QTLs maintaining grain fertility under salt stress detected by exome QTL-seq and interval mapping in barley

Abstract

Enhancing salt stress tolerance is a key strategy for increasing global food production. We previously found that long-term salinity stress significantly reduced grain fertility in the salt-sensitive barley (Hordeum vulgare) accession, ‘OUC613’, but not in the salt-tolerant accession, ‘OUE812’, resulting in large differences in grain yield. Here, we examined the underlying causes of the difference in grain fertility between these accessions under long-term treatment with 150 or 200 mM NaCl from the seedling stage to harvest and identified quantitative trait loci (QTLs) for maintaining grain fertility. In an artificial pollination experiment of the two accessions, grain fertility was significantly reduced only in OUC613 plants produced using pollen from plants grown under NaCl stress, suggesting that the low grain fertility of OUC613 was mainly due to reduced pollen fertility. Using QTL-seq combined with exome-capture sequencing and composite interval mapping of recombinant inbred lines derived from a cross between OUE812 and OUC613, we identified a QTL (qRP-2Hb) for grain fertility on chromosome 2H. The QTL region includes two genes encoding an F-box protein and a TIFY protein that are associated with male sterility, highlighting the importance of this region for maintaining grain fertility under salt stress.

Introduction

Soil salinity is one of the most serious problems limiting crop production. Over 6% of the worldwide land area is affected by salt; approximately 400 and 430 million hectares are covered with saline and sodic soil, respectively (FAOSTAT 2012). The world’s population is steadily increasing and is expected to reach 9.1 billion by 2050, making it necessary to increase annual crop production from the current rate of 210 million tons to 300 million tons (FAO 2009). Since the amount of productive agricultural land available is not expected to increase, yield per unit area in farmlands must be increased to ensure adequate crop production. In saline environments, salt-tolerant crops are more productive than salt-sensitive crops (Munns and Tester 2008). Therefore, one important way to improve global crop production is to develop salt-tolerant crops that perform reliably in salinized soil.

Salt tolerance differs significantly among crop species (Munns et al. 2006, Munns and Tester 2008, Rawson et al. 1988). Among Gramineae crops, for example, barley (Hordeum vulgare) is the most salt tolerant, while rice (Oryza sativa) is the least (Munns and Tester 2008). Moreover, salt tolerance differs among cultivars in barley, wheat (Triticum aestivum), rice, and mung bean (Vigna radiate) (Rawson et al. 1988, Richards et al. 1987, Royo and Aragues 1999, Slavich et al. 1990, Win et al. 2011, Yamanouchi 1989).

The negative effects of soil salinity on plant growth and yield are due to two main factors: osmotic stress and ion toxicity. The osmotic and water-deficit-inducing effects of salt lead to reduced growth (Munns 2005). Under long-term salinity conditions, Na⁺ and Cl⁻ concentrations in plants increase, inhibiting enzyme activity. Two major types of mechanisms mediate tolerance to ion toxicity: those minimizing the entry of salt into the plant, and those minimizing salt levels in the cytoplasm (Munns 2005).

Identifying quantitative trait loci (QTLs) related to salt tolerance is a promising approach for improving salt tolerance in barley (Saisho and Takeda 2011, Sato et al. 2011). Mano and Takeda (1997) and Zhou et al. (2012) identified QTLs associated with salt tolerance at germination and in seedlings, and at the vegetative stage, respectively. Shavrukov et al. (2010) identified a QTL related to Na⁺ exclusion, which they named HvNax3. Rivandi et al. (2011) identified a QTL associated with shoot Na⁺ content named HvNax4. Xue et al. (2009) identified QTLs for agronomic traits, such as dry weight, plant height, and grain yield under salt stress. However, QTLs for traits associated with enhanced grain yield or directly related to grain yield under salt stress in barley have not yet been identified.

Recently, QTL-seq, a technique that combines bulked segregant analysis (BSA) with next generation sequencing, has been developed for rapid mapping of QTLs (Takagi et al. 2013). An advantage of QTL-seq is that it does not require DNA marker development or marker genotyping. Although this method is suitable for species with small genomes such as rice (380 Mbp), it has not been used in species with large genomes such as barley (5.1 Gbp), due to the high cost of whole-genome sequencing and the difficulty in assembling large amounts of sequencing data. However, Hisano et al. (2017) recently demonstrated that QTL-seq could be performed in barley when combined with exome-capture sequencing (termed “exome QTL-seq”) using barley exome-capture method developed by Mascher et al. (2013). The authors used exome QTL-seq to detect a QTL in barley for the black lemma and pericarp (Blp) locus and two QTLs for resistance to net blotch disease (Hisano et al. 2017). However, this technique has yet to be used to detect QTLs for physiological traits that are difficult to evaluate, such as salt tolerance. Increasing the detection power of exome QTL-seq for QTLs responsible for salt tolerance should accelerate QTL detection for complex physiological traits in crop species with large genomes.

We previously detected a significant genotypic difference in grain yield between barley accessions OUE812 (salt tolerant) and OUC613 (salt sensitive) under long-term salinity conditions and found that the difference in grain yield between these accessions is mainly due to differences in grain fertility (ripening percentage) (Hirasawa et al. 2017). In a current study, we investigated the causal factors of reduced grain fertility in barley and identified QTLs responsible for maintaining fertility under long-term salinity conditions via QTL-seq and composite interval mapping using recombinant inbred lines derived from a cross between OUE812 and OUC613.

Materials and Methods

Plant materials

A cross was made between OUE812 (salt-tolerant six-row barley [Hordeum vulgare] landrace collected in Ethiopia) and OUC613 (salt-sensitive two-row barley landrace collected in China). A single seed descent population was developed via self-pollination of 140 F₂ plants to the sixth to seventh generations, referred to hereafter as F₆ and F₇ recombinant inbred lines (RILs), respectively (Supplemental Fig. 1).

Plant cultivation and NaCl treatment

OUC613, OUE812, and the 96 F₆ or F₇ RILs were sown in 1.1 L containers (H 270 mm × W 40 mm × L 100 mm) filled with vermiculite at a rate of six seeds per container on February 4, 2013 and February 4, 2014, with three replications. After seedling establishment, the plants were thinned to one plant per container. The plants were grown under a rain-shield glass roof in vermiculite with 1/10 strength Hoagland solution until 70 days after sowing (DAS), followed by growth in 1/2 strength Hoagland solution until harvest. The full-strength nutrient solution contained 4.0 mM Ca(NO₃)2, 4.0 mM KNO₃, 1.0 mM MgSO₄, 1.0 mM NH₄H₂PO₄, 1.0 mM (NH₄)₂HPO₄, 1 mM NaCl, 36 μM FeNaEDTA, 12.5 μM H₃BO₃, 0.25 μM CuSO₄, 1.0 μM MnSO₄, 1.0 μM ZnCl₂, and 0.4 μM NaMoO₄. NaCl treatment began at 30 DAS. Hoagland solution without additional NaCl (control), with 150 mM NaCl (150 mM NaCl treatment) or with 200 mM NaCl (200 mM NaCl treatment) was supplied twice weekly until harvest. For QTL analysis, 200 mM NaCl and 150 mM NaCl treatment were conducted in 2013 and 2014, respectively. When supplying the nutrient solution, each container was filled completely with solution, and excess solution was thoroughly drained from the bottom of the container. This procedure was repeated twice each time to renew the solution completely and to avoid salt accumulation in the vermiculite.

