Introduction

Curvularia leaf spot is a significant foliar disease of maize that occurs in all maize cultivation regions (Liu et al. 2015a), but especially in north and northeast China, the main maize cultivation regions in China. This disease is caused by the fungus Curvularia lunata, and its severity depends on whether the climatic conditions are favorable for fungus development. The fungus can overwinter in the debris of diseased corn plants left on the soil surface, and conidia produced in the following spring can be spread by wind or splashing of droplets during rain events. The disease is prevalent in areas where dewy mornings are followed by hot humid afternoons and relatively cool nights. The disease occurs mainly during the maize reproductive period and damages the whole plant, which decreases quality and yield. In 1999, an outbreak of maize Curvularia leaf spot in Liaoning Province resulted in the infection of nearly 17,500 hectares of maize, which decreased maize production by 250 million kg (Li et al. 2002).

The development of maize cultivars resistant to Curvularia leaf spot through conventional breeding is one way to control the disease and ensure the security of corn production in China. However, conventional breeding of Curvularia leaf spot resistant cultivars has been difficult because of the complexity of the resistance trait. Although Curvularia leaf spot resistance is a highly heritable trait, it is controlled by many minor quantitative trait loci (QTLs) (Wang et al. 2014; Dong et al. 2015). Improvements in cultivation methods (Li and Mo 2015) and field management strategies have effectively reduced the incidence of Curvularia leaf spot in maize (Zhao et al. 2001). Although Curvularia leaf spot is now well contained in China, the combined effects of climate change, increasing areas of corn monocultures, and the narrow North Chinese resistant germplasm means that the disease still poses a serious threat to Chinese corn production.

Previous studies on the pathogenic mechanisms of C. lunata have identified and characterized some of the important genes and determined which pathways are involved in pathogenesis (Liu et al. 2016; Gao et al. 2017). However, relatively few studies have focused on the location of resistance genes in maize, and the mechanisms of maize disease resistance. Several previous studies have identified QTLs associated with resistance to Curvularia leaf spot, but research on their effects has obtained inconsistent results (Liu et al. 2009; Hou et al. 2013; Liu et al. 2015b). To date, no conserved QTL regions associated with Curvularia leaf spot resistance under different environments have been found.

In this study, to identify consistent QTL intervals related to Curvularia leaf spot resistance, we obtained phenotypic data for maize in three environments (Gongzhuling; Changchun) across two years (2015 and 2016). The results of this study lay the foundation for the identification of resistance genes and the development of markers for molecular-assisted breeding.

Materials and Methods

Plant materials and field trials

The maize cultivar Ma 664 was selected by the breeder Yi-yong Ma (Jilin Agricultural University). This cultivar is strongly resistant to C. lunata (the disease showed the lowest grade, grade I, in a field trial in 2014). The cultivar H4074 was selected by the breeder Shu-yan Guan (Jilin Agricultural University). This cultivar is highly susceptible to C. lunata (disease grade 7 in the field trial). Therefore, a mapping population of 239 F₂ individuals was derived from the cross between Ma664 and H4074. In 2015, the 239 F₂ individuals, the two parent lines, and the F₁ generation were planted in Changchun, Jilin Province. In 2016, the F_2:3 families derived from selfed F₂ individuals were planted in Gongzhuling and Changchun, Jilin Province. Hereafter, the three groups constructed above are abbreviated as follows: F2, Rep1, and Rep2. At each location, the field experiment had a randomized complete block design with three replications for each genotype. Maize was planted with row lengths of 5 m, inter-plant spacing of 25 cm, and row spacing of 60 cm. There was a protective line around each plot, and conventional field management was employed.

Trait evaluation

The F₂ population and two F_2:3 family populations (Rep1, Rep2) at the 12–13 leaf stage were inoculated with C. lunata (provided by the Jilin Academy of Agricultural Sciences). The spore concentration of the inoculum was adjusted to 10–15 spores per field of view under 100x magnification. The inoculum was sprayed onto both sides of whole maize leaves until the solution dripped from the leaves. The evaluation criteria for the disease level were those specified by Hou et al. (2013). The incidence of different disease levels is shown in Fig. 1, and the different disease levels are described in Table 1.

