could be contribute to increase wheat production in different water environments.

Variation of Stem Water-Soluble Carbohydrates under Different Water Condition
The stem WSCs become more important for grain yield in cereal crops under abiotic stress (Blum, 1998;Kiniry, 1993;van Herwaarden et al., 1998).A good capacity for stem reserve and remobilization of WSC has been proposed as a drought adaptive trait in a conceptual model for drought tolerance (Reynolds, 1999).In this study, variation in SWSC among 116 genotypes at two developmental stages has been observed under two water regimes.The phenotypic means for this trait were more affected by drought stress.The means under drought stress were significantly higher than those under well-watered.The present observations were consistent with the view of Zhang et al. (2014) that reported fructan synthesis is induced by drought stress, and that drought tolerant plants can manufacture more fructans.Fructans are the major component of WSC, insert between the head groups of phospholipids, acting as compatible solutes in cells to protect cell membranes and proteins from osmotic damage (Rathinasabapathi, 2000;Vereyken et al., 2001).
In our research, SWSCF and SWSCG under drought stress were overall higher than those under well-watered condition.The tolerant cultivars activate their protection mechanisms faster and more efficiently than the sensitive ones to cope with stress conditions (Goggin & Setter, 2004;Gupta et al., 2011).WSC mobilizes from the stem during the later phase of grain filling and thus can become an important source of assimilate for grain yield in wheat under terminal drought stress conditions (Blum, 1998).Stem WSC accumulation is influenced by environmental factors (Blum, 1998;Ruuska et al., 2008;Ruuska et al., 2006).However, considerable genotypic variation in stem WSC concentration has been observed in wheat (Ruuska et al., 2006;Xue et al., 2008).

Heritability of Stem Water-Soluble Carbohydrates
In the present study, moderate to high broad sense heritability for SWSC at two stages (flowering and grain filling) under WW and DS conditions were estimated ranged from 0.76~0.78 to 0.71~0.80.High heritability indicates potential for phenotypic selection of WSC among families in breeding programs that target adaptation to terminal droughts (Rebetzke et al., 2007).The ability to store and remobilize large amounts of WSC to grain has been suggested as a selection criterion for wheat breeding due to its high heritability and positive linear ship with grain yield (Dreccer et al., 2009;Gupta et al., 2011).

Correlation of Stem Water-Soluble Carbohydrates and Grain Yield Traits
Mobilization of WSC during grain filling can potentially contribute about 20% of the final grain weight under non-stress conditions, and up to 70% or more of grain dry matter under drought stress in wheat (Goggin & Setter, 2004).It has also been reported that stem WSC concentration at anthesis or shortly after anthesis (i.e. at the stem WSC accumulation phase) is a good indicator of positive association between WSC level and grain weight or yield in wheat (Foulkes et al., 2007).Our research exhibited that SWSCF showed significantly negative correlation with AESWSC while positive association with TGWG and TGWM under both normal and stress conditions.In contrast, SWSCG had significant correlation with all traits such as AESWSC, TGWG, TGWM, GFEE and GFEL under both conditions with the exception of TGWG under WW condition that reflected non-significant association between SWSCG and TGWG.It is suggested that SWSCG could play a key function in the subsequent release of carbohydrates from stem to grain.Current results are in conformity by Yang et al. (2007).WSC are recognized as an important source of grain dry matter for grain filling, especially when current photosynthesis is inhibited by drought stress.Water deficit during grain filling stimulates senescence of the whole plant and enhances mobilization of reserved WSC to the grains (Araus et al., 2002;Guoth et al., 2009).
Negative correlation was observed between SWSCG and GFEE, whereas there was a positive association of SWSCG with GFEL under both environments.It is suggested that SWSCG could play an important role in grain filling of wheat.AESWSC exhibited positive correlation with TGWM and GFEL in both drought stress and well-watered conditions.It indicated that accumulation of water-soluble carbohydrates increased the grain filling efficiency at the later stages and that contributed to heavier the grains in term of thousand-grain weight.Xue et al. (2008) also reported a positive and significant relationship with WSC and grain weight in wheat lines.
In this study, thousand-grain weight and grain filling efficiency under drought stress was slightly lower than those under well-watered condition during the early grain filling period.Li et al. (2015) reported higher thousand-grain weight with increased grain filling efficiency under drought stress as compared to well-watered condition during the early grain filling period.Water deficit at grain filling induces carbon mobilization from tillers to the main stem ear (Blum et al., 1994).Rebetzke et al. (2007) reported that wheat progeny with high WSC produced higher grain weight and larger diameter, significantly reducing grain shriveling.WSC accumulation and remobilization are influenced by many factors, making the relationship between WSC and TGW more complex.