Pollination experiment between plants under NaCl treatment and control

OUC613 and OUE812 were grown in 14 L containers with vermiculite in nutrient solution containing 150 mM NaCl or nutrient solution without additional NaCl (0 mM; control). For emasculation, the lateral spikelets of the female parent were cut off, and every other central spikelet was also cut off, leaving five spikelets on each side for a total of 10 spikelets per spike. The anthers of control plants and of those treated with 150 mM NaCl were gently removed with forceps approximately three days before awn emergence. Artificial pollination was performed 3 to 5 days after emasculation. Pollen from male control plants and from those treated with 150 mM NaCl was used to pollinate the pistils of female plants treated with 150 mM NaCl or control. Eight spikes were pollinated for each crossing combination. The pollinated spikes were covered with paraffin bags, and the plants were grown until harvest.

Trait determination

The plants were harvested at the end of the growth period when the leaves had turned completely yellow. The aboveground parts of plants were dried at 80°C in a ventilated oven for at least 3 days before dry weight was measured. For evaluating grain fertility, ripening percentage (RP) was determined as the filled grain number divided by the total floret number. The relative ripening percentage (RRP) was calculated as the ratio (percentage) of RP of a plant under 150 mM or 200 mM NaCl treatment to the average RP of the control plants. Grain weight per plant was calculated at 12.5% moisture content.

QTL-seq

(1) Construction and sequencing of bulked exome capture libraries

Twenty lines with high RRP under salt conditions (high fertility group: HF) and 20 lines with low RRP under salt conditions (low fertility group: LF) were selected. Genomic DNA was isolated from fresh leaves from each F₆ RIL using a DNeasy Plant Mini Kit (Qiagen, Hilden, Germany), and DNA concentration was measured using a Qubit R 2.0 fluorometer (Thermo Fisher Scientific, Waltham, MA, USA). DNA from each RIL was adjusted to a concentration of 20 ng/μL and mixed in an equal ratio to produce two bulked DNA pools per trait from the HF and LF groups, which are referred to hereafter as the HF bulk and LF bulk, respectively. Exome-captured (EC) libraries were prepared from these DNA pools as described previously (Hisano et al. 2017). The EC libraries were sequenced via two runs of MiSeq (Illumina, San Diego, CA, USA) with an MiSeq v2 Reagent Kit, 300 Cycles (Illumina) following the manufacturer’s protocol.

(2) RNA sequencing of OUC613

To isolate RNA from seedling tissues, OUC613 seeds were germinated on moist filter paper in Petri dishes at 20°C in darkness. Shoot and root tissues were sampled from seedlings with 5 cm long shoots. Total RNA was isolated with TRIzol^® Reagent (Life Technologies Japan, Tokyo, Japan), and DNA was removed using an RNase-Free DNase Set (Qiagen). Libraries for RNA-Seq analysis were constructed from RNA samples using a TruSeq RNA Sample Prep Kit v2 (Illumina). The RNA-Seq libraries were sequenced with a MiSeq Reagent Kit v3 (2 × 300 bp cycles) on the MiSeq NGS system according to the MiSeq System User’s Guide (Illumina), and fastq files were generated from both ends of the fragments.

(3) Sequencing data analysis and generation of SNP-index

The QTL-seq pipeline (Takagi et al. 2013) was used to analyze short reads from the EC libraries. Sequencing reads in which more than 10% of sequenced nucleotides had Phred quality scores of <30 were excluded. Provisional exome sequences (PESs) from cv. Morex developed by Hisano et al. (2017) were used as the reference sequence. In brief, the provisional exome sequences were prepared by ordering the gene models of cv. Morex based on the genome information (IBSC 2012). The concatenated sequences of genes with intervals of 200-bp spacer “N” for each chromosome were used as references. To develop pseudo reference sequences (PRSs) for OUC613 (“OUC613 PRS”), the sequence reads obtained from RNA-seq of OUC613 were aligned to Morex PESs, and the detected SNPs were aligned with those from the Morex PESs. After aligning the short reads obtained from both bulk samples to OUC613 PRS, short reads containing mismatches at more than six positions were excluded with a Coval filter (Kosugi et al. 2009), and SNP-index and ΔSNP-index were calculated at all SNP positions. SNP positions with mapping quality of <30, mapping depth of <14, and SNP-index value of <0.3 and/or >0.9 in both bulks were excluded from analysis, since such SNPs might be false positives caused by sequencing and/or alignment errors. The short genomic reads obtained in this study were deposited at DDBJ-BioProject under accession number PRJDB6669.

DNA marker analysis and QTL mapping

Total DNA was extracted from each RIL (96 lines) using a DNeasy Plant Mini Kit (Qiagen). The DNA was analyzed via an Illumina GoldenGate^® Assay (Illumina), and 146 SNPs that were evenly distributed throughout the genome were selected (Supplemental Fig. 1). A tentative linkage map was constructed according to the marker positions reported by Close et al. (2009) (Supplemental Fig. 2). QTLs responsible for ripening percentage were detected via the composite interval mapping procedure of QTL Cartographer v. 2.5. Significance of LOD (logarithm of odds) scores was calculated using 1000 permutation tests (5% level).

Statistical analysis

Genotypic differences in ripening percentage were statistically evaluated by Student’s t-test and Dunnett’s test using JMP v. 12 software (SAS Institute, Cary, NC, USA). Multiple regression analysis was also performed using JMP v. 12 software.

Results

Responses of yield and yield components to salt stress

The dry weights of the aboveground parts of plants and grain weight markedly decreased in response to 200 mM NaCl treatment for both OUE812 and OUC613 (Table 1). To visualize salt effects on growth, photographs of plant at grain filling stage grown in 1/2000 a pot with salt stress were shown in Supplemental Fig. 3. The reduction in grain weight, however, was significantly smaller in OUE812 than in OUC613. As a result, the harvest index (grain weight/aboveground weight) was markedly lower for OUC613 compared to OUE812. The major reason for the marked reduction in grain weight in OUC613 was the reduction in RP (5.1% in OUC613 vs. 79.8% in OUE812). Under 150 mM NaCl treatment, the reduction in dry weight of aboveground parts and grain weight was small in both lines compared with the values obtained under 200 mM NaCl treatment (Table 2). Grain weight was significantly reduced in OUC613 compared with OUE812; this reduction was larger than the reduction in dry weight of aboveground parts due to a marked reduction in RP in OUC613 (Table 2). The RP was lower in OUC613 (60.8%) than in OUE812 (83.8%).