Phenotypic data analysis

Pairwise comparisons of means of the parents’ disease grades were tested for significance with t-tests implemented in SPSS 25 software (http://www.ibm.com/legal/copytrade.shtml). The phenotypic data of the F₂ population and populations of the two F_2:3 families were tested by Descriptive Statistics of S.P.S.S. 25 for normal distribution, where absolute values of kurtosis and skewness of less than 1 confirmed normal distribution.

Molecular data collection and linkage map construction

The DNA was extracted from the plant materials as described by Mu et al. (2010). Simple sequence repeat (SSR) markers covering the entire genome were selected from the maize genome database (http://www.maizegdb.org/) and screened to identify those that were polymorphic between the two parents. These markers were used to genotype the mapping population (F₂ population). Marker linkage analysis and construction of linkage maps were conducted using Ici Mapping 4.0. A limit of detection (LOD) threshold of 2.5 was used to assign markers to the same linkage group. The observed frequencies at each marker were tested against the expected Mendelian segregation ratio of 1:2:1 using a K² test for goodness of fit.

QTL analysis

The DNA extracted from members of the F₂ generation was subjected to PCR amplification and capillary electrophoresis detection using the selected markers. This procedure was used to genotype all members of the F₂ generation. The phenotypic data for the F_2:3 populations and the SSR marker molecular linkage map information were used to identify QTLs using IciMapping 4.0. We used the inclusive composite interval mapping (ICIM) method for single-environment QTL mapping of traits. A QTL was considered to be significantly correlated with resistance to Curvularia Leaf Spot when the LOD score was greater than 2.5. The genetic effects and phenotype contribution rates were analyzed. Each QTL was scored according to its dominance ratio (DR; DR = | d | / | a |). Thus, when DR < 0.2, the QTL was additive; 0.2 < DR < 0.8, the QTL was partially dominant; 0.8 < DR < 1.2, the QTL was dominant; and DR > 1.2, the QTL was super-dominant. Epistatic effects of QTLs were analyzed using the ICIM with epistatic interactions (ICIM-EPI) method.

Results

Genotype analysis and construction of maize genetic linkage map

We tested 650 SSR primer pairs and found that 150 pairs (23.08%) were sufficiently polymorphic between the two parents. The genotype data for the F₂ population were analyzed using SPSS software. The separation of 74.7% SSR markers in the F₂ population was consistent with a 1:2:1 segregation ratio. Of the 150 SSR polymorphic markers, 38 (25.3%) showed segregation distortion. Of those, 20 markers (52.6%) were biased to the female parent, 11 (28.9%) were biased to the male parent, and 7 (18.4%) were biased to the F1 generation.

The markers were unevenly distributed among linkage groups, and the distances between markers ranged from 6.25 and 29.70 cM, with no large gaps Fig. 4. The relative order of the markers was consistent with that on the genetic linkage map at the Maize GDB database, which indicates that our data conformed to QTL positioning requirements. The lengths, numbers, and distances between SSR markers are shown in Table 2.

Table 1: Evaluation criteria of Curvularia Leaf Spot disease grades

Disease grade	Description of corresponding disease grade	Resistance level
1	No lesions or only sporadic lesions, area of lesions accounts for less than 5% of leaf area	Highly resistant (HR)
3	A few lesions, area of the lesions accounts for 6%–10% of the leaf area.	Resistant (R)
5	More lesions, accounting for 11%–30% of the leaf area	Middle resistant (MR)
7	Many lesions, some connected, accounting for 31%–50% of the leaf area.	Susceptible (S)
9	Whole plant is covered with disease spots, lesions are connected and account for more than 50% of the leaf area. Leaves die in late stages of the disease.	Highly susceptible (HS)

Table 2: Lengths, numbers, and distances between SSR markers for each maize linkage group

	Linkage groups										Average	Total
	1	2	3	4	5	6	7	8	9	10
Chain length (CM)	243.0	199.5	180.5	187.5	175.9	131.2	162.0	144.5	139.6	140.0	170.4	1703.7
Number of linkage groups	15	13	10	11	13	10	11	10	10	9	11.2	112
Average spacing (CM)	16.2	15.3	18.1	17.0	13.5	13.1	14.7	14.5	14.0	15.6	15.2

Fig. 1: Disease grades of Curvularia Leaf Spot in maize.

