Introduction

Rice (Oryza sativa L.) is one of the most important food crops in the world. In recent years, with global warming and environmental deterioration, the frequency and range of droughts has increased. Drought as a major challenge results in the reduction of large-scale rice production (Luo 2010). The most effective way to improve drought tolerance of rice is to study physiological responses to drought stress, to screen and identify drought-resistant genes, and to develop drought-tolerant varieties (Luo and Zhang 2001). The response of rice plants to drought stress involves many complex physiological changes and QTLs. To date, more than 800 rice QTLs that respond to drought tolerance have been identified. Among these lines, 338 were associated with root development and tillering under drought stress, 36 were associated with physiological traits (i.e., abscisic acid, osmotic adjustment, and relative water content), and 435 were associated with other traits (http://www.plantstress.com/biotech/index.asp?Flag=1). A water stress experiment was performed on rice at two different periods of the vegetative stage, which identified four drought-sensitive index QTLs using a comparative study of plant height, tiller number, and root thickness (Hemamalini et al. 2000). Further study of root traits revealed 8 drought tolerance QTLs under drought stress (Ali et al. 2000; Kamoshita et al. 2002; Price et al. 2002), while under drought in rice at the seedling and vegetative stages have identified another 18 drought tolerance QTLs (Champoux et al. 1995). Then, 2 novel drought tolerance QTLs were identified at the seedling stage under drought treatment (Teng et al. 2002). As rice has a complex response to drought stress, it is difficult to comprehensively characterize drought resistance by conducting drought treatment experiments only on plants of a particular developmental stage or by studying a limited number of traits under drought stress conditions. However, many drought-responsive QTLs are strongly influenced by the environment, and these QTLs are limited in plant cultivation.

Single segment substitution lines (SSSLs) are ideally suited for QTL identification and have the advantages of line stability and genetic background purity. SSSLs decompose QTLs controlling complex traits into individual Mendelian factors and improve the accuracy of QTL analysis. To date, several rice SSSLs have been developed, and many QTLs underlying traits of biological and economic interest have been detected (Liu et al. 2004, 2008; Ebitani et al. 2005; Xi et al. 2006; Zhu et al. 2009; Teng et al. 2012). In this study, we used SSSLs as a population for the identification and characterization of drought tolerance QTLs in yield-related traits. Through a comparative analysis of QTLs of relative traits under drought stress, drought tolerance QTLs with stable heredity and significant effect were screened to provide a reference for molecular marker-assisted breeding of drought tolerance varieties and identify drought tolerance genes.

Materials and Methods

Experimental materials

A set of single segment substitution lines was derived from a cross of O. sativa indica cv. 9311 as the recipient parent and O. sativa japonica cv. Nipponbare as the donor parent. Our population was composed of 123 lines. The average length of the set lines was 5.58 Mb, covering 63.72% of the rice genome. In this study, 70 substitution lines with productive value were selected for the experimental study.

Field experiment

A drought stress experiment was implemented in an upland field with fertile soil and without shade in winter at Sanya, Hainan. Seeds for the experiment were dried for 48 h at a constant temperature of 50°C followed by 48 h of imbibition, after which they were germinated for 24 h at 30 -35°C and sown. Sublines were sowed evenly in wet seedbeds at 30 days. Thirty plants of each replicate were grown with a single individual per hill in 16.67 × 23.33-cm plots. The experimental field was divided into quarters. Three of these experimental sections contained replicates of the drought treatment, while one-quarter contained a control replicate that experienced regular irrigation. Quarters were isolated from one another using a double layer of ridging film. When the transplanted seedlings were completely green, irrigation was restricted in the treatment plots, while normal irrigation continued in the control plots. This restriction was continued until plant maturation. The 15-15-15NPK compound fertilizer (200 kg/hm²) and urea (150 kg/ hm²) supplement were applied to all plots. Pest and disease prevention were conducted normally for all plots.

The combined ability test for the drought-resistant lines was designed according to NC II. The thermosensitive genic male sterile line Y58S and recessive genic male sterile line ABCG15 were hybridized with SSSLs or 9311. Hybrids of each combination were planted at the Wenjiang experimental farm of Sichuan Agricultural University under normal management. Five plants of each combination were harvested after maturity to investigate yield-related agronomic characters.

Data survey and collection

Days to heading, plant height at maturation of five representative plants was recorded. After maturation, five plants of each line were harvested, threshed, and dried, and the number of effective panicles and spikelets, 1000-grain weight, seed setting rate, and grain yield were scored.

Relative trait values = trait values under drought stress / traits values under irrigation.

Statistical analysis and QTL detection

All statistical analyses were performed using SPSS 7.05 data analysis software. We adopted the method of Liu et al. (2004) to identify QTLs; that is, a t-test was used to analyze the differences between SSSLs and the recipient parent (9311). In this instance, “P ≤0.001” served as the threshold for the existence of QTLs, and QTL additive effects and additive effect contributions were estimated as per Eshed and Zamir (1995).

Additive effect values = (homozygous phenotypic values of SSSLs - phenotypic value of 9311)/2;

Additive effect (%) = (|value of additive effect|/value of control) × 100

QTL substitution mapping methods were performed as per He et al. (2005). The substitution fragment length was calculated according to the method of Young and Tanksley (1989). QTL naming followed the rules of McCouch (2008).