Table 1 Dry weight of the aboveground parts and grain weight, panicle number per plant, and other yield components of the main stems of plants treated with 200 mM NaCl in 2013

NaCl concentration (mM)	Accession	Aboveground weight (g)	Total grain weight (g/plant)	Harvest index (%)	Number of panicles (/plant)	Number of spikelets (/panicle)	Ripening percentage (%)	1000-grain weight (g)
0	OUC613	13.0 ± 4.2	5.9 ± 2.0	44.9 ± 0.6	7.7 ± 2.1	30.7 ± 1.2	100.0 ± 0.0	44.3 ± 1.5
	OUE812	7.7 ± 3.8	4.3 ± 2.2	54.9 ± 3.6	3.0 ± 1.0	39.0 ± 10.4	87.1 ± 5.5	45.6 ± 5.8
		n.s.a	n.s.	**a	*a	n.s.	*	n.s.
200	OUC613	1.2 ± 0.4	0.2 ± 0.1	16.5 ± 8.9	1.7 ± 0.6	18.0 ± 3.7	5.1 ± 4.7	37.2 ± 3.1
		(9)b	(3)	(37)	(22)	(59)	(5)	(84)
	OUE812	1.9 ± 0.5	1.0 ± 0.2	51.0 ± 0.9	1.7 ± 0.6	23.0 ± 1.8	79.8 ± 1.0	38.0 ± 4.7
		(24)	(22)	(93)	(56)	(59)	(92)	(83)
		n.s.	**	**	n.s.	n.s.	***a	n.s.

Mean ± SD (n = 3).

^a  Asterisks indicate significant differences between cultivars: n.s. indicates no significant difference, *, **, and *** indicate p < 0.05, 0.01, and 0.001, respectively (Student’s t-test).

^b  Values in parentheses represent percentages relative to plants treated with 0 mM NaCl.

Table 2 Dry weight of the aboveground parts and grain weight, panicle number per plant, and other yield components of the main stems of plants treated with 150 mM NaCl in 2014

NaCl concentration (mM)	Accession	Aboveground weight (g)	Total grain weight (g/plant)	Harvest index (%)	Number of panicles (/plant)	Number of spikelets (/panicle)	Ripening percentage (%)	1000-grain weight (g)
0	OUC613	9.9 ± 2.0	5.6 ± 1.0	56.9 ± 4.2	6.3 ± 0.6	28.7 ± 6.7	89.4 ± 3.6	52.1 ± 2.1
	OUE812	7.4 ± 0.9	5.2 ± 0.9	69.2 ± 3.8	4.0 ± 0.0	34.3 ± 2.3	94.3 ± 2.6	50.8 ± 5.1
		n.s.a	n.s.	*a	**a	n.s.	n.s.	n.s.
150	OUC613	2.9 ± 0.2	1.1 ± 0.1	37.6 ± 2.2	3.7 ± 0.6	22.0 ± 2.0	60.8 ± 3.7	46.4 ± 0.7
		(29)b	(20)	(66)	(58)	(77)	(68)	(89)
	OUE812	2.9 ± 0.4	1.7 ± 0.2	58.7 ± 1.9	2.3 ± 0.6	24.0 ± 3.0	83.8 ± 10.3	46.7 ± 0.5
		(39)	(33)	(85)	(58)	(70)	(89)	(92)
		n.s.	**	***	*	n.s.	*	n.s.

Mean ± SD (n = 3).

^a  Asterisks indicate significant differences between cultivars: n.s. indicates no significant difference, *, **, and *** indicate p < 0.05, 0.01, and 0.001, respectively (Student’s t-test).

^b  Values in parentheses represent percentages relative to plants treated with 0 mM NaCl.

RP of plants produced by artificial pollination

For OUC613, the RP was 88% in plants obtained from crosses between female and male plants grown under the control condition (0 × 0 mM plants) (Fig. 1A). The RP in plants derived from a cross between female plants grown in 150 mM NaCl and male plants grown in control conditions (150 × 0 mM plants) decreased to 76%, but the reduction was not large enough to induce a significant difference in RP between 0 × 0 mM and 150 × 0 mM plants. The RP decreased to 65% in the 0 × 150 mM plants, and to 50% in the 150 × 150 mM plants. The reduction in RPs in plants with pollen derived from plants grown in 150 mM NaCl caused a significantly lower RP than that of 0 × 0 mM plants. These results indicate that the marked reduction in RP in OUC613 plants treated with 150 mM NaCl could mainly be attributed to pollen sterility. By contrast, for OUE812, the RP was 96% in 0 × 0 mM plants (Fig. 1B), and no significant differences in RP were observed among plants from crosses between female and male plants grown under any conditions.

Fig. 1

Ripening percentage (RP) of artificially pollinated plants. ‘0’ and ‘150’ on the x-axis represent plants grown without additional NaCl and with 150 mM NaCl, respectively. ♀ and ♂ represent female and male plants, respectively. Plants of 0♀ × 0♂ (control) are represented by black bars. Plants with either or both parents were treated with 150 mM NaCl are represented by white bars. (A) Emasculated OUC613 plants were pollinated with pollen from other OUC613 plants. (B) Emasculated OUE812 plants were pollinated with pollen from other OUE812 plants. Error bars represent SD (n = 8). **p < 0.01 and ***p < 0.001 vs. the control, respectively (Dunnett’s test).

QTL analysis of RRP

Even under 0 mM NaCl treatment, some differences in RP were observed among RILs. The RP for RILs grown under control conditions was 89.1 ± 7.6 and 84.1 ± 10.0 in 2013 and 2014, respectively. Therefore, we used RRP for each RIL instead of absolute values of RP in the subsequent QTL analysis. RRP showed a continuous segregation pattern (Fig. 2, Supplemental Fig. 4).

Fig. 2

Histogram of average values of relative ripening percentage (RRP) in F₆-RILs treated with 200 mM NaCl with three replications. White and black bars indicate the number of lines used to produce the salt-resistant bulk in 2013 (higher RRP group) and the salt-sensitive bulk (lower RRP group), respectively. White, black, and gray arrows represent average RRP values for OUE812, OUC613, and RILs treated with 200 mM NaCl, respectively.

In QTL-seq analysis of the F₆ RILs treated with 200 mM NaCl in 2013, after two rounds of sequencing runs using mixed libraries from the HF and LF bulks via Illumina MiSeq, we obtained approximately 30 million reads (amounting to 4.5 Gbp) and ~37 million reads (5.6 Gbp), respectively. After applying the QTL-seq pipeline, 4,384,475 paired reads from the HF bulk and 4,660,760 reads from the LF bulk were mapped to PRS (Table 3). The average depth of the HF and LF bulks was 35.9 and 34.7, respectively (Supplemental Fig. 5). After quality filtering at a level of Coval = 6, we obtained 60,331 SNPs in the HF bulk and 56,414 SNPs in the LF bulk (Table 3). We obtained a ΔSNP-index value from each SNP index between the HF and LF bulks (Supplemental Fig. 6), which was calculated by sliding window analysis and plotted onto the PRS. Fig. 3 shows a plot of the ΔSNP-index values obtained after setting the sliding window size to 150 kb, window sliding size to 50 kb, and Coval to 6. A ΔSNP-index value beyond the 5% level of statistical significance appeared at 2.5 Mb on chromosome 2H of OUC613 PRS (approximately 54 cM), corresponding to SEQ (detected by QTL-seq)-2H in the linkage map. The peak ΔSNP-index was observed at 4 Mb on chromosome 1H of OUC613 PRS (approximately 55 cM), corresponding to SEQ-1H, although it did not exceed the confidence interval line.