Genetic Polymorphism
A total of 1600 alleles were identified from 92 SSR loci when scored on 116 genotypes.High levels of polymorphism were observed for the markers, with a range of 3~49 and average of 17.39 alleles per marker locus, which indicated that the diversity of wheat accessions in this study was relatively high.The microsatellite markers presented high level of PIC in comparison with other markers in wheat (Gupta et al., 2008).It is suggested that genomic SSR markers is powerful for the evaluation of genetic polymorphism, similar results were obtained in 103 wheat accessions (Liu et al., 2010).
The present research results indicated that Xwmc420-4A was associated with LSIT under both drought stress and well-watered conditions, and also associated with LIBSIT, PH, PL, and SWSCF in drought stress.Xwmc89, a marker locus at the same position as Xwmc420, was reported that significantly associated with all grain related QTLs and explained the high proportion of phenotypic variation (Kirigwi et al., 2007).A SSR marker Xwmc48 was identified associating with QTL for grain yield (Kirigwi et al., 2007).In addition, Liu et al. (2010) identified a QTL for PH nearby marker Xwmc420.In current study, Crossa et al. (2007) found a DArT marker wPt8271, which close to Xbarc70 and Xbarc78, was associated with GY, whereas Liu et al. (2010) reported for grains per spike and thousand-kernel weight.

Chromosome 5A
The next important chromosome in this research was 5A.It was tagged by three SSR markers namely Xgwm293 (52 cM), Xgwm415 (56 cM) and Xgwm304 (61 cM).The distance among these markers were less than 10 cM.Yang et al. (2007) detected QTL with flanking marker of Xwmc524~Xgwm595 for TGWG, it was on the same chromosome as our studies but with different position and also was associated with same trait as in this study.Furthermore, (Maccaferri et al., 2011) found Xbarc197 located on chromosome 5A (53 cM) adjacent with Xgwm293 (52 cM) and also nearby Xgwm415 and Xgwm304 (56 and 61 cM apart), was associated with grain yield using association analysis.In our study, Xgwm293, Xgwm415 and Xgwm304 were associated with LIBSIT and TGWG and TGW under DS, while TGW is one of the most important factors affecting grain yield.

Chromosome 2D
Chromosome 2D is important for yield and yield components and also for stem water-soluble carbohydrates.Three SSR markers i.e.Xwmc25 (18 cM), Xwmc112 (28 cM) and Xgwm382 (100 cM) were associated with GY (WW), PH (DS), LIBSIT (DS and WW), SWSCG (DS) and CCG (WW).The distance between two markers were 10 cM so it could be considered as moderately linked according to the classification of Marone et al. (2012) however Xgwm382 was away from these markers.It acted as an independent marker.A marker Xcfd17 associated with WSC under WW with 64 cM was apart from our finding but on the same chromosome 2D (Zhang et al., 2014).
In addition to this, Li et al. (2015) detected a marker Xgwm261-2D was on 23 cM and it was nearby a marker Xwmc112-2D (28 cM) was associated with SWSCG under terminal DS.The distance of these two markers was close and also was associated with the same trait.This finding was strengthened by our result.According to Rebetzke et al. (2007), the QTL on chromosome 2D was also mapped for WSC.It was also supported by Yang et al. (2007), but they reported QTL for WSC away from those markers that we found, and also it was reported at flowering stage.Dodig et al. (2012) identified SSR marker Xgwm484 (41 cM) for chlorophyll content at grain filling but far distance from that marker we reported.In current study, Xgwm382 was located at 100 cM and associated with chlorophyll content at grain filling under well-watered condition.

Chromosome 4B
Chromosome 4B were tagged by two SSR markers i.e.Xgwm149 (31 cM) and Xgwm165 (28 cM) were associated with SP (DS), TGW (DS and WW) and GYP (WW).These two markers are tightly linked with each other.In the present study, Xgwm149 was associated with SP (DS) and TGW in both environments whereas Xgwm165 was associated with GYP under WW condition only while SP, TGW and GYP are the most important traits that affecting grain yield.A QTL was reported for TGW under post-anthesis drought stress on chromosome 4BL (Nezhad et al., 2012).