Results

Performance of SSSLs and parents under drought stress

Under drought stress, plant height decreased and days to heading increased. Compared with the control under normal irrigation, the spikelet number per panicle, number of effective panicles, seed setting rate, and grain yield significantly decreased both in the SSSLs and in parents exposed to drought stress (P < 0.01). Grain yield showed the strongest effect followed by plant height, 1000-grain weight, number of effective panicles, seed setting rate, spikelet number per panicle, and days to heading. Differences for drought tolerance were represented for the same trait between SSSLs (Fig. 1). Trait values and relative trait values of SSSLs were continuously distributed, and the variation coefficient of grain yield—the trait that differed the most strongly—was 29.05%, that of relative grain yield was 34.04%, and the transgressive phenomena appeared in SSSL traits. These findings show that the genes controlling related traits and drought tolerance were quantitative trait loci (QTLs).

Comparison of agronomic traits and combining ability of drought-tolerant lines and 9311

Table 1: Agronomic characteristics of hybrids of drought-tolerant lines combined with two male sterile lines

SSSLs (male)	Male sterile lines (female)	Days to heading (d)	Plant height (cm)	Number of effective panicles	Spikelet number per panicle	1000-grain weight (g)	Seed setting rate (%)	Grain yield per plant (g)
9311(CK)	ABCG15	116.3	115.55	6.1	261.31	26.23	72.45	29.94
9311(CK)	Y58S	109.1	110.17	7.0	192.48	26.81	80.07	26.49
X707	ABCG15	117.0	114.67	4.4	419.84	25.17	71.81	33.09
X699	ABCG15	113.3	118.80	5.1	282.23	26.55	80.67	30.47
X705	ABCG15	117.0	116.73	4.3	384.53	26.06	70.10	29.69
X707	Y58S	108.0	108.00	6.8	180.70	26.01	85.41	27.00
X705	Y58S	107.7	108.87	6.7	169.01	27.40	86.90	26.87
X699	Y58S	107.7	114.20	8.5	128.37	26.99	86.93	25.74

Fig. 1: Variation of seven traits of single segment substitution lines and parents under drought stress (D) and irrigation conditions (I). The box plots overlapped by trait values and normal lines of plant height (PH), days to heading (DH), spikelet number per panicle (SN), effective panicles per plant (EP), seed setting rate (SSR), 1000-grain weight (TGW) and grain yield per plant (GY) of SSSLs and parents. ** Significant at the 0.01 probability level. Nipponbare trait data are not presented due to the abnormal seed setting at the Hainan experimental site in winter

The seed setting rate and grain yield per plant are sensitive indexes to drought stress. The relative seed setting rate and relative grain yield per plant under drought stress can indicate the drought resistance of SSSLs. Seven drought-tolerant lines were identified compared with 9311 and showed better drought tolerance in terms of the seed setting rate and grain yield under drought stress. There was no obvious difference in the agronomic characters in comparisons of lines X646, X699, X707 with 9311. The spikelet number per panicle was significantly lower in X633 than 9311, and the weight per panicle was decreased, which resulted in a significantly lower grain yield than 9311. Moreover, the seed setting rate of each drought-resistant line was more than 80% (Fig. 2A). To evaluate the breeding valuation, two male sterile lines, the recessive genic male sterile line ABCG15 and the thermosensitive male sterile line Y58S, were used to test their combining ability. The results showed that the general combining ability (GCA) values of X705, X707 and X699 were all positive, and larger than or equal to 9311 (Fig. 2B). The combination of ABCG15 × X707 was the best compared with other combinations in terms of yield, and the grain yield reached 33.09 g per plant on average, which was 10.53% higher than the contrasting combination of ABCG15 × 9311. This combination was characterized by moderate days to heading and plant height, with larger panicles reaching 419.84 spikelets per panicle on average, but less effective panicles, a lighter 1000-grain weight, and a lower seed setting rate compared with the control. A huge panicle was clearly observed in the hybrid rice (Table 1). Other excellent combinations, including ABCG15 × X699, Y58S × X707 and Y58S × X705, showed 1.41–1.91% higher grain yield than the control.

In general, the GCA of the substitution lines X699, X705 and X707 was better than or equivalent to 9311, which had a higher seed setting rate and grain yield under drought conditions; these features are helpful for breeding high-quality and drought-tolerant hybrid rice.

QTLs for plant height and its relative

In response to drought stress, eight QTLs for plant height (PH) were detected (Table 2), which were distributed across 14 substitution segments on 7 chromosomes (Chr.1, 4, 5, 7, 8, 9 and 12); qPH-9 was detected only in lines X703 and X713. Among these QTLs, 7 showed positive additive effects (increased plant height), and the additive effect contributions ranged from 37.84 to 13.77%. In contrast, qPH-5b showed negative additive effect (decreased plant height) and contributed an additive effect of approximately 5.48%. Five QTLs for relative plant height (RPH) were detected in 10 substitution segments on 5 chromosomes (Chr. 5, 7, 8, 9 and 11); of these, qRPH-9 was detected only in X703, X707, and X709, and qRPH-5 was detected only in X682. Among these 5 QTLs, 4 showed positive additive effects, and their contributions ranged from 9.72 to 7.64%. In addition, we also found 2 substitution segments (RM4674-RM161 and RM410-RM201) in which RPH and PH QTLs were detected simultaneously.