Phenotype analysis

The incidence of resistance was significantly different between the two parental populations, as determined by independent samples t-tests (Table 3). This confirmed that these parents were suitable for the construction of populations for mapping QTLs related to resistance to Curvularia leaf spot.

The distribution of disease grades in the F2 and F_2:3 families is shown in Table 4. The disease grades were distributed evenly (1, 3, 5, 7 and 9), indicative of a continuous distribution of the resistance phenotype. This confirmed that the resistance of maize to Curvularia leaf spot is a quantitative trait controlled by multiple genes.

A descriptive statistical analysis of the F₂ and two F_2:3 populations was conducted. These analyses (Table 4 and 5) showed that 90% of the F₂ population had a disease level between those of the parents [average value of 5.52, skewness of 0.155, peak of 0.062, and standard deviation (SD) of 1.78]. The overall disease incidence in the population was consistent with a normal distribution (Fig. 2A). For the F_2:3 family Rep1 group, 98% of the population had a disease level between those of the parents (average 4.85, skewness 0.062, kurtosis 0.614, and SD 1.49). The overall disease incidence in the population showed a normal distribution (Fig. 2B). For the F_2:3 family Rep2 group, 98% of the population had a disease level between those of the parents (average 4.90, skewness 0.143, kurtosis −0.037, and SD 1.50). The overall disease incidence in the population showed a normal distribution (Fig. 2C). These results showed that resistance to Curvularia Leaf Spot in maize is a quantitative trait under polygene control, and confirmed that the population was suitable for QTL mapping.

QTL positioning

Table 3: T-tests of disease incidence levels between parents

Year	Levene’s test for equality of variances		T-test for equality of means
	F	Sig	T	Sig
2015	0.924	0.345	-19.008	0.000
2016	7.338	0.011	-16.310	0.000

Table 4: Disease grade distribution in maize parents, and F₁, F₂ and F₂-derived populations

Year	Generation	Disease grade					Total plants	Mean ± SE
		1	3	5	7	9
2015	P1	12	3	0	0	0	15	1.40 ± 0.214
	P2	0	0	3	13	0	15	6.73 ± 0.182
	F1	0	5	10	0	0	15	4.33 ± 0.252
	F2	5	32	122	56	24	239	5.52 ± 0.115
2016	P1	13	2	0	0	0	15	1.27 ± 0.182
	P2	0	0	5	10	0	15	6.33 ± 0.252
	F1	0	6	9	0	0	15	4.20 ± 0.262
	F2:3 rep1	6	53	138	37	5	239	4.85 ± 0.966
	F2:3 rep2	3	60	126	46	4	239	4.90 ± 0.969

Table 5: Descriptive statistics of F₂ population and its two derived F_2:3 families

Segregation population	Mean ± SE	Standard deviation	Skewness ± SE	Kurtosis ± SE
F2	5.52 ± 0.115	1.78	0.155 ± 0.157	0.062 ± 0.314
F2:3rep1	4.85 ± 0.966	1.49	0.062 ± 0.157	0.614 ± 0.314
F2:3rep2	4.90 ± 0.969	1.50	0.143 ± 0.157	-0.037 ± 0.314

Table 6: Results of QTL mapping for resistance to Curvularia lunata (ICIM LOD>2.5)