Chromosome 3B
We detected that Xgwm285 (61 cM) and Xbarc164 (70 cM) tightly linked with each other.They were associated with LIBSIT (DS and WW) and TGWG (DS) on chromosome 3BL.Whereas Xgwm389 was associated with tiller number on 3BS (Dodig et al., 2010).

Chromosome 1D
We identified that Xwmc432-1D (23 cM) was associated with TGWG under DS.An SSR locus Xgwm337 (48 cM) was associated with GW on the same chromosome but far from Xwmc432 (Groos et al., 2003).A QTL for lodging resistance was reported on chromosome 1D in a wheat DH population (Verma et al., 2005).

Chromosome 2A
Marker Xwmc296 (49 cM) on chromosome 2A was associated with LIBSIT in WW condition.It was detected QTL for PL on the same chromosome with different SSR marker Xgwm294 (76 cM) (Brbaklić et al., 2015).

Chromosome 7D
In an interval map of 77 cM on chromosome arm 7DS, SSR marker Xgwm295 was identified for TGWG in DS.
Previously reported QTL linked with TGW under terminal drought stress that close to Xgwm295 marker (Nezhad et al., 2012).

Favorable Alleles for Different Traits
Xbarc78-4A was associated with LIBSIT and PH under DS, however it was only associated with PH under WW condition.It was considered as favorable alleles for these two traits.For well-watered condition, two favorable alleles were also identified on chromosome 2D and 4B that associated with GY and GYP.These traits are important for the contribution to improve the production of wheat.

Conclusions
It is inferred that SWSC were accumulated more at grain filling stage in DS and it is considered as increase fructans for self-protection.High heritability estimated for SWSCG under drought stress.The ability to accumulate WSC and high heritability in grain filling stage suggested as a selection criterion for wheat breeding.Total of seventeen significant marker-trait associations for 13 traits were detected.Chromosomes 2D, 4A and 5A are the most important with respect to traits and loci distance.Xwmc25, Xwmc112 and Xgwm382 were associated with GY (WW), PH (DS), SWSCG (DS) and CCG (WW).The next is 4A found two markers i.e.Xwmc420 and Xgwm397 are moderately linked with each other.Xwmc420 was associated with PH, PL and SWSCF for DS, whereas Xgwm397 was associated with TGW for WW condition.Another chromosome is 5A, three SSR markers namely Xgwm293, Xgwm415 and Xgwm304 were associated with TGW under DS.
Four favorable alleles were identified in two water environments, including Xbarc78-4A 163 increasing plant height in drought stress, but Xbarc78-4A 155 decreased plant height under well-watered condition.Xwmc25-2D 151 and Xgwm165-4B 191 were considered as favorable alleles for increasing grain yield under well-watered condition.All these markers are firstly reported with the traits in our study and expected to be helpful for marker assisted selection in wheat improvement.

Figure
Figure 2. Pop Estimated Ln P

Table 1 .
Mean squares for various traits under two water regimes Note.MS: mean squares; ** and *** indicate P = 0.01 P = 0.001, respectively; CV: coefficient of variation; NS: non-significant; ND: not determined; un bold value shows under DS while bold value shows under WW conditions; DS: drought stress; WW: well-watered; SP: number of spikes per plant; SL: spike length; PL: peduncle length; LSIT: length of second internode from the top; LIBSIT: length of internodes below second internode from the top; NSPS: number of spikelet per spike; SST: sterile spikelet at the top; SSM: sterile spikelet at the middle; SSB: sterile spikelet at the base; NGS: number of grains per spike; GYP: grain yield per plot; TGW: thousand-grain weight; PH: plant height; DTH: days to heading; DTF: days to flowering; FLA: flag leaf area; GY: grain yield per plant; CCF: chlorophyll content at flowering; CCG: chlorophyll content at grain filling; SWSCF: stem water-soluble carbohydrates at flowering stage; SWSCG: stem water-soluble carbohydrates at grain filling stage; AESWSC: accumulation efficiency of stem water-soluble carbohydrates; GFEE: grain filling efficiency at the early stage; GFEL: grain filling efficiency at the late stage; TGWG: thousand-grain weight at grain filling stage.

Table 3 .
Favorable alleles associated with phenotypic traits *