Table 2: QTLs for plant height and relative plant height under drought stress

QTL		Chr.	SSSL	Introgression segment marker	Segment length (Mb)	Plant height		Relative plant height
						Additive effect	Additive effect contribution (%)	Additive effect	Additive effect contribution (%)
qPH-1		1	X638,X641,X649	RM297-RM302-RM319-RM5811	7.62	14.92	23.94
qPH-4		4	X672	RM518-RM3471	8.24	18.34	29.42
qPH-5a		5	X669,X678,X679	RM3348-RM274-RM480	2.8	15.08	24.20
qPH-5b	qRPH-5†	5	X682,X734	RM4674-RM39-RM3351-RM161	2.32	-3.42	5.48	-0.05	6.94
qPH-7		7	X738	RM320-RM432-RM11-RM10-RM455	11.54	8.58	13.77
	qRPH-7	7	X697,X735	RM427	0.22			0.07	9.72
qPH-8		8	X702	RM344-RM331-RM42	10.82	23.58	37.84
	qRPH-8	8	X699,X701	RM25-RM72	5.9			0.07	9.72
qPH-9	qRPH-9	9	X703,X707,X709,X713	RM410-RM257-RM6543-RM278-RM242-OSR28-RM107-RM201	4.16	17.96	28.81	0.06	7.64
	qRPH-11	11	X718,X719	RM4-RM167	5.43			0.06	8.33
qPH-12		12	X729	RM1261-RM519-RM3331	9.48	15.58	25.00

^†QTL for relative plant height

Fig. 2: A. Comparison of major agronomic characteristics between drought-tolerant lines with 9311 planted under normal conditions. B. GCA comparison of major agronomic characteristics of drought tolerance lines. ^*,^** Significant compared with 9311 at the 0.05 and 0.01 probability level, respectively, based on LSD-t tests

QTLs for days to heading and its relative days

Five QTLs for days to heading (DH) were detected in 7 substitution segments on 5 chromosomes (Chr.4, 5, 7, 9 and 10) under drought stress (Table 3). Among these QTLs, 4 loci were associated with lower DH, and the additive effect ranged from 1.92 to 1.57%. One locus (qDH-10) was associated with a higher DH contribution to the additive effect of 2.43%. Six QTLs for relative days to heading (RDH) were detected in 10 substitution segments on 6 chromosomes (Chr. 3, 4, 5, 7, 9 and 10), and their additive effect contributions ranged from 1.92 to 1.44%. Furthermore, 3 substitution segments (RM16792-RM185, RM505 and RM410-RM201) were found in which QTLs for DH and RDH were detected simultaneously.

QTLs for spikelet number per panicle

Under drought conditions, four QTLs for spikelet number per panicle (SN) were detected in 4 substitution segments on 3 chromosomes (Chr. 1, 4 and 5) (Table 4). Three QTLs were associated with a decrease in the SN value, and their additive effect contributions ranged from 9.35 to 10.96%. One QTL (qSN-4) was associated with an increase in the SN value, and it had a positive effect of approximately 17.53%. No QTLs for relative spikelet number were detected.

QTLs for the number of effective panicle and their relative number

Eighteen QTLs for the number of effective panicles (EP) were detected under drought stress and were determined to be located in 26 substitution segments on 11 chromosomes (Table 5). The results showed that all QTLs had negative additive effects, and their additive effect contributions ranged from 9.09 to 21.21%. Among the detected QTLs, qEP-7b showed the largest contribution. Ten QTLs for the relative number of effective panicles (REP) were detected in 16 substitution segments on 8 chromosomes. Nine QTLs showed negative additive effects, and their additive effect contributions ranged from 18.62 to 8.51%, with qREP-4, qREP-5, and qREP-12a providing the greatest contribution. The qREP-1alocus had a positive additive effect and contributed 11.17% to the combined additive effect. EP and REP QTLs were detected simultaneously in 9 substitution fragments.

QTLs for seed setting rate and relative seed setting rate

The seed setting rate is an important index to evaluate plant stress tolerance. Four QTLs associated with a lower seed setting than 9311 were found and were determined to be located in 5 substitution segments on 4 chromosomes (Chr.3, 5, 9 and 12) (Table 6). The contributions of these QTLs showed a negative effect, with the combined additive effect ranging from 19.42 to 18.74%. The qSSR-3 locus showed the largest contribution. Four QTLs were shown to have a negative effect on the relative seed setting rate (RSSR), and these were detected in 5 substitution fragments on 4 chromosomes. Three of these QTLs were located in the same fragment as the SSR QTLs.