Table 3 Number of aligned reads and SNPs detected by exome QTL-seq in F₆-RILs treated with 200 mM NaCl in 2013

	Librarya	1H	2H	3H	4H	5H	6H	7H	Total
No. of aligned reads	HF	570,678	752,937	665,739	438,946	714,920	554,156	687,099	4,384,475
	LF	603,842	798,212	710,168	469,610	760,576	586,498	731,854	4,660,760
No. of SNPs	HF	7,422	10,880	9,226	5,202	9,790	8,029	9,782	60,331
	LF	7,158	10,188	8,524	4,957	9,242	7,123	9,222	56,414

^a  HF and LF indicate bulks from the groups of RILs that showed the highest and lowest RRP, respectively.

Fig. 3

Plots of ΔSNP-index values for each chromosome generated by QTL-seq analysis. The red line indicates the mean ΔSNP-index calculated by sliding window analysis. Yellow lines indicate a 95% of confidence interval. The black arrow shows the peak that crossed the upper confidence interval line, referred to as SEQ-2H. The white arrow shows the peak that did not cross the upper confidence interval line, referred to as SEQ-1H.

We also identified QTLs associated with RRP via composite interval mapping from plants treated with 200 mM NaCl for the F₆ RILs in 2013 and 150 mM NaCl for the F₇ RILs in 2014. Fig. 4 shows the LOD scores for RRP in plants treated with 200 mM NaCl (A) and the additive effects (B). One distinguished QTL where the OUE812 allele enhanced RRP was detected on chromosome 2H (53.5–62.8 cM) (Table 4). We designated this QTL CAR (detected via QTL Cartographer)-2Hb. Two effective regions were also detected on chromosomes 1H (57.0–65.5 cM) and 2H (6.5–26.5 cM), although the LOD scores were not significant. These QTLs are referred to hereafter as CAR-1H and CAR-2Ha, respectively.

Fig. 4

QTL analysis of RRP under salt stress in F₆-RILs. (A) LOD scores for RRP in RILs treated with 200 mM NaCl in 2013. Horizontal line shows the threshold determined by 1000 permutation tests. The QTLs indicated by white, gray and black arrows are referred to as CAR-1H, CAR-2Ha, and CAR-2Hb, respectively. (B) Additive effects. Positive values indicate that the allele from OUE812 enhanced the trait.

Table 4 Quantitative trait loci (QTL) for relative ripening percentage (RRP) using F₆-RILs treated with 200 mM NaCl in 2013

QTL	Chromosome	Positiona (cM)	Nearest marker	LOD	AEb	PVEc (%)
CAR-1H	1H	60.5	2057-1122	2.6	6.2	9.1
CAR-2Ha	2H	20.0	2582-1767	2.5	6.7	7.5
CAR-2Hb	2H	52.5	4630-1036	6.9	13.1	24.8

^a  Peak position of LOD value.

^b  Additive effect of OUE812 allele.

^c  Percentage of total phenotypic variance explained by each QTL.

Fig. 5 shows the LOD scores for RRP under 150 mM NaCl treatment (A) and the additive effects (B). We identified two QTLs that enhance RRP. One of these QTLs, where an allele of OUE812 enhanced RRP, was detected on chromosome 2H (53.5–62.8 cM) (Table 5). This QTL is the same as CAR-2Hb, which was detected in plants treated with 200 mM NaCl. The other QTL, where an allele of OUC613 enhanced RRP, was detected on chromosome 4H (87.5–92.4 cM), which we designated CAR-4H. An effective region was also detected on chromosome 2H (6.5–26.6 cM), although the LOD score was not significant. This QTL is the same as CAR-2Ha, which we detected in plants under 200 mM NaCl treatment. The locations of CAR-1H and CAR-2Hb are quite similar to those of SEQ-1H and SEQ-2H, respectively, which were detected by QTL-seq.

Fig. 5

QTL analysis of RRP under salt stress in F₇-RILs. (A) LOD scores for RRP in RILs treated with 150 mM NaCl in 2014. Horizontal line shows the threshold determined by 1000 permutation tests. The QTLs indicated by white, gray and black arrows are referred to as CAR-2Ha, CAR-2Hb, and CAR-4H, respectively. (B) Additive effects. Positive values indicate that the allele from OUE812 enhanced the trait.

Table 5 Quantitative trait loci (QTL) for relative ripening percentage (RRP) using F₇-RILs treated with 150 mM NaCl in 2014

QTL	Chromosome	Positiona (cM)	Nearest markers	LOD	AEb	PVE c (%)
CAR-2Ha	2H	15.0	2582-1767	1.8	6.9	9.1
CAR-2Hb	2H	52.5	4630-1036	3.0	7.4	10.1
CAR-4H	4H	85.8	4039-1686	4.3	−9.3	16.5

^a  Peak position of LOD value.

^b  Additive effect of OUE812 allele.

^c  Percentage of total phenotypic variance explained by each QTL.

QTL analysis of relative dry weights of the aboveground parts

The analysis of QTL for dry weights of aboveground parts was conducted via the composite interval mapping in the F₆ RILs treated with 200 mM NaCl in 2013 and the F₇ RILs treated with 150 mM NaCl in 2014. In the 200 mM treatment, relative dry weights of aboveground parts (dry weight of plants with 200 mM NaCl treatment/dry weight of control plants) were 24.5%, 9.2% and 19.3% in OUE812, OUC613 and RILs (average), respectively. Two QTLs were detected (Supplemental Table 1); One was on chromosome 3H, where an allele of OUE812 increased dry weight and the other was on chromosome 5H, where an allele of OUC613 increased dry weight. In the 150 mM treatment, relative dry weights of aboveground parts were 36.7%, 31.9% and 43.3% in OUE812, OUC613 and RILs (average), respectively. Any QTL was not detected in the plants with 150 mM NaCl treatment.

Multiple regression analysis of RRP using the newly detected QTLs

To investigate the contributions of the newly detected QTLs and their interactions, we performed multiple regression analysis of the QTLs detected by composite interval mapping. In plants under 200 mM NaCl treatment, the coefficients were significant at the 5% and 10% levels for CAR-2Hb and CAR-1H, respectively (Table 6). The coefficient for CAR-2Ha was not significant. No significant interaction was detected between loci. In plants under 150 mM NaCl treatment, the coefficients were significant at the 10% and 5% levels for CAR-2Ha and CAR-2Hb, respectively (Table 7), whereas the coefficient for CAR-4H was not significant. No significant interaction was detected among loci. Here, we designated the CAR-2Hb to qRP-2Hb as an allele of OUE812, which maintain high RRP in plants with 200 mM and 150 mM NaCl treatments and was detected via QTL-seq as well as composite interval mapping.