Population	Name	Bin	Position	Distance to markers	Marker interval		LOD score	PVE(%)	Additive effect	Dominant effect	Gene action
					Left	Right
F2	qCLS1.10	1.10	212.00	0.22-11.75	bnlg1347a	umc1862	3.65	4.87	-0.26	-0.68	OD
	qCLS3.08	3.08	130.00	26.86-2.84	umc2269	bnlg1108	8.78	14.00	-0.92	0.24	PD
	qCLS8.01	8.01/8.02	26.00	7.12-11.29	umc1483	umc1913	4.29	7.05	-0.67	-0.04	A
	qCLS10.04	10.04	75.00	0.15-19.77	mmp121	umc1506	10.48	14.42	-0.90	-0.40	PD
Rep1	qCLS1.02	1.02	29.00	13.02-3.96	bnlg1014	umc1467	4.64	8.90	-0.58	0.34	PD
	qCLS5.07	5.07	139.00	0.28-10.83	umc1375	umc2013	4.40	6.69	-0.55	0.11	A
	qCLS7.01	7.01/7.02	38.00	5.73-10.43	umc1270	bnlg1247	2.51	4.55	0.45	-0.03	A
	qCLS10.04	10.04	75.00	0.15-19.77	mmp121	umc1506	7.53	11.60	-0.68	-0.29	PD
Rep2	qCLS2.10	2.10	193.00	7.54-6.53	bnlg1893	umc2214	3.97	5.21	-0.48	-0.06	A
	qCLS5.03	5.03	62.00	0.44-15.54	umc1705	bnlg1902	8.56	10.62	-0.68	-0.19	PD
	qCLS6.05	6.05	72.00	2.05-16.45	umc2141	bnlg1732	3.68	4.60	-0.17	-0.58	OD
	qCLS9.01	9.01	7.00	7.00-3.11	umc1957	umc1867	4.67	6.03	-0.45	0.36	PD
	qCLS10.04	10.04	75.00	0.15-19.77	mmp121	umc1506	14.75	19.00	-0.90	-0.21	PD

Note: Bin indicates the corresponding Bin interval on the MaizeGDB map of the QTL; Position indicates QTL position on chromosome (CM); "a-b" in the “Distance to markers” column indicates that the distances of the QTL locus to left and right marks is a and b respectively(CM) ;“Left” and “Right” of the Marker interval indicates the left and right markers of the QTL interval; LOD indicates log10 of the likelihood odds ratio; PVE（%）represents the phenotypic variance percentage that owe to the corresponding QTL.A, D, PD and OD in the “Gene action” column represent additive effects, dominant effects, partial dominance effects and super-dominant effects, respective

Fig. 2: Disease grade distribution in F₂ population and its F_2:3 families.

Four QTL loci were detected by genome-wide scanning of the F₂ population, based on a combination of genotypic and morbidity data. Single QTLs were located on chromosomes 1, 3, 8, and 10 (Fig. 3). The whole genome was analyzed by ICIM-EPI with LOD > 2.5 as the unit. Epistatic interactions among QTLs were not detected. The names, numbers, and effects of QTLs related to leaf spot resistance in the F₂ population are shown in Table 6. The detected QTLs accounted for 40.33% of the phenotypic variation in resistance. The QTLs in the intervals umc2269–bnlg1108 (on chromosome 3) and mmp121–umc1506 (on chromosome 10) made the largest contributions to phenotypic variance (14.00% and 14.42%, respectively). The QTLs associated with the bnlg1347a–umc1862 interval accounted for 8.77% of phenotypic variation (LOD 3.65). The QTL located on umc1843–umc1913 on chromosome 8 accounted for 7.05% of phenotypic variation (LOD 4.29). The additive effect of each QTL was negative, which indicates that all the QTLs related to low disease incidence were from the resistant parent Ma 664. According to the DR ratio proposed by Stuber et al. (1992), qCLS1.10 (DR = 2.62), qCLS3.08 (DR = 0.26), qCLS8.01 (DR = 0.06) and qCLS10. 04 (DR = 0.44) were super-dominant, partially dominant, additive, and partially dominant, respectively.

Fig. 3: QTL mapping results for each maize population

Fig. 4: Maize genetic linkage map showing distribution of QTLs for resistance to Curvularia Leaf Spot in maize

Four QTL loci (Fig. 3) located on chromosomes 1, 5, 7, and 10 were detected in the F_2:3 Rep1 population and accounted for 31.74% of the phenotypic variation in resistance. As shown in Table 6, except for qCLS7.01 (in the umc1270–bnlg1247 interval on chromosome 7), the other QTLs had negative additive effects, which indicates that they were derived from the disease-resistant parent Ma 664, while qCLS7.01 was inherited from the female parent H4074. The QTLs qCLS10.04 (DR = 0.43) and qCLS1.02 (DR = 0.59) were partially dominant, and accounted for 8.9% and 11.60% of the phenotypic variation in resistance, respectively. The qCLS7.01 (DR = 0.07) locus located at umc1270–umc2013 in the umc1375–umc2013 interval and the umc1270–bnlg1247 interval on chromosome 7 showed additive effects, and accounted for 6.69% and 4.55% of the phenotypic variation in resistance, respectively. Among the QTLs, qCLS10.04 on chromosome 10 was detected in the F_2:3 Rep1 and F₂ populations, and was partially dominant. The IGEM-EPI algorithm was used to analyze the epistatic interactions of the whole genome, and no epistatic interactions were detected among these QTLs.