Table 3: QTLs for days to heading and relative days to heading under drought stress

QTL		Chr.	SSSL	Introgression segment marker	Segment length (Mb)	Heading date		Relative days to heading
						Additive effect	Additive effect contribution (%)	Additive effect	Additive effect contribution (%)
	qRDH-3†	3	X656	RM6266-RM426-RM168	5.8			-0.02	1.92
qDH-4	qRDH-4	4	X668	RM16792-RM6314-RM185	3.63	-1.92	1.8	-0.02	1.92
	qRDH-5	5	X669,X676,X678,X679	RM3348-RM274-RM480	2.8			-0.02	1.92
qDH-5		5	X678,X679	RM26-RM31	1.56	-2.04	1.92
qDH-7	qRDH-7	7	X698	RM505	0.72	-1.66	1.57	-0.02	1.44
qDH-9	qRDH-9	9	X712,X713	RM410-RM257-RM6543-RM278-RM242-OSR28-RM107-RM201	4.16	-1.66	1.57	-0.02	1.44
	qRDH-10	10	X714	RM3451-RM333-RM496	2.12			-0.02	1.44
qDH-10		10	X716,X717	RM258-RM171	3.22	2.58	2.43

^†QTL for relative days to heading

Table 4: QTLs for spikelet number per panicle under drought stress

QTL	Chr.	SSSLs	Introgression segment marker	Segment length (Mb)	Additive effect	Additive effect contribution (%)
qSN-1a	1	X630	RM1-RM283-RM8146	1.65	-15.5	10.96
qSN-1b	1	X648	RM5497-RM443	4.01	-13.23	9.35
qSN-4	4	X672	RM3471	5.56	24.8	17.53
qSN-5	5	X678	RM26-RM31	1.56	-14	9.9

Table 5: QTLs for number of effective panicles and relative number of effective panicles under drought stress

QTL		Chr.	SSSL	Introgression segment marker	Segment length (Mb)	Number of effective panicles		Relative number of effective panicles
						Additive effect	Additive effect contribution (%)	Additive effect	Additive effect contribution (%)
qEP-1a	qREP-1a†	1	X633	RM3682-RM11356	2.24	-0.6	9.09	0.11	11.17
qEP-1b	qREP-1b	1	X638,X639,X641,X649	RM297-RM302-RM319-RM5811	7.62	-1.1	16.67	-0.12	12.23
qEP-2		2	X655	RM213-RM208-RM406-RM266-RM138	1.17	-1.1	16.67
qEP-4a		4	X668	RM16792-RM6314-RM185	3.63	-0.8	12.12
qEP-4b	qREP-4	4	X671,X672	RM2416-RM518	3.35	-1.2	18.18	-0.18	18.62
qEP-5a		5	X676,X678	RM274-RM480	1.36	-1.2	18.18
	qREP-5	5	X669,X676	RM421	2.12			-0.18	18.62
qEP-5b		5	X678,X681	RM31	0.91	-1	15.15
qEP-6		6	X695	RM340-RM412-RM345-RM141-RM494	3.51	-0.8	12.12
qEP-7a		7	X698	RM505	0.72	-0.9	13.64
qEP-7b	qREP-7	7	X738	RM320-RM432-RM11-RM10-RM455	11.54	-1.4	21.21	-0.14	14.89
qEP-8	qREP-8	8	X702	RM25-RM72	5.9	-1.2	18.18	-0.08	8.51
qEP-9		9	X712,X713	RM410-RM257-RM6543-RM278-RM242-OSR28-RM107-RM201	4.16	-1.2	18.18
qEP-10a		10	X714	RM3451-RM333-RM496	2.12	-0.85	12.88
qEP-10b	qREP-10	10	X716,X717	RM258-RM171	3.22	-1	15.15	-0.11	11.17
qEP-11a		11	X718	RM4-RM167	5.43	-1	15.15
qEP-11b	qREP-11	11	X723	RM5590-RM5857-RM21-RM1355-RM209-RM229	8.76	-0.6	9.09	-0.09	9.57
qEP-12a	qREP-12a	12	X729	RM1261-RM519	6.95	-1	15.15	-0.18	18.62
qEP-12b	qREP-12b	12	X733	RM17	0.78	-1.2	18.18	-0.12	12.23

^†QTL for relative effective panicle

Table 6: QTLs for seed setting rate and relative seed setting rate under drought stress

QTL		Chr.	SSSLs	Introgression segment marker	Segment length (Mb)	Seed setting rate		Relative seed setting rate
						Additive effect	Additive effect contribution (%)	Additive effect	Additive effect contribution (%)
qSSR-3		3	X656	RM6266-RM426-RM168	5.8	-14.98	19.42
qSSR-5	qRSSR-5^†	5	X669,X679	RM421-RM3348-RM274-RM480-RM26	5.57	-14.94	19.38	-0.09	10.46
	qRSSR-6	6	X685	RM141	0.11			-0.12	13.37
qSSR-9	qRSSR-9	9	X703	RM410-RM257-RM6543-RM278-RM242-OSR28-RM107-RM201	4.16	-14.46	18.74	-0.13	15.12
qSSR-12	qRSSR-12	12	X730	RM1261-RM519-RM3331	9.48	-14.84	19.25	-0.12	13.37