Table 6 Multiple regression analysis of QTLs detected by composite interval mapping for enhancing the relative ripening percentage (RRP) in F₆-RILs treated with 200 mM NaCl in 2013^a

	Region (cM)	Coefficient	SE	t-value	p-value	R2
Intercept		46.3	5.0	9.3	<0.001	0.42
CAR-1H	57.0–65.5	11.0	5.4	2.0	0.051
CAR-2Ha	6.5–26.5	7.0	5.9	1.2	0.247
CAR-2Hb	53.5–62.8	20.9	5.8	3.6	<0.05

^a  RILs carrying the QTL region homozygous for OUE812 or OUC613 were used in the analysis.

Table 7 Multiple regression analysis of QTLs detected by composite interval mapping for enhancing the relative ripening percentage (RRP) in F₇-RILs treated with 150 mM NaCl in 2014

	Region (cM)	Coefficient	SE	t-value	p-value	R2
Intercept		64.6	6.4	10.0	<0.001	0.35
CAR-2Ha	6.5–26.6	12.9	6.3	2.0	0.0503
CAR-2Hb	53.5–62.8	17.8	6.8	2.6	<0.05
CAR-4H	87.5–92.4	6.9	6.2	1.1	0.2742

Genes related to RP

Exome QTL-seq revealed that a QTL that helps to maintain high RP under salt stress is located on chromosome 2H. This significant peak region was detected only at one point in a sliding window during QTL-seq analysis, with the significant region covering a 150-kb area. This region includes 90 genes (Supplemental Table 2), based on IPK Barley BLAST Server analysis (http://webblast.ipk-gatersleben.de/barley_ibsc/). Many genes located in the QTL region on chromosome 2H are highly homologous to those on chromosome 7 in rice. Two of these genes encode proteins associated with male sterility, namely AK357304 and MLOC_12120.2. AK357304 encodes an F-box protein associated with grain fertility under low temperature stress in rice and male sterility in Arabidopsis thaliana (Kim et al. 2010, Saito et al. 2010). Specifically, this protein regulates anther size in rice (Saito et al. 2010) and pollen maturation in Arabidopsis (Kim et al. 2010). A homolog (LOC_Os07g42590) of this gene is located on chromosome 7 in rice (Supplemental Table 2). MLOC_12120.2 encodes a TIFY protein, which is homologous to a jasmonate ZIM-domain protein in rice (LOC_Os07g42370). Jasmonate ZIM-domain proteins have been shown to regulate pollen maturation, anther dehiscence, and filament elongation in Arabidopsis (Song et al. 2011).

Discussion

Grain fertility significantly decreases in response to abiotic stress (Barnabas et al. 2008, Boyer and Westgate 2004, Saini and Westgate 2000). For instance, low or cool temperature stress (Hayase et al. 1969, Nishiyama 1995) and high temperature stress (Nishiyama 1995) reduce grain fertility in rice, wheat (Saini and Aspinal 1982), barley (Sakata et al. 2000), and sorghum (Sorghum bicolor; Brooking 1976), and drought stress reduces grain fertility in wheat (Bingham 1966, Saini and Aspinal 1981), rice (O’Toole and Namuco 1983), and maize (Zea mays; Zinselmeier et al. 1999). Although salt stress also markedly reduces grain yield (Akbar et al. 1972, Francious et al. 1984, Kaddah and Fakhry 1961, Khatun et al. 1995, Richards et al. 1987), little analytical research has been performed on yield components under salt stress. We previously showed that the varietal differences in grain weight detected in barley are mainly due to their differences in RP (Hirasawa et al. 2017). In the current study, we confirmed that the differences in grain weight between barley cultivars result from differences in RP in plants under salt stress (Tables 1, 2).

Analysis of plants obtained by artificial pollination indicated that the RP was mainly determined by pollen sterility (Fig. 1), which might be a major cause of the reduction in ripened grain number in barley under salt stress conditions. Indeed, abiotic stresses such as high and low temperature stress and drought stress cause male sterility rather than female sterility (Hayase et al. 1969, Satake and Yoshida 1978). Pollen sterility is caused by a reduction in the number of mature pollen grains in rice under cool temperature stress (Ito et al. 1970) and in wheat under drought stress (Saini and Aspinall 1981), by a reduction in the number of pollen grains on the stigma in rice under high temperature and drought stress (Ekanayake et al. 1990, Prasad et al. 2006, Rang et al. 2011, Satake and Yoshida 1978), and by a reduction in the germination and elongation rates of pollen in rice under high temperature (Satake and Yoshida 1978) and drought stress (Ekanayake et al. 1990) and in maize under salt stress (Dhingra and Varghese 1985). The results of artificial pollination suggest that the reduction in RP of OUC613 under salt stress might be caused mainly by the reduction in the germination and elongation rate of pollen because excess number of pollen was pollinated. Sakata et al. (2000, 2010) showed that in barley, high temperature stress affects the development of pollen mother cells and microspores due to a reduction in auxin levels. On the other hand, our artificial pollination experiment also suggested that the reduction in RP in response to salt stress might have been partially due to the effects of this treatment on the pistil (Fig. 1). Ovary abortion occurs due to a deficiency in photosynthate levels in maize under water stress (Boyer and Westgate 2004). Pollen sterility from poor anther dehiscence and pollen maturation as well as ovary abortion also might underlie the reduction in grain fertility of OUC613 under salt stress (Ekanayake et al. 1990, Satake and Yoshida 1978).

We identified a QTL with marked effects on maintaining high RP in plants treated with 150 mM and 200 mM NaCl through QTL-seq and QTL interval mapping analyses (Figs. 3–5, Tables 4, 5). The qRP-2Hb allele in OUE 812 might play a distinct role in maintaining pollen fertility under salt stress conditions. The mechanisms underlying pollen sterility in rice at the booting stage in response to cool temperature stress have previously been investigated (Nishiyama 1984, 1995, Oliver et al. 2005, Satake 1976). Although many QTLs for cool temperature damage at the booting stage were revealed (Kuroki et al. 2007, Mori et al. 2011, Saito et al. 2001, Suh et al. 2010), the detailed functions of these QTLs have not yet been clarified. Few studies have focused on identifying QTLs and genes that help to maintain pollen fertility under other abiotic stresses (Xiao et al. 2011). In this study, we identified qRP-2Hb as a QTL that helps maintain grain fertility in plants under salt stress, which might maintain pollen fertility under salt stress. Although both water stress and ion toxicity damage plants under salt stress (Munns and Tester 2008) and QTLs for excluding Na⁺ in plants have been detected (Rivandi et al. 2011, Shavrukov et al. 2010), neither the concentrations of Na⁺ and Cl⁻ in spike organs nor leaf water potential significantly differ between OUC613 plants grown in the presence or absence of additional NaCl (Hirasawa et al. 2017). Perhaps qRP-2Hb is not involved in regulating the balance of salt and/or water, particularly in tissues in the spike (Munns and Rawson 1999). In addition, qRP-2Hb was completely different from the QTL for dry weight of aboveground parts under salt stress (Supplemental Table 1) although there was a similar cultivar difference in dry weight to that in RP. This result supports that qRP-2Hb was associated with pollen fertility. The mechanisms underlying pollen sterility and how qRP-2Hb helps prevent pollen sterility remain to be investigated.