Five QTLs (on chromosomes 2, 5, 6, 9, and 10) were detected in the F_2:3 Rep2 population (Fig. 3) and accounted for 45.46% of the phenotypic variation in resistance. As shown in Table 6, all of the QTLs had negative additive effects, which indicates that the QTLs from the male parent reduced disease incidence and improved resistance. The qS2.10 locus on chromosome 2 was an additive effect QTL (DR = 0.13) located at bnlg1893–umc2214 (LOD 3.97) that accounted for 5.21% of the phenotypic variation. The qCLS6.05 locus on chromosome 6 was super-dominant (DR = 3.41), was located at umc2141–bnlg1732 (LOD 3.68), and accounted for 4.60% of the phenotypic variation. There are still major obstacles for using super-dominant QTLs in crop breeding. Therefore, this QTL needs to be further investigated. The QTLs qCLS9.01 (DR = 0.8) and qCLS10.04 (DR = 0.23) on chromosomes 9 and 10, respectively, were partially dominant. They were located at umc1957–umc1867 and mmp121–umc1506, respectively (LOD 4.67 and 14.75, respectively) and accounted for 6.03% and 19.00% of the phenotypic variation, respectively. qCLS10.04 on chromosome 10 was detected in the F₂ and F_2:3 Rep1 populations and had the same genetic effect in both populations. It had high LOD values and phenotypic variation contribution rates; therefore, we consider that this is a stable QTL related to resistance to Curvularia Leaf Spot in maize. This QTL can serve as the starting point to identify candidate resistance genes through fine positioning mapping. The IGEM-EPI algorithm was used to analyze the epistatic interactions of the whole genome, and no epistatic interactions were detected among the analyzed QTLs.

Discussion

A consistent environment is required to accurately assess the potential of plant genotypes to resist the onset and progress of Curvularia leaf spot, and to determine the magnitude of the genetic factors that contribute to resistance. This is because the development of Curvularia leaf spot is extremely sensitive to environmental conditions. In this QTL mapping study, we obtained phenotypic data for Curvularia leaf spot resistance of maize in two years (2015 and 2016) at sites in Changchun and Gongzhuling. In both years, the summer was humid and relatively hot. These environmental conditions made it possible to assess the level of Curvularia leaf spot resistance in the segregating populations.

We conducted interval mapping at the LOD threshold of 2.5, and detected 11 QTLs related to resistance to Curvularia leaf spot. Of the three main QTLs (at Bin3.08, Bin5.03, and Bin10.04), the QTL on chromosome 10 was consistently detected in three environments. In another study, a stable QTL for resistance to Curvularia leaf spot was detected at the same site (Bin10.04) in analyses of an F_2:3 family population of Shen 137×Huangzao 4 (Hou et al. 2013). The results of that study and our study indicate that Bin10.04 on chromosome 10 is a stable main QTL for resistance to Curvularia leaf spot.

The additive and dominant effects of QTLs can differ among various genetic backgrounds and/or among the same materials in different years. The additive and dominant effects of the consistent QTL located in Bin10.04 differed between the two years and the three environments, but this QTL was inherited dominantly, which is consistent with the findings of Hou et al. (2013). All of the QTLs detected in this study had different additive and dominant effects, but were predominantly additive and partially dominant. Zhao et al. (2002) studied the inheritance of resistance to Curvularia leaf spot using the ADAA model and found that the resistance of maize was mainly additive and dominant.

When we searched the Maize-GDB database, we did not find any Curvularia leaf spot resistance-related QTLs in the marker interval corresponding to the QTL loci located in this study. However, we found that qCLS1.10 (bnlg1347a–umc1862), located in this study, is located in a sugarcane borer resistance QTL region (bnl8.29a–umc106a), and the umc2269–bnlg1108 interval of qCLS3.08 partially overlaps with igc3b–umc63a, a QTL associated with resistance to the European corn borer. In addition, qCLS5.03’s umc1705–bnlg1902 interval contains a gray leaf spot resistance-related QTL (near umc43). Thus, this area may represent a large QTL marker interval that includes many different QTL or the loci targeted in this study, which may be multi-effect QTL.