^†QTL for relative seed setting rate

Table 7: QTLs for grain yield and relative grain yield under drought stress

QTL		Chr.	SSSLs	Introgression segment marker	Segment length (Mb)	Grain yield		Relative grain yield
						Additive effect	Additive effect contribution (%)	Additive effect	Additive effect contribution (%)
	qRGY-1†	1	X633	RM3682-RM11356	2.24			0.22	30.98
qGY-4a		4	X668	RM16792-RM6314-RM185	3.63	-4.164	27.39
qGY-4b		4	X671	RM2416-RM518	3.35	-3.262	21.45
qGY-5	qRGY-5	5	X679	RM3348-RM480-RM26	3.44	-3.624	23.83	-0.18	24.65
qGY-6	qRGY-6	6	X685,X692	RM141	0.11	-3.522	23.16	-0.17	23.94
qGY-7	qRGY-7	7	X735	RM427	0.22	-4.522	29.74	-0.18	26.06
qGY-11a	qRGY-11	11	X722	RM167-RM3701	6.05	-3.342	21.98	-0.15	21.13
qGY-11b		11	X724	RM229-RM26890-RM187	6.38	-2.956	19.44
qGY-12	qRGY-12	12	X728,X729,X730	RM1261-RM519-RM3331	9.48	-4.136	27.2	-0.16	23.24

^†QTL for relative grain yield

QTLs for grain yield and relative grain yield

Compared with other indexes, grain yield was the most important comprehensive index for the evaluation of drought tolerance. We found that 8 QTLs had a negative effect on grain yield (Table 7). These QTLs were located in 11 substitution segments on 6 chromosomes, and their additive effect contributions ranged from 29.74 to 19.44%, in which qGY-7 was the locus with the largest effect. The 6 QTLs associated with relative grain yield (RGY) that were closely related to drought tolerance in 9 substitution segments on 6 chromosomes. Of these QTLs, qRGY-5, qRGY-6, qRGY-7, qRGY-11, and qRGY-12 had negative additive effects, and their contributions ranged from 26.06 to 21.13%. They were located in the same regions as 5 of the QTLs for GY. Finally, one QTL (qRGY-1) was found to have a positive effect, contributing 30.98%.

Fig. 3: Chromosomal locations of QTLs for plant height, days to heading, spikelet number per panicle, effective panicle, seed setting rate, grain yield, and relative traits under drought stress

Discussion

Breeding highly resistant varieties with a higher yield under normal cultivation, less yield loss under drought stress and wide adaptability is an important tactic for rice production to resist adversity. In this study, 25 SSSLs derived from the backbone parent 9311 of two-line hybrid rice lines were found to be more drought tolerant than the recurrent parent 9311, and their comprehensive agronomic characteristics and combining ability were very good. Among them, the excellent drought-tolerant lines X699, X705 and X707 had fine agronomic characteristics and a higher general combining ability of yield than 9311. Specifically, the combination ABCG15 × X707 had great potential to increase production, which was 10.53% higher than the control. These drought resistant lines can be used as parents for breeding two-line hybrid rice with higher yield, wider adaptability and greater drought tolerance.

Discovery of hidden alleles for drought tolerance

Present study found that - SSSLs using Nipponbare as the donor parent and 9311 as the recipient parent - plant height decreased, the number of days before heading increased and the seed setting and yield decreased under drought treatment. Compared with 9311, a majority of the SSSLs showed greater variation after exposure to drought stress. Most of the detected QTLs and relative QTLs had negative effect, meaning that the drought tolerance of the SSSLs declined compared with 9311. However, some segments were also identified that showed enhanced drought tolerance in the SSSLs, such as the RM3682-RM11356 segment on Chr.1. In this region, we found the positive effect loci qREP-1a and qRGY-1 and may indicate the presence of a drought tolerance gene in this region in the Nipponbare donor parent. The phenomenon whereby SSSLs show phenotypes that surpass those of their parents also demonstrates that many genes have cryptic recessive alleles; such alleles have been found to affect drought tolerance, water-logging, salt, cold tolerance, and disease resistance (Xu et al. 2005a; Ali et al. 2006; Lafitte et al. 2006). Moreover, hidden alleles cannot be expressed in their genetic background, where they may be subject to epistatic interference, but can be expressed in specific backcrossed populations or chromosomal introgression lines. In most cases, they are greatly influenced by the genetic background.

Of the 78 QTLs detected in this study (Fig. 3), 56 were found in the same locus or region as reported QTLs (http://www.gramene.org/). Of these 56 QTLs, 36 were found to respond to drought stress (Hemamalini et al. 2000; Babu et al. 2003; Mei et al. 2003; Lafitte et al. 2004; Liu et al. 2005; Wang et al. 2005; Xu et al. 2005b; Yue et al. 2005; Zou et al. 2005; Yue et al. 2006; Bernier et al. 2007; Cui et al. 2008; Subashri et al. 2008; Yue et al. 2008; Zhao et al. 2008). Twelve QTLs of relative traits representing drought tolerance were found at the same locus or chromosomal region as the reported drought tolerance QTLs (Champoux et al. 1995; Teng et al. 2002; Yue et al. 2005, 2006; Zhang et al. 2006). Novel QTLs first identified herein include qEP-1a, qREP-1a, and qRGY-1 in the RM3682-RM11356 region on Chr.1, qGY-4a, qGY-4b, qRDH-4, and qREP-4 in the RM16792-RM185 and RM2416-RM518 regions on Chr.4, and qGY-6, qRSSR-6, and qRGY-6 in the region flanked by RM141 on Chr.6. In addition, qREP-5, qRDH-5, qRPH-5, qRDH-7, qRPH-7, qRGY-7, qREP-8, qRPH-8, qREP-10, qRDH-10, qREP-11, and qREP-12b were first identified on chromosomes 5, 7, 8, 10, 11, and 12.