CAR-4H, a QTL identified in OUC613 that enhances RRP, was identified only under 150 mM NaCl conditions, although it was not significant, according to multiple regression analysis (Table 7). On the other hand, CAR-1H was identified in OUE812 as maintaining RRP under 200 mM NaCl treatment, but not under 150 mM NaCl conditions. These results suggest that the expression of some genes might depend on the degree of salt stress, or that gene expression levels differ during certain stages of pollen development in response to salt stress; however, the mechanisms underlying pollen sterility remain unknown. Indeed, similar observations were reported in previous QTL analyses of rice in response to salt stress (Wang et al. 2012a, 2012b); different QTLs associated with seedling height, N/K ratio in the root, and shoot Na⁺ concentration were observed under salt stress.

QTLs associated with maintaining high grain fertility under salt stress with both strong and weak effects were detected by both QTL-seq and composite interval mapping. This result suggests that genes associated with pollen fertility are present in multiple regions of the genome and that exome QTL-seq is an effective method for detecting links between QTLs and polymorphisms in various genetic regions. Our results support the finding that exome QTL-seq is an effective method for detecting QTLs in barley (Hisano et al. 2017). One advantage of QTL-seq is that it does not require the development of DNA markers and marker genotyping (Takagi et al. 2013). In addition, we can obtain SNP information between two cultivars and narrow down the QTL region using cleaved amplified polymorphic sequence (CAPS) markers based on SNP information obtained by QTL-seq.

To date, genes involved in maintaining high fertility under salt stress have not been identified in barley, rice, and other crops. Many genes in the QTL region on chromosome 2H (150 kb of PES) revealed by QTL-seq is highly homologous to genes in a region on chromosome 7 in rice (Supplemental Table 2). Several loci associated with fertility under high and low temperature stress have been identified on chromosome 7 in rice (Suh et al. 2010, Yamaguchi et al. 1997, Ye et al. 2012). However, none of them are homologous to the genes we identified in barley with annotation related to fertility. Many genes in the QTL region on chromosome 2H is also highly homologous to genes in a region on chromosome 2 in sorghum, but QTLs associated with salt tolerance in sorghum on this chromosome have not been reported to our knowledge. The QTL-seq analysis performed in this study revealed two candidate genes that might regulate grain fertility under salt stress, including genes encoding an F-box protein and a TIFY protein, which is homologous to a jasmonate ZIM-domain protein. They are included in the jasmonate signaling pathway associating with male sterility such as anther dehiscence and the maturation and germination of pollen (Ahmad et al. 2016, Song et al. 2011). Characterizing the expression of these genes under salt stress conditions should be a target of future studies, as should narrowing down the QTL region associated with maintaining grain fertility under salt stress and clarifying the mechanisms underlying pollen sterility under salt stress.

Acknowledgments

We are grateful to Ms. Yuka Motoi for technical support. This work was supported in part by a Grant-in-Aid from the Ministry of Education, Culture, Sport, Science, and Technology (MEXT) of Japan (no. 16K14834) and by MEXT as part of a Joint Research Program implemented at the Institute of Plant Science and Resources, Okayama University, Japan (no. 2637, 2831). Barley seeds were provided by the National Bio-Resource Project of Barley, MEXT of Japan. Computations were partially performed on the NIG supercomputer at the ROIS National Institute of Genetics.