In this study, the F₂ population and F_2:3 families were used as locating groups, and F₂ was used as the mapping population. The phenotypic values of several individuals were substituted for the F₂ representative values, thereby reducing the effects of environmental factors on plant traits and enabling the repeated trial of multiple points. Hou et al. (2013) noted that self-crossing of F₂ individuals yields F_2:3 families with reduced heterozygous genotypes, resulting in low estimates of QTL-associated dominant effects. Here, we compared the genetic effects of the consistent QTL in Bin10.04 in the F₂ population and two F_2:3 families, and we reached the same conclusion. In studies on maize QTLs, the F_2:3 phenotypic mean has often been used instead of the F₂ phenotype to account for the shortcomings of the F₂ generation (Lu et al. 2002; Park et al. 2013; Hou et al. 2015; Liu et al. 2016). Those studies identified genetic loci associated with important agronomic traits of maize using F_2:3 families as the locating populations and successfully mapped stable QTLs. Here, we located QTLs for Curvularia leaf spot resistance in maize that stably exists in different environments. This confirmed the feasibility of using F_2:3 families as QTL-locating populations.

Segregation distortion skews the genotypic frequencies from their Mendelian expectations (Lu et al. 2002). In this study, 38 (25.3%) of 150 of polymorphic markers showed segregation distortion in the F₂ population. Of these 38 markers, 20 (52.6%) were biased towards the female parent, 11 (28.9%) were biased towards the male parent, and seven (18.4%) were biased toward the F₁. This is consistent with the results reported by Lu et al. (2002). Since Mangelsdorf and Jones (1926) first reported segregation distortion in maize, many researchers have detected this phenomenon when studying maize linkage maps (Bentolila et al. 1992; Gardiner et al. 1993; Murigneux et al. 1993; Pereira and Lee 1995). Liu and Yang (2015) constructed a maize genetic linkage map and found that 12 (33.3%) out of 31 polymorphic markers showed segregation distortion; one (8.33%) was biased to the male parent, two (16.67%) were biased towards the heterozygote, and five (41.67%) were unbiased. There are many reasons for the segregation distortion of molecular markers. Lu et al. (2002) studied the segregation of SSR molecular markers in maize and found that most chromosomes have functional genes that cause segregation distortion of markers. This affects the normal separation of alleles and determines the direction of segregation distortion. The ratios of marker segregation distortion are positively correlated with the generation of the population. This is because there is unequal selection between male and female gametes in the process of meiosis and combination of gametes. Molecular markers that show segregation distortion are located in particular regions of chromosomes. In this study, markers on all chromosomes showed segregation distortion, and those showing segregation distortion were located in certain hot spots on chromosomes. The segregation distortion of molecular markers leads to inconsistencies between the marker recombination rate and the genetic distance of the marker, which reduces the accuracy of the genetic linkage map. Some studies have reported that segregation distortion of molecular markers has little effect on the location of loci (Hackett and Broadfoot 2003; Zhang et al. 2010). However, we found that the segregation distortion of markers introduced errors into the genetic linkage map, so that the relative map position of the molecular marker was different from that in Maize-GDB. Therefore, when this occurred, the marker was removed from the genetic linkage map to reduce mapping errors.

The construction of a high-density and precise genetic linkage map is a prerequisite for the accurate detection of QTLs. In this study, we detected stable differences between parents and the separation of F₂ populations in accordance with the 1:2:1 genotypic ratio of 112 SSR markers, which were divided into 10 linkage groups. A linkage map of maize molecular markers was constructed. The relative sequences of the markers were consistent with the genetic linkage map at the Maize-GDB database. The total length of the map was 1,703.7 cM and the average density was 15.2 cM, in accordance with QTL-locating requirements. With the application of single nucleotide (SNP) molecular marker technology in maize gene mapping, more precise genetic linkage maps have been constructed. These maps have made it easier to fine-map maize QTLs (Zou and Song 2003; Pan et al. 2011; Warburton et al. 2015; Song et al. 2017).