Differences in relative trait values in response to drought treatment may explain how rice plants tolerate drought (Lafitte et al. 2004; Yue et al. 2005, 2006, 2008). Notably, the newly detected QTLs qRSSR-6 and qRGY-6, both of which were found to be associated with drought tolerance, contributed the largest additive effects (13.37 and 29.94%). Moreover, these QTLs were located in the 0.11-Mb genetic region and flanked by RM141 on Chr.6. We also found that many QTLs including qREP-5, qRDH-5, qRSSR-5 and qRGY-5 in the RM3348-RM26 region, qRPH-9, qRDH-9 and qRSSR-9 in the RM410-RM201 region, and qREP-12a, qRSSR-12 and qRGY-12 in the RM1261-RM3331 region, were associated with drought resistance, as reported in the same region (Champoux et al. 1995; Teng et al. 2002; Xu et al. 2005b; Zhang et al. 2006; Bernier et al. 2007; Zhou et al. 2013).

Of the traits examined, seed setting and rice grain yield were the most sensitive to drought stress. QTLs affecting the relative seed setting rate were detected in the same four chromosomal segments (RM141, RM3348-RM26, RM410-RM201, and RM1261-RM3331), and QTLs for relative grain yield were found in three of these segments (RM141, RM3348-RM26, and RM1261-RM3331). Moreover, the contributions of these regions to the additive effect for RSSR were all more than 10%, and their additive effect contributions to RGY were all more than 20%. Finally, all of these drought-resistant QTLs were repeatedly detected in the same fragment of different SSSLs. Thus, they were found to have high reproducibility and reliability and, thereby, may be promising candidates for future selective breeding.

In the substituted segments of SSSLs, we found multiple drought-tolerant QTLs clustered in the same region of the chromosomes, and we also detected other stress resistance QTLs (e.g., for cold stress or salt stress). The QTLs qREP-5, qRDH-5, qRSSR-5, and qRGY-5 were found to be associated with drought resistance and were detected in the RM3348-RM26 region. These QTLs were found in the same region as the salt-resistant QTLs (Lin et al. 2004) and cold-resistant QTLs at the plumule stage (Zhou et al. 2013). In addition, qRPH-9, qRDH-9, and qRSSR-9 were detected in the RM410-RM201 region, which is the same area as salt resistance sensitivity QTLs (Lin et al. 2004) and cold resistance QTLs (Andaya and Mackill 2003; Zhou et al. 2013).

These findings indicate considerable genetic overlap between drought-resistant QTLs and other QTLs, including those for salt resistance, cold resistance, and other stress resistances; this phenomenon is common and has been observed in many other organisms (Hu et al. 2006; Karaba et al. 2007; Xiang et al. 2008; Huang et al. 2009). We hypothesize that the genetic overlap areas described in this report are critical stress-resistant areas; they may regulate plant responses to adversity and may jointly regulate plant stress resistance. Examples of such genes include the zinc regulation transporter gene ZIP5, the auxin-induced protein genes iaa19 and iaa18 (which are found in the RM3348-RM26 region), the auxin-induced protein gene iaa26, the ADH activity regulation gene RAD (associated with pollen fertility), and the leaf evergreen gene sgr (which is found in the RM410-RM201region). The discovery of stress-resistance QTLs may improve our understanding of plant responses to adversity, and they may prove useful for the development of new rice cultivars with improved stress tolerance.

Conclusion

In this study, seven SSSLs were identified to be more resistant to drought than 9311 in seed setting rate and grain yield. Lines X699, X705 and X707 with good comprehensive agronomic characteristics had stronger GCA for yield than 9311. A total of 78 QTLs were detected under drought stress, which were located in 30 regions of 12 chromosomes in rice. Drought-tolerance QTLs qREP4, qREP5, qREP12.1, qRSSR9, qRGY1, qRGY5, qRGY6, qRGY7, qRGY11 and qRGY12 were the main effective drought-resistant QTLs, exhibiting an additive effect contribution of more than 15%. Some drought-tolerant QTLs clustered in certain regions on chromosome 5, 6, 9 and 12, which may be the key locus for enhancing drought tolerance in plants.

Author Contributions

Conceptualization, Yong Zhou and Shigui Li; Data curation, Yinghai Wei and Chuncao Song; Formal analysis, Yong Zhou; Funding acquisition, Shigui Li, Shijun Huang and Mei Tang; Investigation, Yong Zhou, Yinghai Wei and Ying Zheng; Project administration, Shigui Li and Shijun Huang; Resources, Shigui Li; Software, Yanjie Peng; Writing original draft, Yong Zhou; Writing, review and editing, Yanjie Peng.

Acknowledgements

This work was supported by the National Science Fund for Distinguished Young Scholars of China (31025017), the Science and Technology Planning of Sichuan Province of China (2016NYZ0028).