Literature Cited

•  Ahmad,  P.,  S.  Rasool,  A.  Gul,  S.A.  Sheikh,  N.A.  Akram,  M.  Ashraf,  A.M.  Kazi and  S.  Gucel (2016) Jasmonates: Multifunctional roles in stress tolerance. Front. Plant Sci. 7: 813.
•  Akbar,  M.,  T.  Yabuno and  S.  Nakao (1972) Breeding for saline-resistant varieties of rice I. Variability for salt tolerance among some rice varieties. Japan. J. Breed. 22: 277–284.
•  Barnabas,  B.,  K.  Jager and  A.  Feher (2008) The effect of drought and heat stress on reproductive processes in cereals. Plant Cell Environ. 31: 11–38.
•  Bingham,  J. (1966) Varietal response in wheat to water supply in the field, and male sterility caused by a period of drought in a glasshouse experiment. Ann. Appl. Biol. 57: 365–377.
•  Boyer,  J.S. and  M.E.  Westgate (2004) Grain yields with limited water. J. Exp. Bot. 55: 2385–2394.
•  Brooking,  I.R. (1976) Male sterility in Sorghum bicolor (L.) Moench induced by low night temperature. I. Timing of the stage of sensitivity. Aust. J. Plant Physiol. 3: 589–596.
•  Close,  T.J.,  P.R.  Bhat,  S.  Lonardi,  Y.  Wu,  N.  Rostoks,  L.  Ramsay,  A.  Druka,  N.  Stein,  J.T.  Svensson,  S.  Wanamaker et al. (2009) Development and implementation of high-throughput SNP genotyping in barley. BMC Genomics 10: 582–594.
•  Dhingra,  H.R. and  T.M.  Varghese (1985) Effect of salt stress on viability, germination and endogenous levels of some metabolites and ions in maize (Zea mays L.) pollen. Ann. Bot. 55: 415–420.
•  Ekanayake,  I.J.,  P.L.  Steponku and  S.K.D.  Datta (1990) Sensitivity of pollination to water deficits at anthesis in upland rice. Crop Sci. 30: 310–315.
• FAO (2009) How to feed the world in 2050. http://www.fao.org/fileadmin/templates/wsfs/docs/expert_paper/How_to_Feed_the_World_in_2050.pdf
• FAOSTAT (2012) http://faostat.fao.org/
•  Francious,  L.E.,  T.  Donovan and  E.V.  Maas (1984) Salinity effects on seed yield, growth, and germination of grain sorghum. Agron. J. 76: 741–744.
•  Hayase,  H.,  T.  Satake,  I.  Nishiyama and  N.  Ito (1969) Male sterility caused by cooling treatment at the meiotic stage in rice plants: II. The most sensitive stage to cooling and the fertilizing ability of pistils. Jpn. J. Crop Sci. 38: 706–711.
•  Hirasawa,  T.,  K.  Sato,  M.  Yamaguchi,  R.  Narita,  A.  Kodama,  S.  Adachi,  T.  Ookawa and  K.  Sato (2017) Differences in dry matter production, grain production, and photosynthetic rate in barley cultivars under long-term salinity. Plant Prod. Sci. 20: 288–299.
•  Hisano,  H.,  K.  Sakamoto,  H.  Takagi,  R.  Terauchi and  K.  Sato (2017) Exome QTL-seq maps monogenic locus and QTLs in barley. BMC Genomics 18: 125.
• International Barley Sequencing Consortium (IBSC) (2012) A physical, genetic and functional sequence assembly of the barley genome. Nature 491: 711–716.
•  Ito,  N.,  H.  Hayase,  T.  Satake and  I.  Nishiyama (1970) Male sterility caused by cooling treatment the meiotic stage in rice plants III. Male abnormalities at anthesis. Jpn. J. Crop Sci. 39: 60–64.
•  Kaddah,  M.T. and  S.I.  Fakhry (1961) Tolerance of egyptian rice to salt: 1. salinity effects when applied continuously and intermittently at different stages of growth after transplanting. Soil Sci. 91: 113–120.
•  Khatun,  S.,  C.A.  Rizzo and  T.J.  Flowers (1995) Genotypic variation in the effect of salinity on fertility in rice. Plant Soil 173: 239–250.
•  Kim,  O.K.,  J.H.  Jung and  C.M.  Park (2010) An Arabidopsis F-box protein regulates tapetum degeneration and pollen maturation during anther development. Planta 232: 353–366.
•  Kosugi,  K.,  S.  Natsume,  K.  Yoshida,  D.  MacLean,  L.  Cano,  S.  Kamoun and  R.  Terauchi (2009) Coval: Improving alignment quality and variant calling accuracy for next-generation sequencing data. PLoS ONE 8: e75402.
•  Kuroki,  M.,  K.  Saito,  S.  Matsuba,  N.  Yokogami,  H.  Shimizu,  I.  Ando and  Y.  Sato (2007) A quantitative trait locus for cold tolerance at the booting stage on rice chromosome 8. Theor. Appl. Genet. 115: 593–600.
•  Mano,  Y. and  K.  Takeda (1997) Mapping quantitative trait loci for salt tolerance at germination and the seedling stage in barley (Hordeum vulgare L.). Euphytica 94: 263–272.
•  Masher,  M.,  T.A.  Richmond,  D.J.  Gerhardt,  A.  Himmelbach,  L.  Clissold,  D.  Sampath,  S.  Ayling,  B.  Steuernagel,  M.  Pfeifer,  M.  D’Ascenzo et al. (2013) Barley whole exome capture: a tool for genomic research in the genus Hordeum and beyond. Plant J. 76: 494–505.
•  Mori,  M.,  K.  Onishi,  Y.  Tokizono,  H.  Shinada,  T.  Yoshimura,  Y.  Numao,  H.  Miura and  T.  Sato (2011) Detection of a novel quantitative trait locus for cold tolerance at the booting stage derived from a tropical japonica rice variety Silewah. Breed. Sci. 61: 61–68.
•  Munns,  R. and  H.M.  Rawson (1999) Effect of salinity on salt accumulation and reproductive development in the apical meristem of wheat and barley. Aust. J. Plant Physiol. 26: 459–464.
•  Munns,  R. (2005) Genes and salt tolerance: bringing them together. New Phytol. 167: 645–663.
•  Munns,  R.,  R.A.  James and  A.  Lauchii (2006) Approaches to increasing the salt tolerance of wheat and other cereals. J. Exp. Bot. 57: 1025–1043.
•  Munns,  R. and  M.  Tester (2008) Mechanisms of salinity tolerance. Annu. Rev. Plant Biol. 59: 651–681.
•  Nishiyama,  I. (1984) Climatic influence on pollen formation and fertilization. In:  Tsunoda,  S. and  N.  Takahashi (eds.) Biology of Rice, Japan. Sci. Soc. Press, Tokyo/Elsevier, Amsterdam, pp. 153–171.
•  Nishiyama,  I. (1995) Damage due to extreme temperatures. In:  Matsuo,  T.,  K.  Kumazawa,  R.  Ishii,  K.  Ishihara and  H.  Hirata (eds.) Science of the Rice Plant Vol. 2: Physiology, Food and Agriculture Policy Research Center, Tokyo, pp. 769–812.
•  Oliver,  S.N.,  J.T.  Van Dongen,  S.C.  Algfed,  E.A.  Mamun,  X.  Zhao,  H.S.  Saini,  S.F.  Fernandes,  C.L.  Blanchard,  B.G.  Sutton,  P.  Geigenberger et al. (2005) Cold-induced repression of the rice anther-specific cell wall invertase gene OSINV4 is correlated with sucrose accumulation and pollen sterility. Plant Cell Environ. 28: 1534–1551.
•  O’Toole,  J.C. and  O.S.  Namuco (1983) Role of panicle exsertion in water stress induced sterility. Crop Sci. 23: 1093–1097.
•  Prasad,  P.V.V.,  K.J.  Boote,  L.H.  Allen,  J.E.  Sheehy and  J.M.G.  Thomas (2006) Species, ecotype and cultivar differences in spikelet fertility and harvest index of rice in response to high temperature stress. Field Crops Res. 95: 398–411.