Molecular marker-assisted screening of stress-resistant crops is extremely important and is a very effective breeding method. The QTL detected in Bin10.04 in this study has potential uses in marker-assisted selection in breeding, but further research is required to identify the gene(s) responsible for resistance. It can be risky to use QTL-linked markers for breeding when their effects on other agronomic traits are unknown, or when the mechanism of the genetic effect is unclear (Tanksley and Hewitt 1988). Epistatic effects can also influence molecular marker-assisted selection. Vasal et al. (1970) found that epistatic interactions, especially those among dominant genes, are major genetic effects related to leaf spot resistance. Epistatic interactions among QTLs were not detected in this study. Appropriate combinations of populations and genetic mating design are essential to accurately detect the presence of epistatic effects among QTLs. Therefore, further research to analyze the genetic effects of QTL at Bin10.04 and the interaction with the environment are of great significance for the molecular marker-assisted selection of maize lines resistant to Curvularia leaf spot.

Conclusion

We detected three main QTLs, one each on chromosomes 3, 5, and 10 (Bin3.08, Bin5.03, and Bin10.04). The QTL on chromosome 10 was detected in three environments. This locus may contain a stable resistance gene. Further research is required to identify and characterize this gene.

Acknowledgments

We thank Wang Piwu (Biotechnology Center, Jilin Agricultural University) and Prof. Guan for support and help. This research was funded by Jilin Province Science and Technology Research Projects (no. 20170204005NY), the Special Fund Project of Provincial Grain Production and Development in Jilin Province (no. 2015001) and Five-Year Scientific Research Project of the Education Department of Jilin Province. We thank Jennifer Smith, PhD, from Liwen Bianji, Edanz Group China (www.liwenbianji.cn/ac), for editing the English text of a draft of this manuscript.

Author Contributions

Jian-Bo Fei and Zhao-Xu Dong planted maize populations; Zhi-Bo Liu and Dong-Liang Jin collected phenotypic data; Jing Qu, Si-Yan Liu, Yi-yong Ma, and Shu-Yan Guan obtained genotype data, Jian-bo Fei wrote the manuscript with contributions from Zhao-Xu.

References

Bentolila S, T Hardy, C Guitton, G Freyssinet (1992). Comparative genetic analyses of F2 plants and another culture derived plants of maize. Genome 35:575‒582

Dong HY, YM Xue, LJ Wang, KJ Liu, XD Xu (2015). Resistance identification and evaluation of introduced maize germplasms to 3 kinds of maize leaf spot. Plant Prot 41:167‒170

Gao SG, NI Xuan, YY Li, KH Fu, CJ Yu, JX Gao, M Wang, YQ Li, J Chen (2017). Sod, gene of Curvularia lunata, is associated with the virulence in maize leaf. J Integr Agric 16:874‒883

Gardiner JM, EH Coe, S Melia-Hancock, DA Hoisington, S Chao (1993). Development of a core RFLP map in maize using an immortalized F2 population. Genetics 134:917‒930

Hackett CA, LB Broadfoot (2003). Effects of genotyping errors, missing values and segregation distortion in molecular marker data on the construction of linkage maps. Heredity 90:33‒38

Hou J, YX Xing, Y Zhang, YF Tao, ML Xu (2013). Identification of quantitative trait loci for resistance to curvularia leaf spot of maize. Maydica 58:266‒273

Hou X, Y Liu, Q Xiao, B Wei, X Zhang, Y Gu, YB Wang, J Chen, YF Hu, HM Liu, J Zhang, YB Huang (2015). Genetic analysis for canopy architecture in an F_2:3 population derived from two-type foundation parents across multi-environments. Euphytica 205:421‒440

Li JT, SX Mo (2015). Effects of two-plants orientation cultivation pattern on the maize yield and occurrence and epidemic of Curvularia leaf spot. J Maize Sci 2015:137‒140

Li Y, FC Dai, RL Jing, TY Wang, JY Du, JZ Jia (2002). QTL analysis of resistance to Curvularia lunata in maize. Chin Agric Sci 2002:1221‒1227

Liu CN, JL Dai, LJ Wu, XP Xing, HL Li, HX Yuan (2015a). Genetic analysis of resistance to Curvularia leaf spot of maize (Zea mays L.) inbred line Qi319. J Henan Agric Univ 49:234‒238