References

Ali AJ, JL Xu, AM Ismail, BY Fu, CHM Vijaykumar, YM Gao, J Domingo, R Maghirang, SB Yu, G Gregorio, S Yanaghihara, M Cohen, B Carmen, D Mackill, ZK Li (2006). Hidden diversity for abiotic and biotic stress tolerances in the primary gene pool of rice revealed by a large backcross breeding program. Field Crops Res 97:66‒76

Ali ML, MS Pathan, J Zhang, G Bai, S Sarkarung, HT Nguyen (2000). Mapping QTLs for root traits in a recombinant inbred population from two indica ecotypes in rice. Theor Appl Genet 101:756‒766

Andaya VC, DJ Mackill (2003). QTLs conferring cold tolerance at the booting stage of rice using recombinant inbred lines from a japonica X indica cross. Theor Appl Genet 106:1084‒1090

Babu RC, BD Nguyen, V Chamarerk, P Shanmugasundaram, P Chezhian, P Jeyaprakash, SK Ganesh, A Palchamy, S Sadasivam, S Sarkarung, LJ Wade, HT Nguyen (2003). Genetic analysis of drought resistance in rice by molecular markers association between secondary traits and field performance. Crop Sci 43:1457‒1469

Bernier J, A Kumar, V Ramaiah, D Spaner, G Atlin (2007). A large-effect QTL for grain yield under reproductive-stage drought stress in upland rice. Crop Sci 47:507–516

Champoux MC, G Wang, S Sarkarung, DJ Mackill, JC O'Toole, N Huang, SR McCouch (1995). Locating genes associated with root morphology and drought avoidance in rice via linkage to molecular markers. Theor Appl Genet 90:969‒981

Cui K, J Huang, Y Xing, S Yu, C Xu, S Peng (2008). Mapping QTLs for seedling characteristics under different water supply conditions in rice (Oryza sativa). Physiol Plantarum 132:53‒68

Ebitani T, Y Takeuchi, Y Nonoue, T Yamamoto, K Takeuchi, M Yano (2005). Construction and evaluation of chromosome segment substitution lines carrying overlapping chromosome segments of indica rice cultivar ‘kasalath’in a genetic background of japonica elite cultivar ‘Koshihikari’. Breed Sci 55:65‒73

Eshed Y, D Zamir (1995). An introgression line population of Lycopersicon pennellii in the cultivated tomato enables the identification and fine mapping of yield-associated QTL. Genetics 141:1147‒1162

He FH, ZY Xi, RZ Zeng, A Talukdar, GQ Zhang (2005). Mapping of heading date QTLs in rice (Oryza sativa L.) using single segment substitution lines. Sci Agric Sin 38:1505‒1513

Hemamalini GS, HE Shashidhar, S Hittalmani (2000). Molecular marker assisted tagging of morphological and physiological traits under two contrasting moisture regimes at peak vegetative stage in rice (Oryza sativa L.). Euphytica 112:69‒78

Hu H, M Dai, J Yao, B Xiao, X Li, Q Zhang, L Xiong (2006). Overexpressing a Nam, Ataf, and Cuc (Nac) transcription factor enhances drought resistance and salt tolerance in rice. Proc Natl Acad Sci USA 103:12987‒12992

Huang XY, DY Chao, JP Gao, MZ Zhu, M Shi, HX Lin (2009). A previously unknown zinc finger protein, dst, regulates drought and salt tolerance in rice via stomatal aperture control. Genes Dev 23:1805‒1817

Kamoshita A, J Zhang, J Siopongco, S Sarkarung, H Nguyen, LJ Wade (2002). Effects of phenotyping environment on identification of quantitative trait loci for rice root morphology under anaerobic conditions. Crop Sci 42:255‒265

Karaba A, S Dixit, R Greco, A Aharoni, KR Trijatmiko, N Marsch-Martinez, A Krishnan, KN Nataraja, M Udayakumar, A Pereira (2007). Improvement of water use efficiency in rice by expression of hardy, an Arabidopsis drought and salt tolerance gene. Proc Natl Acad Sci USA 104:15270‒15275

Lafitte HR, Z Li, CHM Vijayakumar, YM Gao, Y Shi, J Xu, B Fu, SB Yu, J Ali, J Domingo, R Maghirang, R Torres, D Mackill (2006). Improvement of rice drought tolerance through backcross breeding: Evaluation of donors and selection in drought nurseries. Field Crops Res 97:77‒86

Lafitte HR, AH Price, B Courtois (2004). Yield response to water deficit in an upland rice mapping population: Associations among traits and genetic markers. Theor Appl Genet 109:1237‒1246

Lin HX, MZ Zhu, M Yano, JP Gao, ZW Liang, WA Su, XH Hu, ZH Ren, DY Chao (2004). QTLs for Na⁺ and K⁺ uptake of the shoots and roots controlling rice salt tolerance. Theor Appl Genet 108:253‒260

Liu G, Z Zhang, H Zhu, F Zhao, X Ding, R Zeng, W Li, G Zhang (2008). Detection of QTLs with additive effects and additive-by-environment interaction effects on panicle number in rice (Oryza sativa L.) with single-segment substitution lines. Theor Appl Genet 116:923‒931

Liu GM, WT Li, RZ Zeng, ZM Zhang, GQ Zhang (2004). Identification of QTLs on substituted segments in single segment substitution lines of rice. Acta Agron Sin 31:1395‒1400

Liu HY, GH Zou, GL Liu, SP Hu, MS Li, XQ Yu, HW Mei, LJ Luo (2005). Correlation analysis and QTL identification for canopy tem-perature, leaf water potential and spikelet fertility in rice under contrasting moisture regimes. Chin Sci Bull 50:317‒326