•  Rang,  Z.W.,  S.V.K.  Jagadish,  Q.M.  Zhou,  P.Q.  Craufurd and  S.  Heuer (2011) Effect of high temperature and water stress on pollen germination and spikelet fertility in rice. Field Crops Res. 70: 58–65.
•  Rawson,  H.M.,  R.A.  Richards and  R.  Munns (1988) An examination of selection criteria for salt tolerance in wheat, barley and triticale genotypes. Aust. J. Agric. Res. 39: 759–772.
•  Richards,  R.A.,  C.W.  Dennett,  C.O.  Qualset,  E.  Epstein,  J.D.  Norlyn and  M.D.  Winslow (1987) Variation in yield of grain and biomass in wheat, barley, and triticale in a salt-affected field. Field Crops Res. 15: 277–287.
•  Rivandi,  J.,  J.  Miyazaki,  M.  Hrmova,  M.  Pallotta,  M.  Tester and  N.C.  Collins (2011) A SOS3 homologue maps to HvNax4, a barley locus controlling an environmentally sensitive Na⁺ exclusion trait. J Exp. Bot. 62: 1201–1216.
•  Royo,  A. and  R.  Aragues (1999) Salinity-yield response functions of barley genotypes assessed with a triple line source sprinkler system. Plant Soil 209: 9–20.
•  Saini,  H.S. and  D.  Aspinall (1981) Effect of water deficit on sporogenesis in wheat (Triticum aestivum L.). Ann. Bot. 48: 623–633.
•  Saini,  H.S. and  D.  Aspinall (1982) Abnormal sporogenesis in wheat (Triticum aestivum L.) induced by short periods of high temperature. Ann. Bot. 49: 835–846.
•  Saini,  H.S. and  M.E.  Westgate (2000) Reproductive development in grain crops during drought. Adv. Agron. 68: 59–96.
•  Saisho,  D. and  K.  Takeda (2011) Barley: Emergence as a new research material of crop science. Plant Cell Physiol. 52: 724–727.
•  Saito,  K.,  K.  Miura,  K.  Nagano,  Y.  Hayano-Saito,  H.  Araki and  A.  Kato (2001) Identification of two closely linked quantitative trait loci for cold tolerance on chromosome 4 of rice and their association with anther length. Theor. Appl. Genet. 103: 862–868.
•  Saito,  K.,  Y.  Hayano-Saito,  M.  Kuroki and  Y.  Sato (2010) Map-based cloning of the rice cold tolerance gene Ctb1. Plant Sci. 179: 97–102.
•  Sakata,  T.,  H.  Takahashi,  I.  Nishiyama and  A.  Higashitani (2000) Effects of high temperature on the development of pollen mother cells and microspores in barley Hordeum vulgare L. J. Plant Res. 113: 395–402.
•  Sakata,  T.,  T.  Oshino,  S.  Miura,  M.  Tomabechi,  Y.  Tsunaga,  N.  Higashitani,  Y.  Miyazawa,  H.  Takahashi,  M.  Watanabe and  A.  Higashitani (2010) Auxins reverse plant male sterility caused by high temperatures. Proc. Natl. Acad. Sci. USA 107: 8569–8574.
•  Satake,  T. (1976) Determination of the most sensitive stage to sterile-type cool injury in rice plants. Res. Bull. Hokkaido Natl. Agric. Exp. Sta. 113: 1–43.
•  Satake,  T. and  S.  Yoshida (1978) High temperature-induced sterility in indica rices at flowering. Jpn. J. Crop Sci. 47: 6–17.
•  Sato,  K.,  T.J.  Close,  P.  Bhat,  M.  Munoz-Amatriain and  G.J.  Muehlbauer (2011) Single nucleotide polymorphism mapping and alignment of recombinant chromosome substitution lines in barley. Plant Cell Physiol. 52: 728–737.
•  Shavrukov,  Y.,  K.G.  Narendra,  J.  Miyazaki,  M.N.  Baho,  K.J.  Chalmers,  M.  Tester,  P.  Langridge and  N.C.  Collins (2010) HvNax3—a locus controlling shoot sodium exclusion derived from wild barley (Hordeum vulgare ssp. spontaneum). Funct. Integr. Genomics 10: 277–291.
•  Slavich,  P.G.,  B.J.  Read and  B.R.  Cullis (1990) Yield response of barley germplasm to field variation in salinity quantified using the EM-38. Aust. J. Exp. Agric. 30: 551–556.
•  Song,  S.,  T.  Qi,  H.  Huang,  Q.  Ren,  D.  Wu,  C.  Chang,  W.  Peng,  Y.  Liu,  J.  Peng and  D.  Xie (2011) The Jasmonate-ZIM domain proteins interact with the R2R3-MYB transcription factors MYB21 and MYB24 to affect jasmonate-regulated stamen development in Arabidopsis. Plant Cell 23: 1000–1013.
•  Suh,  J.P.,  J.U.  Jeung,  J.I.  Lee,  Y.H.  Choi,  J.D.  Yea,  P.S.  Virk,  D.J.  Mackill and  K.K.  Jena (2010) Identification and analysis of QTLs controlling cold tolerance at the reproductive stage and validation of effective QTLs in cold-tolerant genotypes of rice (Oryza sativa L.). Theor. Appl. Genet. 120: 985–995.
•  Takagi,  H.,  A.  Abe,  K.  Yoshida,  S.  Kosugi,  S.  Natsume,  C.  Mitsuoka,  A.  Uemura,  H.  Utsushi,  M.  Tamiru,  S.  Takuno et al. (2013) QTL-seq: rapid mapping of quantitative trait loci in rice by whole genome resequencing of DNA from two bulked populations. Plant J. 74: 174–183.
•  Wang,  Z.,  J.  Cheng,  Z.  Chen,  J.  Huang,  Y.  Bao,  J.  Wang and  H.  Zhang (2012a) Identification of QTLs with main, epistatic and QTL × environment interaction effects for salt tolerance in rice seedlings under different salinity conditions. Theor. Appl. Genet. 125: 807–815.
•  Wang,  Z.,  Z.  Chen,  J.  Cheng,  Y.  Lai,  J.  Wang,  Y.  Bao,  J.  Huang and  H.  Zhang (2012b) QTL analysis of Na⁺ and K⁺ concentrations in roots and shoots under different levels of NaCl stress in rice (Oryza sativa L.). PLoS ONE 7: e51202.
•  Win,  K.T.,  A.Z.  Oo,  T.  Hirasawa,  T.  Ookawa and  Y.  Hirata (2011) Genetic analysis of Myanmar Vigna species in responses to salt stress at the seedling stage. Afr. J. Biotechnol. 10: 1615–1624.
•  Xiao,  Y.H.,  Y.  Pan,  L.  Luo,  H.  Demg,  G.  Zhang,  W.  Tang and  L.  Chen (2011) Quantitative trait loci associated with pollen fertility under high temperature stress at flowering stage in rice (Oryza sativa). Rice Sci. 18: 204–209.
•  Xue,  D.,  Y.  Huang,  X.  Zhang,  K.  Wei,  S.  Westcott,  C.  Li,  M.  Chen,  G.  Zhang and  R.  Lance (2009) Identification of QTLs associated with salinity tolerance at late growth stage in barley. Euphytica 16: 187–196.
•  Yamaguchi,  Y.,  R.  Ikeda,  H.  Hirasawa,  M.  Minami and  A.  Ujihara (1997) Linkage analysis of thermosensitive genetic male sterility gene, tms-2 in rice (Oryza sativa L.). Breed. Sci. 47: 371–373.
•  Yamanouchi,  M. (1989) The mechanisms of salinity tolerance in rice plants. Relationships between the varietal difference of salinity tolerance and characteristics of absorption and translocation of sodium ion (2). Jpn. J. Soil Sci. Plant Nutr. 60: 210–219.
•  Ye,  C.,  M.A.  Argayoso,  E.D.  Redoña,  S.N.  Sierra,  M.A.  Laza,  C.J.  Dilla,  Y.  Mo,  M.J.  Thomoson,  J.  Chin,  C.B.  Delaviña et al. (2012) Mapping QTL for heat tolerance at flowering stage in rice using SNP markers. Plant Breed. 131: 33–41.
•  Zhou,  G.,  P.  Johnson,  P.R.  Ryan,  E.  Delhaize and  M.  Zhou (2012) Quantitative trait loci for salinity tolerance in barley (Hordeum vulgare L.). Mol. Breed. 29: 427–436.
•  Zinselmeier,  C.,  B.R.  Jeong and  J.S.  Boyer (1999) Starch and the control of kernel number in maize at low water potentials. Plant Physiol. 121: 25–36.