Liu T, J Hu, Y Zuo, Y Jin, J Hou (2015b). Identification of microRNA-like RNAs from Curvularia lunata associated with maize leaf spot by bioinformation analysis and deep sequencing. Mol Genet Genomics 291:587‒596

Liu LH, YH Pei, P Mu, EY Zhang, XM Guo, HG Yan, XY Song (2009). SSR linkage map construction and QTL analysis for Curvularia leaf spot disease in maize. J North Chin Agric 24:27‒31

Liu YH, Q Yi, XB Hou, XG Zhang, JJ Zhang, MH Liu, YF Hu, YB Huang (2016). Comparative quantitative trait locus mapping of maize flowering-related traits in an F_2:3 and recombinant inbred line population. Genet Mol Res 15; Article gmr8465

Liu ZJ, J Yang (2015). Occurrence and control of curvularia leaf spot in corn. N Agric 2015:28‒28

Lu H, J Romero-Severson, R Bernardo (2002). Chromosomal regions associated with segregation distortion in maize. Theor Appl Genet 105:622‒628

Mangelsdorf PC, DF Jones (1926). The expression of Mendelian factors in the gametophyte of maize. Genetics 11:423–455

Mu CH, FJ Zhang, WC Li, Q Sun, ZH Ding, L Wang, ZD Meng (2010). A method of genomic DNA extraction of maize. J Maize Sci 2010:170‒172

Murigneux A, S Baud, M Beckert (1993). Molecular and morphological evaluation of doubled-haploid lines in maize. 2. Comparison with single-seed-descent lines. Theor Appl Genet 87:278‒287

Pan Q, F Ali, X Yang, J Li, J Yan (2011). Exploring the genetic characteristics of two recombinant inbred line populations via high-density SNP markers in maize. PLoS One 7; Article e52777

Park KJ, KJ Sa, HJ Koh, JK Lee (2013). QTL analysis for eating quality-related traits in an F_2:3 population derived from waxy corn × sweet corn cross. Breed Sci 63:325–332

Pereira MG, M Lee (1995). Identification of genomic regions affecting plant height in sorghum and maize. Theor Appl Genet 90:380‒388

Song W, Z Shi, J Xing, MX Duan, AG Su, CH Li, RY Zhang, YX Zhao, MJ Luo, JD Wang, JR Zhang (2017). Molecular mapping of quantitative trait loci for grain moisture at harvest in maize. Plant Breed 136:28–32

Stuber CW, SE Lincoln, DW Wolff, T Helentjaris, ES Lander (1992). Identification of genetic factors contributing to heterosis in a hybrid from two elite maize inbred lines using molecular markers. Genetics 132:823‒839

Tanksley SD, J Hewitt (1988). Use of molecular markers in breeding for soluble solids content in tomato - a re-examination. Theor Appl Genet 75:811‒823

Vasal SK, CL Moore, P Udom (1970). Inheritance of resistance to Curvularia Leaf Spot in maize. Sabrao Newslett 2:81‒89

Wang LJ, HY Dong, L Kejie, L Liyun, X Xiude (2014). Identification and evaluation of maize germplasms resistance to Curvularia leaf spot. Crops 2014:72‒75

Warburton ML, JD Tang, GL Windham, LK Hawkins, SC Murray, W Xu, D Boykin, A Perkins, WP Williams (2015). Genome–wide association mapping of Aspergillus flavus and aflatoxin accumulation resistance in maize. Crop Sci 55:1857–1867

Zhang LY, SQ Wang, HH Li, QM Deng, AP Zheng, SC Li, P Li, ZL Li, JK Wang (2010). Effects of missing marker and segregation distortion on QTL mapping in F2 populations. Theor Appl Genet 121:1071‒1082

Zhao BX, XC Yu, XL Zhao, ZM Liu (2001). The causal factors and integrated control techniques of Curvularia leaf spot in Corn. Shandong Agric Sci 2001:28‒28

Zhao J, GY Wang, J Hu, XH Zhang, JR Dai (2002). Analysis of the ADAA genetic model for Curvularia lunata leaf spot resistance. Crops 28:127‒130

Zou YP, GE Song (2003). A novel molecular marker–SNPs and its application. Chin Biodivers 11:370‒382