Luo LJ (2010). Breeding for water-saving and drought-resistance rice (Wdr) in China. J Exp Bot 61:3509‒3517

Luo LJ, QF Zhang (2001). The status and strategy on drought resistance of rice (Oryza sativa L.). Chin J Rice Sci 15:209‒214

McCouch SR (2008). Gene nomenclature system for rice. Rice 1:72‒84

Mei HW, LJ Luo, CS Ying, YP Wang, XQ Yu, LB Guo, AH Paterson, ZK Li (2003). Gene actions of QTLs affecting several agronomic traits resolved in a recombinant inbred rice population and two testcross populations. Theor Appl Genet 107:89‒101

Price AH, KA Steele, BJ Moore, RGW Jones (2002). Upland rice grown in soil-filled chambers and exposed to contrasting water-deficit regimes: II. Mapping quantitative trait loci for root morphology and distribution. Field Crops Res 76:25‒43

Subashri M, S Robin, KK Vinod, S Rajeswari, K Mohanasundaram, TS Raveendran (2008). Trait identification and QTL validation for reproductive stage drought resistance in rice using selective genotyping of near flowering rils. Euphytica 166:291‒305

Teng B, R Zeng, Y Wang, Z Liu, Z Zhang, H Zhu, X Ding, W Li, G Zhang (2012). Detection of allelic variation at the wx locus with single-segment substitution lines in rice (Oryza sativa L.). Mol Breed 30:583–595

Teng S, Q Qian, DL Zeng, Y Kunihiro, K Fujimoto, DN Huang, LH Zhu (2002). Analysis of gene loci and epistasis for drought tolerance in seedling stage of rice (Oryza sativa L.). Acta Genet Sin 29:235‒240

Wang XS, J Zhu, L Mansueto, R Bruskiewich (2005). Identification of candidate genes for drought stress tolerance in rice by the integration of a genetic (QTL) map with the rice genome physical map. J Zhej Univ Sci B 6:382‒288

Xi ZY, FH He, RZ Zeng, ZM Zhang, XH Ding, WT Li, GQ Zhang (2006). Development of a wide population of chromosome single-segment substitution lines in the genetic background of an elite cultivar of rice (Oryza sativa L.). Genome 49:476‒484

Xiang Y, N Tang, H Du, H Ye, L Xiong (2008). Characterization of Osbzip23 as a key player of the basic leucine zipper transcription factor family for conferring abscisic acid sensitivity and salinity and drought tolerance in rice. Plant Physiol 148:1938‒1952

Xu JL, YM Gao, BY Fu, ZK Li (2005a). Identification and screening of favorable genes from rice germplasm in backcross introgression populations. Mol Plant Breed 3:619‒628

Xu JL, HR Lafitte, YM Gao, BY Fu, R Torres, ZK Li (2005b). QTLs for drought escape and tolerance identified in a set of random introgression lines of rice. Theor Appl Genet 111:1642‒1650

Young ND, SD Tanksley (1989). RFLP analysis of the size of chromosomal segments retained around the Tm-2 locus of tomato during backcross breeding. Theor Appl Genet 77:353‒359

Yue B, W Xue, L Luo, Y Xing (2008). Identification of quantitative trait loci for four morphologic traits under water stress in rice (Oryza sativa L.). J Genet Genom 35:569‒575

Yue B, W Xue, L Xiong, X Yu, L Luo, K Cui, D Jin, Y Xing, Q Zhang (2006). Genetic basis of drought resistance at reproductive stage in rice: separation of drought tolerance from drought avoidance. Genetics 172:1213‒1228

Yue B, L Xiong, W Xue, Y Xing, L Luo, C Xu (2005). Genetic analysis for drought resistance of rice at reproductive stage in field with different types of soil. Theor Appl Genet 111:1127‒1136

Zhang X, S Zhou, Y Fu, Z Su, X Wang, C Sun (2006). Identification of a drought tolerant introgression line derived from dongxiang common wild rice (O. Rufipogon Griff.). Plant Mol Biol 62:247‒259

Zhao XQ, JL Xu, M Zhao, R Lafitte, LH Zhu, BY Fu, YM Gao, ZK Li (2008). QTLs affecting morph-physiological traits related to drought tolerance detected in overlapping introgression lines of rice (Oryza sativa L.). Plant Sci 174:618‒625

Zhou Y, XB Zhu, H Yuan, Y Zheng, P Qin, YH Wei, YP Wang, SJ Huang, SG Li (2013). Characterization of cold tolerance and identification of cold tolerance QTLs for rice single segment substitution lines at plumule and seedling stages. Chin J Rice Sci 27:381‒388

Zhu W, J Lin, D Yang, L Zhao, Y Zhang, Z Zhu, T Chen, C Wang (2009). Development of chromosome segment substitution lines derived from backcross between two sequenced rice cultivars, indica recipient 93-11 and japonica donor nipponbare. Plant Mol Biol Rep 27:126‒131

Zou GH, HW Mei, HY Liu, GL Liu, SP Hu, XQ Yu, MS Li, JH Wu, LJ Luo (2005). Grain yield responses to moisture regimes in a rice population: association among traits and genetic markers. Theor Appl Genet 112:106‒113